Skip to content

Clinical trials landscape

TL;DR — The VCI trial landscape is fragmented across blood-pressure implementation, small-vessel perfusion, post-stroke cognitive rehabilitation, neuromodulation, supplements/drugs, and observational biomarker cohorts. Historical VaD drug trials produced small cognitive effects without consistent function, encouraging movement toward mechanism-defined populations (Battle 2021, PMID 33704781; Markus 2022, PMID 35969390). Completed landmark trials include SPS3 cognition (NCT00059306), SPRINT MIND (NCT01206062), FINGER (NCT01041989), PASTIS (NCT02450253), LACI-2 (NCT03451591), and ASPREE cognition (NCT01038583). A live ClinicalTrials.gov v2 query on 2026-08-31 found additional current or recently active studies spanning intensive BP implementation, choline alfoscerate, butylphthalide, stimulation, and cognitive training. A repeat sweep on 2026-09-02 identified a phase-3 successor to LACI-2 (IMPACT, NCT07252544, n=3,156) that the earlier query missed. Registry status is administrative and can lag reality; "unknown" is not "completed," and a two-day-old absence claim about a registry is not evidence of absence.

Landmark completed trials (live-resolved)

NCT Question Phase Enrollment Status Key cognitive result
NCT00059306 SPS3: BP target and dual antiplatelet after lacunar stroke 3 3,020 actual completed CASI Z-score change did not differ by antiplatelet or BP target (Pearce 2014, PMID 25453457)
NCT01206062 SPRINT: intensive vs standard SBP N/A 9,361 actual completed dementia HR 0.83 (0.67–1.04); MCI HR 0.81 (0.69–0.95) (Williamson 2019, PMID 30688979)
NCT01041989 FINGER multidomain prevention N/A 1,200 estimated unknown annual NTB difference 0.022 (0.002–0.042) (Ngandu 2015, PMID 25771249)
NCT01038583 ASPREE aspirin 100 mg — 19,114 actual unknown dementia-trigger HR 1.03 (0.91–1.17) (Ryan 2020, PMID 32213642)
NCT02450253 PASTIS single-dose tadalafil in SVD 2 55 actual completed no significant subcortical CBF increase (Pauls 2022, PMID 35135037)
NCT03451591 LACI-2 ISMN ± cilostazol after lacunar stroke 2/3 363 actual completed combination composite aHR 0.58 (0.36–0.92); ISMN cognitive-impairment aOR 0.55 (0.36–0.86) (Wardlaw 2023, PMID 37222252)

Every NCT ID and field above was re-fetched from the ClinicalTrials.gov API on 2026-08-31 during independent audit. FINGER and ASPREE registry status is “unknown” despite published results, illustrating lag.

Other live-registry VCI studies (2026-08-31)

NCT Intervention/question Phase Enrollment Status
NCT04797403 patient-centered intensive BP control to minimize cognitive decline N/A 1,306 actual active, not recruiting
NCT05050604 choline alfoscerate in VCI 4 418 estimated unknown
NCT02993367 butylphthalide in vascular cognitive impairment-no dementia 2/3 200 estimated unknown
NCT06579664 intermittent theta-burst stimulation in CSVD cognitive impairment N/A 58 estimated recruiting
NCT04944004 computer-based executive training after stroke N/A 100 estimated unknown
NCT07208019 electroacupuncture for post-stroke cognitive impairment N/A 340 estimated recruiting
NCT07150169 individualized magnetic stimulation after stroke N/A 25 estimated not yet recruiting
NCT05649800 VESPAR: vascular mechanisms in stroke/depression/dementia/delirium observational 140 estimated recruiting
NCT02264899 MEMENTO vascular components of dementia N/A 332 actual completed
NCT05829421 managing daily tasks after post-stroke cognitive impairment N/A 8 actual completed
NCT04545138 cognitive plus treadmill training after stroke N/A 8 actual completed

NCT02993367 is registered as phase 2/3, not phase 2 alone. A broader condition query also returned many incidental Alzheimer, fatigue, and biorepository records; those were not tabulated unless VCI-relevant.

