Desmopressin Withdrawal Strategy for Pediatric Enuresis: A Meta-analysis
CONTEXT: A high relapse rate after discontinuation of desmopressin treatment of pediatric enuresis is consistently reported. Structured withdrawal strategies have been used to prevent relapse.
OBJECTIVE: To assess the efficacy of a structured withdrawal strategy of desmopressin on the relapse-free rate for desmopressin responder pediatric enuresis.
DATA SOURCES: Systematic literature search up to November 2015 on Medline, Embase, Ovid, Science Direct, Google Scholar, Wiley Online Library databases, and related references without language restriction.
STUDY SELECTION: Related clinical trials were summarized for systematic review. Randomized controlled trials on the efficacy of structured versus abrupt withdrawal of desmopressin in sustaining relapse-free status in pediatric enuresis were included for meta-analysis.
DATA EXTRACTION: Eligible studies were evaluated according to Cochrane Collaboration recommendations. Relapse-free rate was extracted for relative risk (RR) and 95% confidence interval (CI). Effect estimates were pooled via the Mantel–Haenszel method with random effect model.
RESULTS: Six hundred one abstracts were reviewed. Four randomized controlled trials (total 500 subjects) of adequate methodological quality were included for meta-analysis. Pooled effect estimates compared with the abrupt withdrawal, structured withdrawal results to a significantly better relapse-free rate (pooled RR: 1.38; 95% CI, 1.17–1.63; P = .0001). Subgroup analysis for a dose-dependent structured withdrawal regimen showed a significantly better relapse-free rate (pooled RR: 1.48; 95% CI, 1.21–1.80; P = .0001).
LIMITATIONS: The small number of studies included in meta-analysis represents a major limitation.
CONCLUSIONS: Structured withdrawal of desmopressin results in better relapse-free rates. Specifically, the dose-dependent structured withdrawal regimen showed significantly better outcomes.
- CI —
- confidence interval
- NNT —
- number needed to treat
- RCT —
- randomized controlled trial
- RR —
- relative risk
The International Children’s Continence Society defines pediatric enuresis as a symptom and condition of incontinence among children ages ≥5 years that occurs exclusively during sleeping periods, with >1 episode per month, and a frequency of 3 episodes over 3 months.1 Based on the previous literature, the prevalence of enuresis in the general pediatric population is 8% to 15% in children age 6 years and ∼2% in adolescents.2–5 Desmopressin, a vasopressin analog that reduces urine production, is currently the first-line pharmacologic agent recommended for the treatment of pediatric nocturnal enuresis.5 Treatment results with desmopressin are favorable, with acceptable adverse effects; however, as reported by several studies, the relapse rate is high.5,6 Therefore, a structured withdrawal (time dependent and dose dependent) of desmopressin therapy was being studied in several retrospective studies. Although these results showed that structured withdrawal is more time consuming and costly, a higher clinical success rate was achieved.7–9 Despite its popularity, available literature showed inconsistent results when compared with abrupt discontinuation of desmopressin.10–13 Therefore, there is an imminent need to assess the clinical applicability of structured desmopressin withdrawal compared with abrupt discontinuation. Our aim is to assess the efficacy of different structured withdrawal strategies and an abrupt withdrawal strategy to prevent relapse and sustain dry nights for desmopressin-responsive patients with pediatric nocturnal enuresis.
The protocol of this meta-analysis was made in consultation with a topic expert and methodologist; it was also registered in the PROSPERO registry (CRD42015017895) and complies with the recommendations of Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.14 Meta-analysis and risk of bias assessment were performed according to the Cochrane Collaboration recommendations.15
Identification of the Literature
Two independent reviewers (M.E.C. and J.M.S.) identified published medical literature of human studies about desmopressin in the treatment of pediatric nocturnal enuresis. This search was carried out in March 2015, and an updated search was conducted in November 2015, for the following electronic databases: Medline, Embase, Science Direct, Ovid, Google Scholar, and Wiley Online Library. The reviewers searched the clinicaltrial.gov Web site for possible unpublished trials, and inquiries were sent to the authors about the incomplete data. The search strategy used was as follows: (desmopressin AND enuresis). No restriction on language was imposed. A comprehensive search for eligible studies was carried out to minimize reporting bias, publication bias, within-trial reporting bias, and their potential impact. Hand-searching of review articles and cross-references of reference lists that met our inclusion criteria were also performed for potentially relevant titles. External peer reviewers were asked to identify additional relevant studies that were not included in the draft.
