Discontinuation Rates of Tadalafil Alone and in Combination with a-Blockers in the Treatment of Male Lower Urinary Tract Symptoms with or without Coexisting Erectile Dysfunction: A Systematic Review and Meta-Analysis

Purpose We examined the discontinuation rates of tadalafil alone and in combination with a-blockers (ABs) for the treatment of male lower urinary tract symptoms (LUTS), with or without erectile dysfunction (ED). Materials and Methods We searched the EMBASE, PubMed, Web of Science, Scopus, Cochrane Library, and ClinicalTrials.gov databases for studies published until May 15, 2022. The discontinuation rates associated with LUTS medications were subsequently analyzed by meta-analysis. Results Forty-four studies, including 1724 discontinued patients, were included. The combined discontinuation rate was 12.78% (95% confidence interval (CI) 9.89–15.98%), and the discontinuation rates because of adverse events and lack of efficacy were 4.56% (95% CI 3.39–5.90%) and 3.30% (95% CI 1.53–5.72%), respectively. Conclusions The discontinuation rate of tadalafil alone or in combination with ABs for LUTS with or without ED was relatively low and varied according to the study type. Patients receiving monotherapy or combination therapy were similarly likely to abandon treatment. Treatment with a fixed-dose combination was associated with better persistence than with a free-dose combination. These data may help guide clinicians in selecting drug regimens when making decisions. Factors associated with treatment withdrawal need to be determined through high-quality clinical studies to reduce the drug discontinuation rate, which will ultimately reduce healthcare costs and improve patient outcomes.


Introduction
Several preclinical and clinical trials have demonstrated a strong correlation between erectile dysfunction (ED) and lower urinary tract symptoms (LUTS) [1], several of which have also shown that LUTS caused by benign prostatic hyperplasia (BPH) is an independent risk factor for ED [1]. Treatment approaches aim to avoid BPH-related complications, stop BPH progression, and improve symptoms and quality of life. Te currently available medical therapies for LUTS include 5a-reductase inhibitors (5ARIs), a-blockers (ABs), and their combination therapy (CT) [2,3]. Although efective, these treatments can exert side efects associated with sexual dysfunction [4].
Tadalafl, a phosphodiesterase type 5 inhibitor (PDE5i), has been shown to be efective in once-daily and/or ondemand treatment for ED, with proven efcacy in multiple controlled clinical trials, including those in LUTS patients with and without ED [5,6]. Tadalafl acts directly on the micturition phases, not just penile erection, by increasing nitric oxide levels in the smooth muscle to relax the prostate and bladder neck [7,8]. Te efcacy of PDE5is in combination with ABs for remission of LUTS has also been assessed. Currently, research has shown this treatment to have benefcial additive efects on both sexual function and LUTS compared to monotherapy [9]. Hence, the opportunity to treat LUTS with or without ED by using tadalafl alone or in combination with ABs may lead to new and increasingly tailored treatment strategies.
However, several clinical surveillance studies have suggested that the discontinuation rates of medical treatment for LUTS, including overactive bladder or BPH, are quite high [10,11]. In these studies, dissatisfaction and therapy failure were identifed as the main reasons for drug withdrawal. To our knowledge, longitudinal information regarding the use of tadalafl alone or in combination with ABs in patients with LUTS is limited. Tus, we performed this systematic review and meta-analysis to investigate the discontinuation rate of tadalafl alone or in combination with ABs in male LUTS patients with or without ED.

Study Protocol.
Tis study protocol was registered on the International Platform of Registered Systematic Review and Meta-Analysis Protocols (INPLASY, registration number: INPLASY202260105) [12].

Search Strategy.
Te English-language literature was systematically reviewed until May 15, 2022, in accordance with the criteria of the Preferred Reporting Project for Systematic Reviews and Meta-Analysis (PRISMA) [13]. Te PRISMA checklist is provided in Table S1. Te EMBASE, PubMed, Scopus, Web of Science, Cochrane Library, and ClinicalTrials.gov databases were searched to identify studies reporting the efects of tadalafl alone or in combination with ABs to treat LUTS patients with or without coexisting ED. We performed an extensive search using Medical Subject Headings (MeSH) terms and related keywords: "tadalafl," "lower urinary tract symptoms," "Cialis," and "bladder overactive" (Table S2). Additionally, references in the retrieved articles were manually searched to attempt to identify any studies not found in the original literature search.

