Comparative assessment of efficacy and safety of approved oral therapies for overactive bladder: a systematic review and network meta-analysis

ABSTRACT Purpose: To compare the effectiveness and safety of marketed oral drugs for overactive bladder based on a systematic review and network meta-analysis approach. Methods: Pubmed, Embase, Web of Science, and the Cochrane Register of Clinical Trials databases were systematically searched. The search time frame was from database creation to June 2, 2022. Randomized controlled double-blind trials of oral medication for overactive bladder were screened against the protocol's entry criteria. Trials were evaluated for quality using the Cochrane Risk of Bias Assessment Tool, and data were statistically analyzed using Stata 16.0 software. Result: A total of 60 randomized controlled double-blind clinical trials were included involving 50,333 subjects. Solifenacin 10mg was the most effective in mean daily micturitions and incontinence episodes, solifenacin 5/10mg in mean daily urinary urgency episodes and nocturia episodes, fesoterodine 8mg in urgency incontinence episodes/d and oxybutynin 5mg in voided volume/micturition. In terms of safety, solifenacin 5mg, ER-tolterodine 4mg, mirabegron, vibegron and ER-oxybutynin 10mg all showed a better incidence of dry mouth, fesoterodine 4mg, ER-oxybutynin 10mg, tolterodine 2mg, and vibegron in the incidence of constipation. Compared to placebo, imidafenacin 0.1mg showed a significantly increased incidence in hypertension, solifenacin 10mg in urinary tract infection, fesoterodine 4/8mg and darifenacin 15mg in headache. Conclusion: Solifenacin showed better efficacy. For safety, most anticholinergic drugs were more likely to cause dry mouth and constipation, lower doses were better tolerated. The choice of drugs should be tailored to the patient's specific situation to find the best balance between efficacy and safety.


INTRODUCTION
Overactive bladder (OAB) consists of four closely related symptoms: urgency, frequency, urge urinary incontinence (UUI) and nocturia, which have no significant impact on the patient's life safety but seriously reduce the quality of life.Studies have shown (1) that OAB can have varying degrees of impact on six aspects of daily life: recreational life, psychological problems, isolation, sexual desire, and work efficiency, causing a heavy economic burden on patients and society.The prevalence of OAB is high, ranging from 7% to 27% in men and 9% to 43% in women, and the prevalence of OAB increases with age (2,3).However, the pathophysiological mechanisms involved in the symptoms of OAB syndrome are varied and treatment is difficult (4).For this reason, more and more scholars have been conducting research on the pathogenesis of OAB from different perspectives in recent years and are constantly exploring new treatments for OAB.Treatment options for OAB are divided by "lines of therapy" based on levels of invasiveness.Lifestyle modification and pelvic floor physical therapy are the tenets of the first line of therapy.Second line therapy consists of drug therapy with anticholinergics and/or beta-3 agonists.Third line therapies include intravesical botulinum toxin injection, sacral neuromodulation, and percutaneous tibial nerve stimulation (5,6).
For decades, antimuscarinics such as tolterodine (TOL) and solifenacin (SOL) have been the main pharmacological treatment for OAB, but their lack of bladder specificity has led to a high incidence of adverse events such as dry mouth and constipation, ultimately limiting their effectiveness.In recent years, β3-adrenoceptor agonists, which are highly selective, have been developed as a potential treatment for OAB.Pharmacological assays have shown that β3-adrenoceptor agonists participate in beta adrenergic-mediated bladder relaxation, thus exerting their effect (5).They have been shown to be effective and well tolerated (7,8).
Different treatment modalities have their advantages and limitations, and it is essential to choose the right treatment modality for the specific patient in clinical practice.The wide choice of drugs available for OAB treatment and the lack of head-to-head clinical trials between drugs has led to controversy over the best drug choice.Given that one previously published study (9) had too many drug doses (including unapproved doses) grouped together, and the outcome indicators were not combined in a reasonable manner, the potential for bias is too high and the robustness of the final study results is questionable.Therefore, this study proposes to conduct a precise network meta-analysis of approved oral drugs, including only oral drugs with approved dosages and only outcome indicators with the same observation period, in order to reduce the heterogeneity of the introduced studies and provide a basis for the selection of therapeutic drugs in clinical practice.