This meta-analysis includes only randomized controlled trials (RCTs) that compare the efficacy of structured versus abrupt withdrawal (as control) of desmopressin among desmopressin-responsive patients with pediatric nocturnal enuresis. Clinical trials (ie, non-RCTs, comparisons of desmopressin therapy with other medical agents or behavioral therapy, and clinical trials with no control group) were excluded for meta-analysis but included in the systematic review, with assessment of the outcome of relapse-free rate, tabulated in a supplemental table.
The primary outcome measure determined in this meta-analysis was the rate of sustained response at ≥1 month after withdrawal from desmopressin therapy. We applied intention to treat analysis with all randomly assigned subjects included for analysis of effect estimate. The dropout cases for all treatment groups were considered relapses. The subgroup analysis divided the structured withdrawal therapies into dose dependent and time dependent. Dose dependent is described as decreasing the effective dosage until the cessation of medication. Time dependent is described as sustaining the effective dosage while increasing the time interval of taking the medication until complete cessation of medication.
Evaluation of the Literature
Two reviewers (M.E.C. and J.M.S.) independently evaluated the citations and abstracts. The reviewers flagged article titles that focused on desmopressin used in pediatric nocturnal enuresis where intervention involves withdrawal of desmopressin and assessment of sustained response rate after withdrawal. Articles that either reviewer flagged were ordered, as well as articles whose abstracts and title relevance could not be determined. Two physician reviewers then independently reviewed each full-text article obtained and determined whether inclusion criteria were met. Reviewers were knowledgeable about principles of critical appraisal. Any discrepancies were resolved through consensus, and differences were reviewed by a senior physician researcher if there was disagreement.
Extraction of data on details about the study characteristics and primary outcome assessment were done. The risk of bias on quality of the study design and execution of studies were assessed independently by the 2 reviewers according to the recommendation of the Cochrane Collaboration; specifically, the randomization sequence generation; allocation concealment; blinding of participants, providers, and outcome assessors; completeness of outcome data; selective outcome reporting; and other potential sources of bias were assessed. A funnel plot was generated to explore the possibility of publication bias.15
Data Synthesis and Data Classification
For both the meta-analysis and systematic review, details of the included study trials, such as study source, study population, intervention of different structured desmopressin withdrawal regimens, study outcome, and data results were summarized in a table. The RevMan program, downloaded from www.cochrane.org, was used for data analysis, forest plot construction, and funnel plot construction.
Measures of Treatment Effect
The dichotomous data of sustained response rates as described by individual study trials were expressed as relative risk (RRs) with 95% confidence intervals (95% CIs). RR as an effect measure is appropriate for outcome assessment among prospective study models, particularly when sustained responder and relapse rate are considered. Effect estimates from individual studies were then pooled into the RevMan program for meta-analysis via the Mantel–Haenszel method with random effect model. The number needed to treat (NNT) to maintain response rate was calculated for overall effect estimate and each subgroup analysis.
Assessment of Heterogeneity
We analyzed the heterogeneity between the treatment effects of different studies by looking at the points on the forest plot, the overlap of CIs (a poor overlap indicates heterogeneity), and the χ2 statistical test for heterogeneity. A low P value (or a large χ2 statistic relative to its degrees of freedom) shows evidence of heterogeneity of treatment effects. A P value of .10 rather than .05 was used to show heterogeneity because only a small number of trials with small sample sizes were included in this meta-analysis. The I2 statistic was used to quantify the variations between the studies. If a value >50% was found, significant heterogeneity was assumed, and we determined the source of heterogeneity among the study characteristics by considering the clinical and methodological characteristics of the included studies in the meta-analysis.
Figure 1 summarizes the literature search process from various sources. Six hundred one titles were available after duplicates were removed. Vast numbers of articles were retrieved that discussed desmopressin for enuresis; however, the studies rarely applied and mentioned the effect of the withdrawal strategies, so a large percentage of the records were excluded. Twenty-two eligible studies remained after screening based on title and abstract. Eighteen articles were excluded from meta-analysis but were included for systematic review and summarized in Supplemental Table 3. Four RCTs10–13 that compare the efficacy of structured and abrupt withdrawal of desmopressin after response to therapy were included for this meta-analysis. The detailed study characteristics and risk of bias assessment of these included studies are summarized in Table 1 and Table 2, respectively. There were no major discrepancies between the reviewers in relation to study inclusion, critical appraisal, and data extraction. A Chinese full-text article10 was included, which was fully evaluated with no difficulty, because the author reviewers were fluent in Chinese.