Study Selection.
Based on the population, intervention, comparison, outcomes, and study design (PICOS) approach, the trials included in our study met the following criteria: (P) male patients with LUTS with or without ED; (I) consuming medication: tadalafl alone or in combination with ABs; (C) none; (O) evaluating the incidence of drug discontinuation: studies must report patients discontinuing medication, regardless of reasons; and (S) prospective, retrospective, observational, and randomized clinical trials (RCTs) were included. We placed no restrictions on the sample size, drug dose, or follow-up duration. We excluded any studies which were repeated publications, or those for which the original data for the literature could not be provided, or those for which the authors did not receive a response after contacting the authors. We further excluded any non-human studies, case series, case reports, commentaries, and reviews.

Study Outcomes.
Tis study aimed to measure the discontinuation rate of the main treatment drug (tadalafl alone or in combination with ABs) for LUTS. Discontinuation was defned as the nonpersistence of the main treatment drug prescribed at the start of the frst treatment, regardless of the reason. Te discontinuation rate was calculated by dividing the number of patients who discontinued treatment by the total number of initial index patients. Further, the discontinuation rate due to adverse events (AEs) was defned to include only patients who discontinued treatment because of AEs in the numerator. Similarly, the discontinuation rate due to lack of efcacy was defned to include only patients who discontinued treatment because of drug inefcacy in the numerator.

Data Extraction and Quality Assessment.
Te authors CQ and M-YJ separately extracted data from each included study and crosschecked the results after extraction. Any disagreements between reviewers regarding data extraction were resolved through consensus discussion with a third reviewer. Te information extracted from the included studies was as follows: (1) publication date, (2) name of the frst author, (3) country of study, (4) type of design, (5) drug regimen and dose received by the patient, (6) number of participants per drug regimen, (7) treatment period, and (8) data on total number of patients who discontinued treatment, discontinuation due to AEs, and no efcacy.
Bias assessments of observational studies were rated on the Newcastle-Ottawa Scale, whereas RCTs were evaluated using the Cochrane risk of bias assessment tool.

Statistical Analyses.
Te statistical software R version 4.0.3 (package "meta") was used for all statistical analyses.
Due to diferences in the methodological and clinical perspectives in the included studies, we surmised that the heterogeneity among the included studies may be large. Hence, we used a random-efects model to obtain pooled estimates and a 95% confdence interval (CI) of the discontinuation rate, and the variance was stabilized by an arcsine transformation [14,15]. Heterogeneity was assessed using Cochran's value and I 2 . Te heterogeneity test was cut of at signifcant Cochran Q-values (P < 0.1) and I 2 > 50%, as an I 2 value of 30-50% was suggested as a cutof value for moderate heterogeneity [16]. A prediction interval (PI) for the proportion in a new study was calculated if the argument prediction and comb. random were TRUE [17].
A cumulative meta-analysis was conducted to determine whether the discontinuation rate stabilized with an increasing sample size. A sensitivity analysis was further performed to phase out each study to test the reliability of the discontinuation rate.

Search Results and Study Characteristics.
Te initial literature search revealed 2361 studies, of which 44 comprising a total of 1724 discontinued patients were deemed eligible. Te PRISMA literature selection fowchart is shown in Figure 1.

Cumulative Analysis of Discontinuation Rate.
Cumulative meta-analysis results showed that, with an increase in sample size, point estimation tended to be stable and CI gradually decreased, indicating that the greater the sample size, the higher the accuracy of the results ( Figure S2).

Sensitivity Analysis of Discontinuation Rate.
When each study was sequentially excluded from the analysis, the sensitivity analysis results (12.08% to 13.14%) remained in good agreement with the discontinuation rate, indicating that the study results were robust ( Figure S3).