MATERIALS AND METHODS
The software involved in this study included EndNote X8 (literature management and article writing) (Thomson Research Soft), Excel 2019 (data extraction and collation) (Microsoft Office), Review Manager 5.3 (methodological quality evaluation) (The Cochrane Collaboration, Copenhagen), and Stata 16.0 (network meta--analysis [NMA], heterogeneity assessment and inconsistency testing, surface under the cumulative ranking curve [SUCRA] plots) (Stata Corporation).The study was written according to the NMA extension for Priority Reporting Entry for Systematic Evaluation and Meta-Analysis (PRIS-MA).This study is registered with PR0SPERO (registration number CRD42021233959).

Search strategies
Two reviewers searched independently in the following database: PubMed, Embase, Web of Science and Cochrane Library.Both mesh terms and free terms were used in the search.Details of search strategies are provided in Supplementary Table-1 (see Page 1).

Inclusion criteria
(1) Study population: patients ≥18 years of age with a diagnosis of OAB according to symptoms or urodynamic studies.
(2) Intervention: any drug approved for the treatment of OAB, or placebo as control, or ano-ther drug for the treatment of OAB as control.

Exclusion criteria
Trials without any access to full text (eg, conference abstracts, etc.), with incomplete data, lack of relevant outcome indicators, data not publicly available and duplicate publications were excluded.Studies with non-oral antimuscarinic or intravesical administrations were also excluded.

Literature screening and data extraction
(1) Literature Screening: the literature was screened using EndNote X8 software to electronically check the literature retrieved from the systematic search and the manual search to eliminate duplicate literature.Then, two investigators independently read the titles and abstracts of the literature to exclude those that did not meet the inclusion criteria.After that, the remaining literature was read further in full to exclude those that did not meet the inclusion criteria, and the reasons for exclusion were recorded.Finally, both sides cross-checked the included literature and jointly decided on the inclusion of the literature, and in case of disagreement, a third investigator was consulted to decide on the inclusion of the literature.
(2) Data extraction: data extraction was performed using Excel 2019 software, which included: authors and year of publication, sample size, interventions, baseline characteristics of the study population, and outcome indicators of the literature.Two researchers worked independently and discussed and resolved any disagreements or consulted a third researcher to decide.If incomplete information or disagreements were encountered in the literature study, the authors of the literature could be contacted for information.

Methodological quality evaluation
The risk of bias was assessed in the included literature using the Cochrane Risk of Bias Assessment Tool (10) in Review Manager 5.3 software, including seven aspects: random sequence generation, allocation concealment, blinding of investigators and subjects, blinded evaluation of study outcomes, completeness of outcome data, and selective reporting of study results and other biases.For each study element, the investigator made a risk of bias assessment profile according to "low risk", "high risk" and "unclear".

Statistical Analysis
We used the frequentist framework to perform a random effect network meta-analysis.The mean difference (MD) was used as an effect indicator for continuous variables, and odds ratio (OR) was used as an effect indicator for dichotomous variables.A 95% confidence interval (CI) was calculated for each effect size, and differences were considered statistically significant when P<0.05.Uncertainty in the effect of heterogeneity was defined as the inconsistency between the CI of the relative treatment effect and its prediction interval (11).The global inconsistency model was used to assess the consistency of the entire network and was considered good at p > 0.05 (12).A loop-specific approach was used to assess the presence of local inconsistencies in each closed loop.The node splitting method was used to assess the inconsistency of the model with separating evidence on a particular comparison into direct and indirect evidence (13).Funnel plots were plotted to evaluate the presence of publication bias.

Study selection and basic characteristics
Through systematic search, 60 randomized, controlled, double-blind studies involving a total of 50,333 subjects were finally included.The literature search and screening process is shown in Figure -1, and the basic characteristics of the included studies are shown in Table-1.

Evaluation of the quality of the included studies' literature
A total of 60 randomized, controlled, double-blind studies were included, including 7 four--arm studies, 18 three-arm studies and 35 double--arm studies.The overall risk of bias was generally low.The risk of bias was assessed as shown in Supplementary Table-2 (see Page 2).