Pooled effect estimates from the 4 RCTs with no interstudy heterogeneity (χ2 = 3.82, P = .43; I2 = 0%) showed that compared with an abrupt withdrawal regimen, patients with pediatric nocturnal enuresis who responded to desmopressin and managed with a structured withdrawal approach had a significantly better sustained response rate (pooled RR: 1.38; 95% CI, 1.17–1.63; P = .0001) (Fig 2). The significant difference was sustained in the subgroup analysis of studies that applied dose-dependent structured withdrawal (pooled RR: 1.48; 95% CI, 1.21–1.80; P = .0001), with no heterogeneity noted between studies (χ2 = 1.49; P = .47; I2 = 0%). However, subgroup analysis on homogenous studies (χ2 = 0.80, P = .37; I2 = 0%) that applied a time-dependent structured withdrawal regimen showed no significant difference between treatment groups (pooled RR: 1.18; 95% CI, 0.88–1.59; P = .27). NNT for overall structured versus abrupt withdrawal is 7; for dose-dependent withdrawal it is 5, and for time-dependent withdrawal it is 12.
On visual inspection, given that the studies are evenly plotted on both sides of the average line, the funnel plot demonstrates a low probability of publication bias (Fig 3). Common study methodological limitations included risk of bias for inadequate information or allocation concealment, lack of blinding of both patient and outcome assessor, and failure to perform an intention to treat analysis (Table 2). These prompted downgrading of the overall quality of evidence from all individual outcomes. Failure of a study to report intention to treat analysis raised concerns about reporting bias.11 The overall small sample size of each study and some inconsistency across studies led to downgrading for inconsistency.
No reported adverse events were noted related to the process of desmopressin withdrawal among the treatment groups. Although, Tang et al10 and Tullus et al13 described some self-limiting nonserious symptoms associated with the desmopressin treatment (Table 1). In general, desmopressin was illustrated by most of the included studies as well tolerated, with a good safety profile for pediatric nocturnal enuresis.
Desmopressin is a synthetic analog of human endogenous antidiuretic hormone, arginine vasopressin; its major action is to reduce urine production.16 Nørgaard et al17 postulated that some children with nocturnal enuresis produce high volumes of urine with low osmolarity at night, with no associated increase in secretion of antidiuretic hormone, hence the applicability of desmopressin given at bedtime to supplement antidiuretic hormone and increase overnight urine osmolality and decrease urine volume, leading to dry nights.18–20 According to several literature reviews, desmopressin treatments render immediate response; however, they are associated with high relapse rates after abrupt withdrawal.5,6 Critical care studies have postulated that suppressed endogenous vasopressin caused by vasopressin infusion and sudden withdrawal may lead to sustained vasopressin deficiencies.21 Endocrine withdrawal syndrome is an established physiologic phenomenon after discontinuation of hormone therapy, which results into typical hormone deficiency syndrome and a generic withdrawal syndrome; therefore, it is strongly recommended to gradually taper the medication to ease withdrawal symptoms from the hormonal imbalance state.22 Applying aforementioned clinical and physiologic principles, one could elucidate the basis of high relapse rates after abrupt withdrawal in pediatric nocturnal enuresis. Recently, structured gradual withdrawal of desmopressin has been examined by several trials as a method to sustain long-term success and prevent relapse.7–13,23,24 Some studies demonstrated that desmopressin, when withdrawn structurally, has a better sustained response rate when compared with enuresis alarm alone.8,25,26 However, to date no study has clearly illustrated and explained the actual biological mechanism of improved response. A plausible reason for the better response among desmopressin structured withdrawal may be its continuing stimulation and maturation of the innate production of antidiuretic hormones from the posterior pituitary and sustaining of the circadian rhythm. However, current literature does not show sustained urine osmolality levels among desmopressin responders after structured withdrawal.18–20 In contrast, some studies demonstrate that desmopressin might have >1 mechanism of action in nocturnal enuresis treatment, such as central brain signaling, sleep arousability, circadian rhythm, and renal tubular proteins; gradual withdrawal may increase sustainability of the response.22,27–31 These concepts could be a topic for future studies to explore and explain the efficiency of structured withdrawal treatment schemes.