Live registry sweep, 2026-08-31 — what is actually running

A ClinicalTrials.gov v2 query on 2026-08-31 across the conditions vascular cognitive impairment, cerebral small vessel disease, and lacunar stroke returned the studies below in addition to those already tabulated. Every field was read from the API in this session.

NCT Study Type / phase N Status Note
NCT05734378 PRO-SVD: prognosis of cerebral small vessel disease (Bern) observational 900 est recruiting (2020-12 → 2030-11) co-primary endpoints are intracerebral haemorrhage and ischaemic stroke, not cognition; covers SVD, CAA, and CADASIL in one cohort
NCT04229056 COMPEX: computer-assisted self-training for executive function after stroke, cardiac arrest, or Parkinson disease RCT, N/A phase 307 est recruiting (2020-06 → 2026-06) active comparator is unspecific computer-based stimulation — one of the few cognitive-rehabilitation trials with a credible control
NCT03815292 MMH-MAP in MCI during early rehabilitation after ischaemic stroke (Materia Medica Holding) phase 3 276 actual completed (2018-10 → 2020-01) primary outcome is the proportion improving by ≥1 MoCA point — a responder definition with no established clinical anchor
NCT06176404 Stellate ganglion block in SVD with dysphagia and cognitive impairment RCT, N/A phase 84 actual completed (2021-02 → 2023-05) primary outcome is the Penetration-Aspiration Scale, not cognition
NCT07111559 rt-PA versus dual antiplatelet therapy in hyperacute single-perforator infarction (Nippon Medical School) phase 4 500 est recruiting (2025-08 → 2029-03) the largest lacunar-specific acute trial currently running; primary outcome is excellent functional outcome
NCT04298866 WMH subtypes on 7 T ultra-high-resolution MRI (AP-HP) imaging, N/A phase 100 est unknown (2021-03 → 2023-06) primary outcome is the proportion of patients with a different form of WMH — directly tests the heterogeneity thesis in pathophysiology
NCT06416371 Retinal vessel leakage in SVD, sub-study of Mild Stroke Study 3 (Edinburgh) observational 40 est recruiting (2025-01 → 2029-02) fluorescein angiography with ultrawide-field imaging; the accessible-surrogate junction named in D9
NCT07755631 BASIC-S integrated lifestyle intervention for vascular risk reduction feasibility RCT, N/A phase 120 est not yet recruiting (start 2026-10) primary outcomes are recruitment and retention rates only

The distribution is itself the finding. Of the studies above, one is a phase-3 drug trial and it is already complete; none of the recruiting interventional studies has a cognitive primary endpoint in a vascular-dementia population. The two largest recruiting studies are observational (PRO-SVD, n=900) or acute-stroke functional (NCT07111559, n=500). No registered phase-3 successor to LACI-2 was found under that query; a repeat query on 2026-09-02 did find one (IMPACT, NCT07252544 — see the next section), so the 2026-08-31 absence was a query-coverage artifact rather than a fact about the field. LACI-2 is itself registered as ISRCTN14911850 rather than on ClinicalTrials.gov, so the ISRCTN registry may hold further trials these sweeps cannot see.

Live registry sweep, 2026-09-02 — the phase-3 pipeline that has appeared since

A second ClinicalTrials.gov v2 query on 2026-09-02 across the same three conditions returns interventional studies the 2026-08-31 sweep did not, including the phase-3 successor whose absence that sweep recorded. The correction matters: an absence claim in a registry sweep decays within days.