Effectiveness indicators
Mean daily micturitions Forty-two RCTs (14-56) reported micturition's/d, including 2 studies in 4 arms, 12 studies in 3 arms and 30 studies in double arms, containing a total of 15 treatment measures and a total sample size of 32,317 cases (Figure -2).Initial overall inconsistency testing showed a p-value <0.05 and partial p-values <0.05 in ring inconsis-tency, so subgroup regression analysis of the data according to the proportion of female patients showed that all inconsistency testing p-values were >0.05.For the subgroup with ≥50% female, all interventions were significantly more effective than placebo compared to placebo, except for oxybutynin (OXY)5mg-TID, with SOL10mg-QD being the most effective and significantly better than the majority of interventions.For the subgroup with less than 50% women, SOL10mg-QD remained the most effective, with statistically significant differences in efficacy compared to propiverine (PRO) 20mg-QD, mirabegron (MIR) 50mg-QD, extended-release tolterodine (ER--TOL) 4mg-QD and PBO.Results of the NMA are reported in Supplementary Table

Mean daily incontinence episodes
Twenty-three RCTs (14-16, 25-44) reported incontinence episodes/d, including two 4-arm studies, eight 3-arm studies and 14 two-arm studies, comprising a total of 14 treatment measures and a total sample size of 15,632 cases (Figure -2).Among these studies, since the inclusion criteria for the Dmochowski 2003 et al. (54).study was "patients at least 18 years of age taking current pharmacologic treatment for OAB", this study had significant clinical heterogeneity with other study populations, and the data were analyzed after excluding this study.The results showed that SOL10mg-QD was the most effective, followed by SOL5mg-QD and SOL5/10mg-QD.Results of the NMA are reported in Supplementary Table-4 (see Page 7). Figure -3 shows the mean values of SU-CRA for interventions on micturitions.

Mean daily urgency incontinence episodes
Twenty-nine RCTs (15-19, 22-30, 37-51) reported urgency episodes/d, including three 4-arm studies, eight 3-arm studies and 18 two-arm studies, containing a total of 14 treatment measures and a total sample size of 17,859 cases (Figure -2).The results showed that FES8mg-QD was the most effective in reducing mean daily urgency incontinence episodes, followed by SOL10mg-QD, with no statistically significant difference between the two, but both showed significant improvements in efficacy compared to most other interventions.FES8mg-QD was significantly more effective than FES4mg-QD and FES4/8mg-QD; while the difference in efficacy between SOL10mg-QD and SOL-5mg-QD and SOL5/10mg-QD was not statistically significant.All interventions were significantly more effective than placebo and the differences were statistically significant, except for TOL2mg--BID which showed no significant improvement in efficacy differences compared to placebo.Results of the NMA are reported in Supplementary Table-6 (see Page 11). Figure -3 shows the mean values of SUCRA for interventions on micturitions.

Safety outcomes
Fifty-five RCTs (14, 16-18, 20-31, 33-41, 43-53, 56-73) reported dry mouth, and to exclude nocebo effect on study outcomes, two articles (14,25) with significantly higher data in the placebo group than in other studies were excluded.Therefore, two 4-arm studies, 17 three-arm studies, and 34 two-arm studies, containing a total of 19 treatment measures and a total sample size of 45,756 cases, were considered (Figure -4).The results showed that the interventions with the lowest incidence of dry mouth were VIB-QD, MIR50mg-QD and PBO respectively.Constipation was reported in 50 RCTs, including two 4-arm studies, 18 three--arm studies, and 30 two-arm studies, containing a total of 19 treatment measures and a total sample size of 45,674 cases.The incidence of constipation was not significantly higher for FES4mg-QD, ER--OXY10mg-QD, TOL2mg-BID, and VIB-QD compared with placebo, while the incidence of constipation was higher for the remaining interventions than for the placebo group.A total of nine interventions were included for hypertension, of which only IMI0.1 mg-BID caused a significant difference in the incidence of hypertension compared with placebo and other treatments, and the remaining seven were not significantly different compared with placebo.For headache, 17 interventions were included, and only FES4/8mg-QD and CR-DAR-15mg-QD were found to exhibit a significantly higher incidence compared to placebo.A total of 18 interventions were included for urinary tract infections, and their incidence with only SOL10mg--QD differed statistically significantly from placebo.

