EUVIMEDEUVIMED entdecken
Systematische ÜbersichtMetaanalyse

Efficacy and Safety of Bumetanide (NKCC1 Inhibitor) for Core Symptoms in Childrenwith Autism Spectrum Disorder: An Updated Systematic Review and Meta-Analysis(2020–2026)

Arbind Kumar Choudhary

Pharmacology · 17.11.2026

DOI: 10.1159/pha/aejag012 ↗

Kurz gefasst

Systematische Übersicht in der Zeitschrift Pharmacology, online erschienen am 17.11.2026. Verfasst von Arbind Kumar Choudhary.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Bumetanide, a selective inhibitor of the Na⁺-K⁺-2Cl⁻ co-transporter 1 (NKCC1), is hypothesised to restore hyperpolarising γ-aminobutyric acid (GABA)-mediated inhibition and has been investigated as a disease-modifying therapy for the core symptoms of autism spectrum disorder (ASD). Early small randomised controlled trials (RCTs) reported benefits on the Childhood Autism Rating Scale (CARS); however, two large Phase III trials (SIGN-1 and SIGN-2; Fuentes et al., 2023) failed on their primary endpoints. An updated systematic review and meta-analysis of RCTs, with the search extended from January 2020 to May 2026, was conducted to clarify the efficacy, safety and moderators of effect of bumetanide for core ASD symptoms in children.

Methods

Seven databases (PubMed/MEDLINE, Embase, Web of Science, Cochrane CENTRAL, PsycINFO, ClinicalTrials.gov and the WHO ICTRP) were searched from January 2020 to 9 May 2026. A single investigator screened records, extracted data and assessed risk of bias using the Cochrane Risk of Bias 2 tool, each step being carried out twice at a separation of several weeks with the second pass made blind to the first. The primary outcome was change in CARS total score. Secondary outcomes included the Social Responsiveness Scale, second edition (SRS-2), the Autism Diagnostic Observation Schedule (ADOS), Clinical Global Impression scales and pre-specified safety endpoints. Mean differences (MD) and standardised mean differences (SMD; Hedges' g) were pooled using a DerSimonian–Laird random-effects model, and absolute effects and numbers needed to harm (NNH) were derived for safety outcomes. All numerical data were verified against the primary reports and trial registries. Post hoc analyses examined trial design features, mean change within each arm, and how the trials reported co-occurring conditions and excitation/inhibition-relevant behavioural phenotypes. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach was used to assess certainty of evidence.

Results

Ten RCTs (about 1,020 children) were included, of which six (710 children analysed) provided arm-level CARS data. The pooled MD favoured bumetanide (MD −1.27; 95% CI −2.05 to −0.49; I² = 63%; SMD −0.47, 95% CI −0.80 to −0.14) but was well below the minimal clinically important difference of approximately 4.5 points. The four earlier trials gave MD −1.83 (95% CI −2.85 to −0.82; SMD −0.74), whereas the two Phase III trials, whose primary endpoint was the CARS2, gave MD −0.19 (95% CI −1.12 to +0.73; SMD −0.04). Children receiving bumetanide improved by a similar amount in both trial generations (3.43 and 3.57 points); children in the control arms improved by 1.28 points in the earlier trials and 3.38 points in the Phase III trials. Larger trials found smaller mean differences, and age did not significantly modify the effect (p = 0.11). No trial pre-specified attention-deficit/hyperactivity symptoms, impulsivity or seizure frequency as an outcome. Bumetanide increased hypokalaemia (RR 7.77; 95% CI 3.59 to 16.83; NNH 7), polyuria (RR 3.31; 1.47 to 7.43; NNH 4) and serious adverse events (RR 2.25; 1.03 to 4.91; NNH 24). Certainty of evidence was Low for CARS and ranged from Low to High for safety outcomes.

Conclusions

Bumetanide produces, at most, a small CARS reduction that falls below clinical meaningfulness in unselected children, is absent in the Phase III trials, and is accompanied by an unfavourable tolerability profile. The difference between trial generations lies in how much the control arms improved rather than in the response to bumetanide, so the earlier positive estimates are not a reliable guide to efficacy. The available data do not, however, allow a clean separation of ‘the drug does not work’ from ‘the trials were not built to detect what the drug does’: the endpoint used across the programme is a diagnostic severity instrument with limited sensitivity to the excitation/inhibition-linked behaviours that the mechanism should most plausibly modify, and a machine-learning reanalysis of the Phase III data has since identified cross-validated responder subgroups comprising up to 40% of participants. Current evidence does not support routine clinical use. Future trials should prioritise biomarker-stratified designs with blinded or objective outcome assessment, pre-defined responder criteria and endpoints matched to the proposed mechanism.

Zitieren

APA

Choudhary, A. K. (2026). Efficacy and Safety of Bumetanide (NKCC1 Inhibitor) for Core Symptoms in Childrenwith Autism Spectrum Disorder: An Updated Systematic Review and Meta-Analysis(2020–2026). Pharmacology. https://doi.org/10.1159/pha/aejag012

Vancouver

Choudhary AK. Efficacy and Safety of Bumetanide (NKCC1 Inhibitor) for Core Symptoms in Childrenwith Autism Spectrum Disorder: An Updated Systematic Review and Meta-Analysis(2020–2026). Pharmacology. 2026. doi:10.1159/pha/aejag012

BibTeX
@article{Choudhary2026,
  author = {Choudhary, Arbind Kumar},
  title = {Efficacy and Safety of Bumetanide (NKCC1 Inhibitor) for Core Symptoms in Childrenwith Autism Spectrum Disorder: An Updated Systematic Review and Meta-Analysis(2020–2026)},
  journal = {Pharmacology},
  year = {2026},
  doi = {10.1159/pha/aejag012}
}
Kostenlos starten →