NCT Study Phase N Status / dates Primary outcome
NCT07252544 IMPACT: isosorbide mononitrate and butylphthalide in acute lacunar stroke (Beijing Tiantan Hospital), 2×2 placebo-controlled 3 3,156 not yet recruiting (2025-12 → 2028-02) proportion with post-stroke disability at 6 months; co-primary safety outcome of moderate-or-worse headache
NCT07180472 Cilostazol versus aspirin or clopidogrel for stroke recurrence and SVD progression (Chengdu Medical College) 4 632 not yet recruiting (2025-09 → 2027-09) composite of new ischaemic stroke, TIA, cognitive impairment, functional disability and death at 360 days
NCT07692399 Edaravone dexborneol sublingual tablets for blood–brain barrier dysfunction in CADASIL (Huashan Hospital) 2 60 not yet recruiting (2026-07 → 2027-12) change in whole-brain BBB water-exchange rate (kw) on DP-pCASL MRI at 6 and 12 months
NCT07012629 European randomized study of antithrombotic and lipid-lowering treatment in covert brain infarcts (Aarhus) 3 1,652 recruiting (2025-11 → 2040-01) MACCE and major/fatal bleeding at 12 and 36 months
NCT04700826 DaRe2THINK: DOACs to prevent stroke, death and cognitive decline in a broader AF community (Birmingham) 4 3,000 recruiting (2021-06 → 2031-01) composite time-to-first-event at 5 years
NCT06512376 Hereditary cerebral small vessel diseases registry — trial-ready cohort (Beijing Tiantan Hospital) observational 100 active, not recruiting (2022-07 → 2027-07) epidemiological, clinical and radiographic features
NCT06933212 Mediterranean diet with extra-virgin olive oil or walnuts versus low-fat control in CADASIL and CAA (Istituto Neurologico Carlo Besta) observational 86 recruiting (2024-10 → 2027-07) stroke incidence and cognitive decline
NCT05885295 IC3: Imperial Comprehensive Cognitive Assessment in Cerebrovascular Disease observational 700 recruiting (2021-12 → 2028-05) cognition at 1 year after stroke
NCT07660263 Genomics-based multimodal prediction model for post-stroke vascular dementia observational 300 recruiting prediction model development

Four observations follow, and they revise the picture in the section above.

  1. IMPACT is the phase-3 successor to LACI-2, and it is not in Europe. A 3,156-participant 2×2 factorial of isosorbide mononitrate and butylphthalide in acute lacunar stroke takes both the LACI-2 nitrate signal and the Chinese butylphthalide literature — the two evidence streams that the treatment page describes as mutually unread — into one adequately powered design. Its primary outcome is disability rather than cognition, which continues the pattern below.
  2. Cognition is entering composites rather than standing alone. NCT07180472 places cognitive impairment inside a five-part composite; NCT04700826 places cognitive decline in its title and a composite in its endpoint. A composite that mixes stroke, disability, death and cognition cannot show that a drug protects cognition specifically, but it is how cognition is currently getting funded.
  3. The one trial with a mechanism-specific primary endpoint targets the barrier, in the monogenic model. NCT07692399 measures BBB water exchange by DP-pCASL in CADASIL — the design MINERVA pioneered, applied to the mechanism Walsh 2021 showed CADASIL does not reproduce. If the CADASIL barrier phenotype is genuinely absent, this trial's endpoint may have little to move.
  4. Covert infarcts are becoming an entry criterion. NCT07012629 randomizes people with covert brain infarcts — the population Fang 2024 and Kühne 2022 show accumulates silently — to antithrombotic and lipid strategies. Its primary outcomes are vascular events and bleeding rather than cognition, so it will bound the safety question that the covert-lesion literature raises without answering the cognitive one.

Trial-ready cohorts are also appearing on the genetic side (NCT06512376), which is the infrastructure prerequisite for the monogenic-enrichment strategy argued in inherited forms.