Inconsistency and heterogeneity check
Initially, in improving mean daily micturition's and voided volume per micturition, overall inconsistency testing showed inconsistency (P value < 0.05) and inconsistency in individual rings (95% CIs not including 1), and subgroup analysis based on race, duration of disease, and other factors did not reveal significant improvement.Therefore, subgroup regression analysis of the data according to the proportion of female patients showed that the overall inconsistency and ring inconsistency p values were >0.05.In terms of reducing mean daily incontinence episodes, sensitivity analysis showed that the study by Dmochowski 2003 et al. (54).was significantly different from other studies, considering that the inclusion criteria for the study were "patients at least 18 years of age taking current pharmacologic treatment for OAB".Therefore, this study showed significant clinical heterogeneity with other study populations in the efficacy index of reduction in the number of incontinence episodes.Thus, analysis of the data upon excluding this study would show no inconsistency.The global inconsistency model showed well with p>0.05 (Figures 6-8).The result of local inconsistency showed that most loops were consistent according to the 95%CI.The test for inconsistency using node-splitting model revealed no significant difference between direct and indirect comparisons (P>0.05).

Publication bias
A funnel plot was established to assess the publication bias.There was no significant evidence of publication bias for outcomes based on a Begg funnel plot (Figure -9).

DISCUSSION
OAB is a chronic syndrome that is not life--threatening and does not progress to uncontrollable functional impairment but has serious impacts on the patient's quality of life.Therefore, current research is increasingly focused on the impact of interventions on the quality of life of patients Ten OAB therapeutic agents were included in this study, involving a total of 19 interventions grouped by different doses administered, and the NMA results show that solifenacin had a relatively good overall efficacy and a significant advantage in improving patients' symptoms.Solifenacin 10mg was the most effective in reducing the number of voiding and incontinence; solifenacin 5/10mg was the most effective in reducing urinary urgency and nocturia; solifenacin 10mg ranked second in both urgency incontinence and voided volume.In terms of safety, the incidence of dry mouth events with solifenacin 5mg was not significantly different from placebo and was significantly lower than other anticholinergic drugs.Solifenacin is a competitive antagonist of M3 receptors and is highly specific and selective for bladder M3 receptors.The results of past studies have shown that solifenacin has a weaker blocking effect on salivary secretion than other anticholinergic drugs and that it inhibits salivary secretion at 3.6-6.5 times the effective concentration at which it produces an effect in the bladder (74,75), which is consistent with the results of the present study.However, in the case of constipation, the results of this study showed that even a small dose of solifenacin (5mg) increased the incidence of constipation.Constipation has the greatest effect on patient satisfaction (76).Therefore, the results suggest that solifenacin is not recommended for the clinical treatment of patients with OAB who are prone to constipation.
Different interventions have different pharmacological characteristics, and different doses may affect the efficacy of treatment, in addition to their safety.Therefore, it is necessary to select the appropriate medication and dose according to the individual patient's condition so that the patient's quality of life can be maximized.This NMA analyzed the incidence of dry mouth, constipation, nasopharyngitis, headache, hypertension, and urinary tract infection in the included studies and showed that anticholinergic drugs may increase the incidence of dry mouth and constipation, while imidafenacin may increase the risk of hypertension, and FES4/8mg-QD and CR--DAR15mg-QD increase the incidence of headache compared to placebo.SOL10mg-QD may increase the risk of urinary tract infections.
Before choosing a treatment plan, the benefits of the treatment plan for the patient and the possible risks and complications should be fully considered, and decisions should be made after weighing the pros and cons.In terms of efficacy, vibegron and mirabegron are superior to placebo and comparable to anticholinergics; although they do not show an efficacy advantage over anticholinergic drugs, their greatest advantage is in terms of safety, with both drugs showing good tolerability.In particular, vibegron and mirabegron have a significant advantage over cholinergic receptor antagonists with respect to dry mouth.As potent β3 agonists, vibegron and mirabegron relax the detrusor muscle by activating β3 receptors, thereby increasing bladder capacity and prolonging the interval between voiding without affecting bladder voiding activity.The selectivity for β3 receptors over other β receptor subtypes also suggests that both drugs are effective and well-tolerated novel drugs for OAB patients (77,78).
In the voided volume per micturition outcome indicator, there was inconsistency between the direct and indirect comparison results of SOL-10mg-QD and PRO20mg-QD (p-value 0.017).Although the direct and indirect comparisons were significantly different, the results of the two interventions compared pointed towards the same direction, suggesting that SOL10 mg-QD was superior to PRO20 mg-QD, varying only in the degree of their difference, so the results were considered to be somewhat reliable.
Because of the overall inconsistency in this NMA study in terms of decreasing micturition/d and increasing voided volume/ micturition, a subgroup regression analysis was performed.Despite the differences between male and female in the anatomy and physiology of the lower urinary tract system and the potential mechanisms of action that may lead to OAB-like symptoms (79), none of the clinical studies included "gender" as an analyzable data in detail, but simply expressed whether the proportion of women was ≥50%, so only subgroups of women ≥ and <50% were analyzed in this study.The results of the subgroup analysis suggest that the results of imidafenacin are opposite in the subgroups with greater than and less than 50% women, so it is speculated that the efficacy of imidafenacin in men and women may vary, which would need to be confirmed by the results of more single-sex studies.
To control for homogeneity in the included studies, strict entry row criteria were established, and all 12-week efficacy indicators were used as the endpoints examined in this study, which avoided the introduction of clinical heterogeneity due to different study periods.Some limitations still exist in this study: 1.Because the quality of life measurements used in different studies are not uniform, this indicator of quality of life has not been analyzed and compared.Clinical endpoints can assess the effectiveness of symptom treatment from an objective perspective, but further research is needed to determine whether these symptom changes are relevant to the improvement of patients' quality of life.2. No subgroup analysis of age was performed in this study.Existing studies have shown differences in the effectiveness of solifenacin versus mirabegron in elderly and non-differentiated age groups (80).However, only 2 of the studies included in this study enrolled elderly subjects, so subgroup analysis could not be performed.3.No comparative study of long--term medication use was conducted in this study.Overactive bladder requires long-term medication treatment, and the data from the 12-week study used in this study are not representative of its true efficacy and safety.