In our study, despite a thorough literature search, we were able to gather only 4 RCTs with adequate methodological quality that examined the sustained response and the relapse-free rate of desmopressin-responsive patients with pediatric nocturnal enuresis (Fig 1). Likewise, the pooled studies included did not have interstudy heterogeneity (Table 1, Fig 2). Eighteen non-RCTs were also identified to summarize the clinical experience described in the current literature. However, because of the low methodological quality with widely variable withdrawal schemes and inconsistent results described in these studies, no definite statement could be inferred from these studies (Supplemental Table 3).
Based on the pooled estimates from the included RCTs for meta-analysis, we recommend structured withdrawal for patients with desmopressin-responsive pediatric nocturnal enuresis to sustain long-term success and relapse-free intervals (Figs 2 and 3). This recommendation was consistent with a recent review by Alloussi et al,6 which summarized 13 retrospective heterogeneous studies. Despite the low-level evidence available at that time, they recommended against abrupt termination of desmopressin therapy. Because of insufficient data, they were unable to clearly state whether dose-dependent or time-dependent withdrawal is more likely to render sustained success. In our subgroup analysis of dose-dependent structured withdrawal, we were able to show a significantly better relapse-free rate over abrupt withdrawal. Specifically, an RR of 1.48 (95% CI, 1.21–1.80) for dose-dependent withdrawal can be translated as an NNT of 5 to prevent 1 relapse. This finding strongly justifies its application over abrupt termination of desmopressin. However, this significance was not illustrated in our additional analysis for the time-dependent structured withdrawal subgroup, with an RR of 1.18 (95% CI, 0.88–1.59) and an NNT of 12 to prevent 1 relapse. A recent comparative study by Ohtomo et al24 substantiated our findings. Their study showed that gradual dose-dependent and time-dependent structured withdrawal have significantly better outcomes with lower relapse rates when compared with only time-dependent structured withdrawal (17% vs 56%, respectively; P = .026).24 Overall, we propose that the structured withdrawal of desmopressin for pediatric nocturnal enuresis, specifically a dose-dependent withdrawal scheme, should be recommended by clinicians who treat patients with complete response to prevent wet night relapse.
The safety profile of desmopressin in the treatment of pediatric nocturnal enuresis was studied in a recent multinational trial by Van Harzeele et al,32 which found that desmopressin treatment is well tolerated in children. The same findings were reported in the 4 studies included in our meta-analysis (Table 1); likewise, structured withdrawal is not associated with any adverse event, as reported in all 18 clinical studies included for qualitative analysis and 4 randomized trials (Table 1, Supplemental Table 3).
This meta-analysis has the strength of including only adequate-quality RCTs with no interstudy heterogeneity. Likewise, no publication bias was noted. Subgroup analysis was done to assess the difference between dose-dependent and time-dependent structured versus abrupt withdrawal to infer comparison. The only limitation of this meta-analysis is the small number of studies available for effect estimation, despite a great effort by the authors to identify related trials. We recommend more high-quality RCTs to assess head-on comparison of 2 different structured withdrawal regimens to support our findings. Moreover, future studies should explore the biological mechanism of this improved response, as by comparing urine osmolality between treatment groups.
The pooled effect estimates in our meta-analysis showed that structured withdrawal of desmopressin results in a better relapse-free rate compared with abrupt withdrawal of the medication. Specifically, the dose-dependent structured withdrawal regimen showed significantly better outcomes.
- Accepted April 22, 2016.
- Address correspondence to Stephen S. Yang, MD, PhD, Division of Urology, Taipei Tzu Chi Hospital, Medical Foundation, New Tapei, Taiwan and Buddhist Tzu Chi University, Hualien, Taiwan. Email:
This trial has been registered at the PROSPERO registry for systematic reviews (http://www.crd.york.ac.uk/PROSPERO) (identifier CRD42015017895).
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Austin PF,
- Bauer SB,
- Bower W, et al
- Bayne AP,
- Skoog SJ
- Higgins JPT,
- Green S
- Vande Walle J,
- Rittig S,
- Bauer S,
- Eggert P,
- Marschall-Kehrel D,
- Tekgul S; American Academy of Pediatrics; European Society for Paediatric Urology; European Society for Paediatric Nephrology; International Children’s Continence Society
- Evans J,
- Lottmann H
- Robben JH,
- Sze M,
- Knoers NV,
- Eggert P,
- Deen P,
- Müller D
- Lee YJ,
- Lee JE,
- Choi HJ, et al
- Copyright © 2016 by the American Academy of Pediatrics