Development lanes

Lane Examples Principal endpoint problem
Risk reduction SPRINT MIND, PROGRESS, SPS3 BP, ASPREE long latency, competing death, non-VaD endpoints
Perfusion/endothelium PASTIS tadalafil, LACI-2 ISMN/cilostazol surrogate validity
Repurposed drugs cholinesterase inhibitors, memantine, nimodipine mixed phenotypes, small scale effects
Neuromodulation iTBS, individualized stimulation sham/blinding/durability
Rehabilitation cognitive and dual-task training transfer to daily function
Homocysteine VITATOPS B-vitamins lowered tHcy without MMSE benefit (Hankey 2013, PMID 23765945)
Biomarker cohorts MEMENTO, VESPAR causality and assay harmonization

PROGRESS assigned 6,105 people with prior stroke/TIA to perindopril ± indapamide versus placebo: dementia 6.3% vs 7.1% (RRR 12%, 95% CI −8% to 28%); cognitive decline 9.1% vs 11.0% (RRR 19%, 4–32%), driven by recurrent-stroke-associated events (Tzourio 2003, PMID 12742805). VITATOPS secondary analysis found B-vitamins lowered tHcy (10.2 vs 14.2 μmol/L) without changing MMSE (−0.22 vs −0.25; difference 0.03, 95% CI −0.13 to 0.19) in 2,214 cognitively unimpaired participants (Hankey 2013, PMID 23765945).

Historical pharmacology lesson

Eight cholinesterase-inhibitor trials enrolled 4,373 participants; mean cognitive changes were small and functional/global outcomes inconsistent (Battle 2021, PMID 33704781). Donepezil 10 mg had the largest network cognitive estimate but more adverse events. Modern trials should therefore avoid a cognition-only success definition.

SVD trial framework

FINESSE emphasizes coherent entry phenotypes, harmonized imaging, plausible duration, and pairing intermediate with clinical outcomes (Markus 2022, PMID 35969390). STRIVE-2 supplies the imaging vocabulary (Duering 2023, PMID 37236211). PRESERVE network analysis in 82 lacunar-stroke participants found intensive vs standard BP improved DTI-based weighted global efficiency (P=0.002) without changing conventional WMH, volume, or DTI histogram metrics (Pflanz 2022, PMID 35977831).

Design domain Minimum standard
Population explicit stroke/SVD/CAA and mild/major phenotype
Copathology AD biomarker characterization where feasible
Cognition domain-sensitive battery and reliable change
Function daily activity/participation measure
Imaging STRIVE-compatible acquisition and central QC
Safety stroke, ICH, syncope/falls, mortality
Registration timely status and posted results

What the field's failures have taught, trial by trial

Six well-conducted studies since 2016 have returned null or ambiguous primary results. Read together they form a design curriculum rather than a record of futility.

Trial Primary result Design lesson
COMCID (Saito 2023, PMID 38048134) cilostazol 50 mg bid for up to 96 weeks in 159 people with MCI: MMSE change −1.8 vs −1.3 for placebo, null a drug with a strong stroke-prevention signal (20 trials, 10,505 participants; McHutchison 2020, PMID 32646330) can be null for cognition. Vascular-event prevention is not a cognitive surrogate
CONIVaD (Salvadori 2021, PMID 33855653) nimodipine + choline alphoscerate vs nimodipine + placebo, 62 randomized, all endpoints null adherence is the binding constraint: >75% of assigned drug was taken by 96% for a twice-daily agent and by 15% for the thrice-daily one. Regimen frequency should be treated as a design parameter, not an afterthought
MINERVA (Brown 2024, PMID 38629936) minocycline for 3 months in 44 SVD patients: no change in microglial PET signal (RR 1.01) or BBB permeability (RR 0.97) experimental-medicine designs can falsify a mechanistic claim in 3 months and 44 patients — far cheaper than a clinical-endpoint trial that would have been null for unknowable reasons
PASTIS (Pauls 2022, PMID 35135037) and OxHARP (Webb 2024, PMID 38832504) single-dose tadalafil: no subcortical CBF effect; sildenafil: pulsatility unchanged (P=0.18) but reactivity and perfusion improved the choice of physiological primary endpoint decides whether the drug looks active. Pre-registering the wrong mechanism reads as a negative trial
rIVR + exercise 2×2 factorial (Zhang 2026, PMID 41870419) 513 randomized to exercise and/or systolic <130 mm Hg plus atorvastatin for 24 months; no effect of either or both on PACC the design closest to the vascular-dementia causal hypothesis, executed with a usual-care arm, was flatly null. This is the strongest existing challenge to the assumption that modifying vascular risk in later life modifies cognition
US POINTER (Baker 2025, PMID 40720610) structured vs self-guided multidomain lifestyle: both arms improved; difference 0.029 SD/y when the control arm is genuinely active, the effect shrinks to a fraction of what advice-controlled trials report. Cognition rising in every arm makes an inactive control uninterpretable

