← Back
NeuroTrials.ai
Neurology Clinical Trial Database

DRAIN

Safety, tolerability, and efficacy of acetazolamide in idiopathic normal pressure hydrocephalus (DRAIN) in Sweden: a randomised, double-blind, placebo-controlled, phase 2 trial

Year of Publication: 2026

Authors: Johan Virhammar, Oskar Fasth, Maria Ekblom, ..., Dag Nyholm

Journal: Lancet Neurology

Citation: Lancet Neurol 2026; 25: 550-59

Link: https://doi.org/10.1016/S1474-4422(26)00126-2


Clinical Question

Does low-dose acetazolamide improve gait function compared with placebo in patients with idiopathic normal pressure hydrocephalus awaiting shunt surgery?

Bottom Line

Low-dose acetazolamide (up to 500 mg/day) did not improve gait or any secondary clinical outcome in idiopathic normal pressure hydrocephalus and was poorly tolerated, with high rates of adverse events and treatment discontinuation. These findings do not support its use as pharmacological treatment for iNPH; ventriculoperitoneal shunt surgery remains the only established effective treatment.

Major Points

  • First randomised, double-blind, placebo-controlled trial of any pharmacological treatment in iNPH
  • Acetazolamide did NOT improve the primary outcome (composite gait score): adjusted between-group difference 0.09 units (95% CI -3.61 to 3.79, p=0.96)
  • No benefit on any secondary clinical outcomes (iNPH grading scale, EQ-5D-5L, patient-reported outcomes)
  • Adverse events numerically more frequent with acetazolamide (80% vs 60%; p=0.22, not statistically significant)
  • Treatment discontinuation due to adverse events 4.5x higher with acetazolamide (36% vs 8%; p=0.037)
  • Minor MRI changes (ventricular volume, periventricular white matter lesions, callosal angle) seen in MRI subgroup but without clinical benefit
  • Findings argue against routine use of acetazolamide in iNPH; shunt surgery remains the only established effective treatment

Design

Study Type: Investigator-initiated, randomised, double-blind, placebo-controlled, phase 2 trial

Randomization: 1

Blinding: Double-blind (participants, investigators, clinical staff, and outcome assessors all masked)

Allocation: 1:1 computer-generated permuted block randomisation (block size 2 in MRI subgroup, block size 4 in non-MRI subgroup); concealed via sequentially numbered identical patient-specific drug packs

Enrollment Period: Feb 17, 2022 to Nov 18, 2024

Follow-up Duration: Until admission for shunt surgery or maximum 9 months

Centers: 1

Countries: Sweden

Sample Size: 50

Analyzed: 41

Analysis: Modified intention-to-treat (mITT) — all randomly assigned participants with baseline and at least one post-baseline gait assessment. Linear regression adjusted for prespecified covariates or Fisher's exact test. Prespecified per-protocol sensitivity analysis also performed.

Power Calculation: Based on observational data from 26 untreated iNPH patients showing mean 2.6% (SD 10.2) deterioration in composite gait score over 6-16 months. To detect a 10 percentage point between-group difference with 80% power and two-sided α of 0.05, 17 patients per group required. Planned sample size up to 50 allowing for dropouts.

Registration: ClinicalTrials.gov NCT04975269; EudraCT 2020-004132-22


Inclusion Criteria

  • Adults aged 50-82 years
  • Probable idiopathic normal pressure hydrocephalus per international guidelines
  • Characteristic imaging findings: ventriculomegaly with narrow callosal angle OR disproportionately enlarged subarachnoid space hydrocephalus (DESH)
  • MMSE score >20 OR iNPH grading scale cognitive domain score >30
  • Radiological features consistent with iNPH

Exclusion Criteria

  • Concomitant disease likely to have major effect on symptoms
  • Very mild gait impairment (no disturbed gait pattern)
  • Severe gait dysfunction requiring walking aids
  • Kidney, liver, or heart failure
  • Low electrolyte concentrations
  • Initiation of daily use of selected drugs that could interfere with treatment or safety assessments

Baseline Characteristics

Total randomized: 50

Median age (years): 76 (IQR 73-79)

Male: 30 (60%)

Female: 20 (40%)