CONCLUSIONS
Individualized treatment based on the characteristics of the patient is crucial.Anticholinergic drugs carry a risk of increased incidence of dry mouth and constipation, with lower doses carrying a lower risk.Solifenacin (10mg, 5mg/10mg) has significant advantages in improving patient symptoms.However, even low doses of solifenacin (5mg) can increase the incidence of constipation.In addition, imidafenacin may increase the risk of hypertension, FES4/8mg and CR-DAR-15mg may increase the incidence of headaches, and SOL10mg-QD may increase the risk of urinary tract infections.These drugs should be used with caution in patients at risk for these side effects.Although the efficacy of mirabegron and vibegron is not superior to anticholinergic drugs, they are better tolerated by OAB patients. - 3 (see Page 5).

Figure- 3
shows the mean values of SUCRA for interventions on micturitions.

Figure 1 -
Figure 1 -Flow Chart of Literature Search and Screening.

Figure 2 -
Figure 2 -Evidence Network Plot for Micturitions with Female Proportion>50% (A), Micturitions with Female Proportion≤50% (B), Incontinence (C), Urgency (D), Urgency Incontinence (E), Nocturia (F), Voided Volume/micturition with Female Proportion>50% (G), Voided Volume/micturition.with Female Proportion ≤ 50%.Lines connect the interventions that have been studied in head-to-head (direct) comparisons in the eligible randomized controlled trials.The width of the lines represents the cumulative number of randomized controlled trials for each pairwise comparison, and the size of every node is proportional to the number of randomized participants (sample size).A (see Page 9).Figure-3 shows the mean values of SUCRA for interventions on micturitions.
Figure-5 shows the mean values of SUCRA for interventions on AEs.Results of the NMA are reported in Supplementary Tables 9-14 (see Page 17-30).Figure-6 shows the mean values of SUCRA for interventions on safety outcomes.

Figure 4 -Figure 5 -Figure 6 -
Figure 4 -Evidence Network Plot for Dry Mouth (A), Constipation (B), Nasopharyngitis (C), Hypertension (D), Urinary Tract Infection (E), Headache (F).Lines connect the interventions that have been studied in head-to-head (direct) comparisons in the eligible randomized controlled trials.The width of the lines represents the cumulative number of randomized controlled trials for each pairwise comparison, and the size of every node is proportional to the number of randomized participants (sample size).A