A systematic mapping review of therapeutic clinical trials in dengue

Dengue is a growing public health threat with increasing case numbers globally. Despite the substantial burden, there are no licensed therapeutics for patients with dengue. To inform the design of large-scale practice-changing clinical trials and to assess the feasibility of an individual patient data platform for meta-analysis, we conducted a systematic mapping review of clinical trials evaluating dengue therapeutics. Our aims were to characterise published and registered dengue therapeutic trials, describe their endpoints, and assess study design quality and internal validity to inform feasibility of meta-analysis and future research.

We systematically searched Ovid MEDLINE, Ovid EMBASE, WHO ICTRP and ClinicalTrials.gov for prospective clinical trials evaluating therapeutics in patients with symptomatic dengue. Two independent reviewers screened records using Covidence. Data were extracted into a REDCap database, and risk of bias was assessed using the ROB-2 and ROBINS-I tools to describe trial design rigour. Descriptive analyses summarised the interventions, trial characteristics, study populations, and primary endpoints. This systematic review was pre-registered with PROSPERO (CRD42023469022).

A total of 121 clinical studies were identified, comprising 72 published trials and 49 registered but unpublished studies. Interventions were categorised according to the authors' proposed mechanism of action: antiviral (n = 10), host-directed (HDT, n = 34), supportive (n = 31), or undefined (n = 46). Aside from the studies of supportive therapies (n = 31) and unpublished studies (n = 37) which were only reviewed for their primary outcomes, 53 publications remained for review of therapeutic efficacy. Methodological concerns were common - 24 of 53 published trials (45%) were classified as having high or critical risk of bias. Corticosteroids were the most frequently evaluated intervention, involving a total of 944 randomised patients. The primary endpoints used in both antiviral and HDT trials were highly heterogeneous, limiting comparability. The combination of methodological concerns and non-standardised endpoints precluded meta-analysis for any intervention. No single treatment had sufficient or consistent evidence to support recommendations for use in clinical practice.

Our findings highlight a remarkably sparse evidence base for dengue therapeutics and a lack of standardised, clinically meaningful endpoints. These factors have hindered progress in evaluating candidate treatments and limited the potential for individual patient data meta-analyses. Large, high-quality trials - powered for harmonised and clinically relevant endpoints - are urgently needed to advance the development of effective therapies for dengue.