Acetazolamide arm n: 25

Placebo arm n: 25

mITT Acetazolamide: 20

mITT Placebo: 21


Arms

FieldAcetazolamideControl
N2525
InterventionOver-encapsulated 250 mg acetazolamide tablets, started at one capsule orally at bedtime for 2 weeks, then titrated to one capsule twice daily (up to 500 mg/day) if tolerated. Dose could be maintained at one capsule daily or reduced to every other day if poorly tolerated.Matching placebo capsules with identical appearance, same titration schedule as acetazolamide arm
DurationUntil shunt surgery admission or maximum 9 months (median 121 days, IQR 58-200)Until shunt surgery admission or maximum 9 months (median 187 days, IQR 115-214)

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Change in composite gait score (unweighted mean of time and steps from 10 m walk, Timed Up and Go, and 3 m backward walk; higher values = worse performance; negative change = improvement) from baseline to end-of-treatment visitPrimaryPlacebo arm change in composite gait scoreAcetazolamide arm change in composite gait scoreAdjusted between-group difference 0.09 units0.96
iNPH grading scale total score (gait, balance, cognition, continence)SecondaryNo clinical benefit with acetazolamide vs placebo
Patient-reported perceived treatment effect (10-item study-specific PROM)SecondaryNo clinical benefit with acetazolamide vs placebo
Patient-reported side effects (10-item study-specific PROM)SecondaryNot improved with acetazolamide vs placebo
EQ-5D-5L total score and visual analogue scale (health-related quality of life)SecondaryNo clinical benefit with acetazolamide vs placebo
MRI subgroup: lateral ventricular volume, total and periventricular white matter hyperintensity volume, callosal angleSecondaryMinor between-group differences in ventricular and periventricular white matter lesion volumes and callosal angle with acetazolamide, but not accompanied by clinical benefit; clinical significance uncertain
Any adverse eventSafetyAcetazolamide: 20/25 (80%) · Placebo: 15/25 (60%)
Treatment discontinuation due to adverse eventsSafetyAcetazolamide: 9/25 (36%) · Placebo: 2/25 (8%)
Serious adverse events (none considered treatment-related)SafetyAcetazolamide: 2 (urinary tract infection; venous thrombosis) · Placebo: 2 (pulmonary embolism with pneumonia; transient loss of consciousness)
Any AE - AcetazolamideAdverse20/25 (80%)
Any AE - PlaceboAdverse15/25 (60%)
Discontinuation due to AE - AcetazolamideAdverse9/25 (36%)
Discontinuation due to AE - PlaceboAdverse2/25 (8%)
Serious AEs - AcetazolamideAdverseUrinary tract infection; venous thrombosis
Serious AEs - PlaceboAdversePulmonary embolism with pneumonia; transient loss of consciousness
Treatment-related SAEsAdverseNone

Subgroup Analysis

Prespecified MRI subgroup (target 24-26 patients) underwent brain MRI at baseline and 3-month follow-up. Showed minor between-group differences favoring acetazolamide in ventricular volume, periventricular white matter hyperintensity volume, and callosal angle, but these imaging changes were not accompanied by clinical benefit and their clinical significance is uncertain.


Criticisms

  • Single-center trial at Uppsala University Hospital limits generalizability
  • Small sample size (n=50 randomized, n=41 in mITT) — though adequately powered per prespecified calculation
  • Variable treatment duration (median 121 vs 187 days) because timing depended on clinical surgical scheduling; shorter exposure in acetazolamide arm partly due to discontinuations
  • Low-dose regimen (up to 500 mg/day) selected for tolerability — higher doses not tested and could theoretically have different efficacy
  • Primary outcome reported as absolute change rather than prespecified percentage change (post-hoc harmonization with secondary outcomes)
  • Composite gait score was a novel, study-specific measure rather than an established clinical endpoint
  • Population limited to patients with probable iNPH awaiting shunt surgery — findings may not generalize to other iNPH subgroups
  • Phase 2 trial — confirmatory phase 3 trial not feasible per authors given tolerability and risk of delaying shunt surgery

Funding

Swedish Society for Medical Research, Swedish Society of Medicine, Neuro Sweden, Region Uppsala, Thureus Foundation for Geriatric Research, and Swedish Brain Foundation

Based on: DRAIN (Lancet Neurology, 2026)

Authors: Johan Virhammar, Oskar Fasth, Maria Ekblom, ..., Dag Nyholm

Citation: Lancet Neurol 2026; 25: 550-59

Content summarized and formatted by NeuroTrials.ai.