Discussion
LUTS symptoms and ED severity tend to increase with age [60]. Hence, as with other chronic diseases, long-term use of LUTS and ED medications is important to improve the patient's symptoms and quality of life [61]. Although tadalafl alone and in combination with ABs has been proven to be efective and well tolerated in clinical studies, patient outcomes in clinical practice are not always consistent with research fndings. One reason for this may be the relatively high rates of premature drug discontinuation [10]. Tis is the frst report to investigate the discontinuation rate of tadalafl alone or in combination with ABs for the treatment of male LUTS with or without ED. Te total discontinuation rate was 12.78%, of which the discontinuation rates due to AEs and inefectiveness were 4.56% and 3.30%, respectively.
Overall, the discontinuation rates observed herein appear to be relatively low, despite the inclusion of observational studies and RCTs. Interestingly, the discontinuation rate was much higher in observational studies than in RCTs.

International Journal of Clinical Practice
Te lower discontinuation rate in clinical studies may be due to both better adherence to recommendations made in the clinical trial setting and increased patient motivation [33]. Additionally, in clinical trials, patients are closely observed and receive medication free of charge during the study period [62]. Tis can improve the incidence of drug persistence. In contrast, in real-world clinical practice, patients often pay for drugs, which may result in higher expectations for their efcacy as well as a decreased tolerance for side efects. Moreover, a previous study also showed that selfreported economic status was related to long-term medication persistence [63]. In this study, respondents were asked whether their household income was sufcient to meet their needs. Persistence was 30% higher among those who answered "sufcient" or "more than sufcient." Knowledge of these factors has allowed more illuminating clinical research to be conducted in real-world practice settings; hence, the actual incidence of treatment discontinuation can be calculated more accurately. Nevertheless, it is inevitable that the discontinuation rate will be underestimated in retrospective studies. Te reasons for this include that patients who were excluded due to unrecorded follow-up examinations could not be analyzed, and it is not possible to know whether they discontinued treatment or continued treatment at another hospital. In some cases, discontinuation data are obtained through prescription records or selfreports rather than independent audits of pill counts or other, more accurate, verifcation methods. Treatment persistence may be overestimated if patients fail to refll their prescriptions.
Even more interestingly, patients who were treated with CT or monotherapy had similar discontinuation rates (12.87% vs. 12.16%). Tese data contradict the results of prior research evaluating long-term treatment for LUTS, which showed a lower discontinuation rate for CT (doxazosin + fnasteride) (18%) compared to doxazosin (27%) and fnasteride (24%) monotherapy [64]. In addition, the discontinuation rates reported in our study were relatively lower than those calculated in this study. However, it is difcult to compare the two studies owing to the diferences in drug regimens and study design (meta-analysis vs. prospective randomized study), and these diferences should therefore be interpreted with caution. Notably, discontinuation of tadalafl or ABs was easier and more regulated than Records removed before screening: Figure 1: PRISMA fow chart illustrating the study selection process. International Journal of Clinical Practice  International Journal of Clinical Practice that of 5ARIs or antimuscarinics (AMs). In fact, tadalafl and ABs both show rapid activity and can be easily and efectively retaken [26,65]. However, compelling evidence in the literature suggests that treatment persistence is inversely correlated with the complexity of medication regimens. With this in mind, socalled FDC drugs (compilation of two or more drugs in a single tablet/pill) have been developed to treat multiple clinical conditions (e.g., hyperlipidemia and hypertension) or one disease (e.g., asthma, diabetes mellitus) in a complementary manner [66]. When only RCTs were included, a single tablet in an FDC regimen has a lower discontinuation rate than the use of two tablets in the freedose combination for the treatment of LUTS. Our fndings are consistent with the results from two retrospective observational cohorts of men with LUTS treated with AM + AB in Spain and the Netherlands, showing FDC had higher persistence rates than free-dose combinations [67,68]. Tis advantage has been extensively demonstrated in other pathologies. A meta-analysis of hypertension data published between 2000 and 2017 showed that patients who received FDC had higher treatment persistence for their hypertension medication than those who received a free-dose combination [69]. Unlike the overall discontinuation rate, the discontinuation rate due to AEs was higher with CT than with tadalafl monotherapy. Although most AEs are self-limiting, they directly afect a patient's perception of treatment satisfaction. Te least wanted undesirable efect was ejaculatory dysfunction (retrograde or diminished ejaculation), which is a well-known side efect of selective ABs [70]. Tis has previously been shown to be the main factor associated with high satisfaction in the tadalafl only group compared with the PDE5-I combination or tadalafl combination groups (both of which include the use of ABs) [71]. In our study, 9 patients in the CT group (tadalafl + tamsulosin) discontinued due to ejaculatory dysfunction. Interestingly, none received tamsulosin or silodosin before taking the CT [36]. However, other AEs related to tadalafl may have a smaller efect on drug withdrawal, in line with previous  )   13  145  82  13  3  15  50  33  90  14  7  11  6  17  1  6  4  82  14  39  44  74  8  9  11  4  128  17  18  3  7  8  4  1  672  7  6  10  18  8  10  3  7  2   267  846  466  161  51  171  406  306  676  306  51  105  67  363  15  88  122  953  69  158  97  155  47  31  80  81  427  90  84  139  38  35  44  38  1393  58  66  167  158  136  99  56        treatment of LUTS and not ED, which is a factor in the targeting of comorbid conditions. Finally, because of the inclusion of diferent study types, we did not perform any analysis of any specifc factors that could signifcantly alter or afect the treatment discontinuation rates. In fact, treatment discontinuation was associated with objective clinical data and provider factors, as well as demographic data and subjective symptoms in patients with LUTS. Tus, a largescale, prospective, randomized trial should be performed to further investigate the factors infuencing treatment discontinuation.