Two structural conclusions follow. First, the field's most informative recent trials were mechanistic, not clinical — MINERVA, OxHARP, and PASTIS each produced an interpretable answer on tens of participants, whereas the large clinical-endpoint trials produced ambiguity on hundreds. Second, every null clinical trial in this table is compatible with either "the drug does not work" or "the endpoint could not detect it," and none was designed to distinguish those. That is the gap FINESSE addresses by pairing intermediate imaging outcomes with clinical ones (Markus 2022, PMID 35969390).

Endpoint readiness: which markers could carry a trial

The MarkVCID consortium's staged validation gives the clearest available answer on which imaging measures are trial-ready and which are not (Wilcock 2021, PMID 33480172).

Candidate endpoint Strongest supporting evidence Readiness
WMH volume universally used; correlates with cognition weakest of the quantitative options — PSMD explains cognition beyond it (Luckey 2024, PMID 39569745), and WMH regression has no cognitive correlate while growth does (Bahrani 2023, PMID 37840499)
PSMD β −0.8 (95% CI −1.2 to −0.4) with general cognition in 396 participants, replicated in three independent cohorts including CHARGE (n=6,172) (Luckey 2024, PMID 39569745) biologically validated; clinical validation pending
MRI free water associated with executive composite across eight cohorts and predicted accelerated decline over ~1.3 years (P=0.0026) (Maillard 2022, PMID 36523847) the only kit with a demonstrated longitudinal prediction
Cerebrovascular reactivity prespecified hypothesis confirmed at all three sites, n=263 (Liu 2024, PMID 38951718) validated as a cross-sectional correlate; note that CVR predicts white-matter damage locally but not globally (Pham 2026, PMID 40832975)
Global SVD factor moved with intensive BP control at Cohen's d −0.40 with a dose-response (Charisis 2026, PMID 42614618) responsive to intervention; not yet linked to cognitive benefit
Patient-reported outcome CADA-PRO validated in CADASIL (α 0.95, ICC 0.88) but correlated only with mood scales in other SVD (Di Folco 2024, PMID 39234671) usable in monogenic disease only

Responsiveness to treatment and validity as a surrogate are different properties, and no marker in this table has yet been shown to mediate a cognitive treatment effect — the formal requirement for surrogacy, and the content of OQ-7.

Registry interpretation

  • Enrollment is planned or actual as recorded, not independently audited.
  • “Unknown” commonly means required updates are overdue.
  • Phase “N/A” is typical for behavioral/device studies.
  • A completed registry does not guarantee results publication.
  • Condition indexing can include broad or incidental populations.

Priority trial designs

  1. BP-target trials enriched by SVD burden but protected against orthostatic harm.
  2. Recurrent-stroke prevention trials with prespecified cognitive outcomes.
  3. Confirmatory ISMN/cilostazol trials after LACI-2 secondary signals.
  4. Biomarker-stratified symptomatic trials separating AD-positive mixed disease.
  5. Rehabilitation trials measuring independent daily function.
  6. CAA net-benefit studies for antithrombotic and anti-amyloid decisions.