Conclusion
Te discontinuation rate of tadalafl alone or in combination with ABs for LUTS with or without ED was relatively low and varied according to the study type. Patients treated with CT or monotherapy were similarly likely to abandon treatment. Furthermore, treatment with FDC appeared to have better persistence than the free-dose combination. Tese data may help guide clinicians in decision-making for drug regimens. At the same time, further studies are required to assess the factors afecting the discontinuation incidence and help develop strategies to reduce their occurrence.

Data Availability
All data generated or analyzed during this study are included in this article (and its Supplementary material fles).

Ethical Approval
Tis retrospective review of patient data did not require ethical approval in accordance with local/national guidelines. Written informed consent from participants was not required in accordance with local/national guidelines.

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Authors' Contributions
Research conception and design was performed by Songxi Tang and Qiang Chen. Data acquisition was performed by Qiang Chen and Yinjun Mao. Statistical analysis was performed by Qiang Chen and Yinjun Mao. Data analysis and interpretation was performed by Yinjun Mao and Huiliang Zhou. Drafting of the manuscript was performed by Qiang Chen and Yinjun Mao. Critical revision of the manuscript was performed by Songxi Tang and Huiliang Zhou.Supervision was performed by Songxi Tang. Approval of the fnal manuscript was given by all authors. Qiang Chen and Yinjun Mao contributed equally to this work. Table S1: PRISMA 2020 checklist. Table S2: Search strategy. Table S3: Newcastle-Ottawa Scale assessment results of included observational studies. Table S4: Risk of bias of the RCTs calculated with the Cochrane risk of bias assessment tool. Figure S1: Forest plot illustrating subgroup analysis of the incidence of discontinuation with tadalafl alone or in combination with ABs for LUTS with or without ED. (A) Drug regimen, (B) Drug regimen (discontinuation due to AEs), (C) Study design, (D) Combination form. ABs: a-blockers; AEs: adverse events; CI: confdence interval; CT: combination therapy; ED: erectile dysfunction; FDC: fxed-dose combination; LUTS: lower urinary tract symptoms; RCT: randomized clinical trial. Figure S2: Cumulative meta-analysis of the discontinuation rate. CI: confdence interval. Figure S3: Sensitivity analysis of the discontinuation rate. CI: confdence interval. (Supplementary Materials)