Open questions

  • Will IMPACT's 2×2 design finally test the nitrate and butylphthalide signals against each other in one adequately powered population? (NCT07252544)
  • Can a composite endpoint containing cognition alongside stroke, disability and death ever demonstrate a cognition-specific effect? (NCT07180472; NCT04700826)
  • If CADASIL does not reproduce the sporadic-SVD barrier phenotype, is a BBB-permeability endpoint viable in CADASIL? (NCT07692399; Walsh 2021, PMID 34000009)
  • Which imaging marker is ready for treatment-response use? (Duering 2023, PMID 37236211; Pflanz 2022, PMID 35977831)
  • Can LACI-2 secondary signals survive a phase 3 cognitive primary endpoint? (Wardlaw 2023, PMID 37222252)
  • Can small mechanistic studies produce a credible go/no-go signal? (Markus 2022, PMID 35969390; Pauls 2022, PMID 35135037)
  • Can common data elements make rehabilitation and drug trials comparable? (Skrobot 2018, PMID 29055812)
  • Why does no currently recruiting interventional trial in this registry sweep carry a cognitive primary endpoint in a vascular-dementia population?
  • Is a ≥1-point MoCA improvement (NCT03815292) a defensible responder definition, and what would anchor it? (Battle 2021, PMID 33704781)
  • Does a single-registry sweep systematically miss European trials registered on ISRCTN, as the LACI-2 registration suggests? (Wardlaw 2023, PMID 37222252)

References

  1. Battle CE, et al. Cholinesterase inhibitors for vascular dementia and other VCI. Cochrane Database Syst Rev. 2021;2:CD013306. PMID 33704781
  2. Markus HS, et al. FINESSE framework for SVD trials. JAMA Neurol. 2022;79:1187-1198. PMID 35969390
  3. Duering M, et al. STRIVE-2. Lancet Neurol. 2023;22:602-618. PMID 37236211
  4. Skrobot OA, et al. VICCCS guidelines. Alzheimers Dement. 2018;14:280-292. PMID 29055812
  5. Pearce LA, et al. Cognitive function in SPS3. Lancet Neurol. 2014;13:1177-85. PMID 25453457
  6. Williamson JD, et al. SPRINT MIND probable dementia. JAMA. 2019;321:553-561. PMID 30688979
  7. Ngandu T, et al. FINGER. Lancet. 2015;385:2255-63. PMID 25771249
  8. Ryan J, et al. ASPREE aspirin and dementia. Neurology. 2020;95:e320-e331. PMID 32213642
  9. Pauls MMH, et al. PASTIS tadalafil. Alzheimers Dement. 2022;18:2393-2402. PMID 35135037
  10. Wardlaw JM, et al. LACI-2. JAMA Neurol. 2023;80:682-692. PMID 37222252
  11. Tzourio C, et al. PROGRESS dementia and cognition. Arch Intern Med. 2003;163:1069-75. PMID 12742805
  12. Hankey GJ, et al. VITATOPS cognitive secondary analysis. Stroke. 2013;44:2232-9. PMID 23765945
  13. Pflanz CP, et al. PRESERVE network integrity and BP intensity. Neurology. 2022;99:e1945-e1953. PMID 35977831
  14. Saito S, et al. Efficacy and safety of cilostazol in mild cognitive impairment: a randomized clinical trial (COMCID). JAMA Netw Open. 2023;6:e2344938. PMID 38048134
  15. McHutchison C, et al. Cilostazol for secondary prevention of stroke and cognitive decline. Stroke. 2020;51:2374-2385. PMID 32646330
  16. Salvadori E, et al. Nimodipine and choline alphoscerate in cognitive impairment due to cerebral small vessel disease: the CONIVaD trial. Drugs Aging. 2021;38:481-491. PMID 33855653
  17. Brown RB, et al. MINERVA: minocycline to reduce inflammation and blood-brain barrier leakage in small vessel disease. Alzheimers Dement. 2024;20:3852-3863. PMID 38629936
  18. Webb AJS, et al. Cerebrovascular effects of sildenafil in small vessel disease: the OxHARP trial. Circ Res. 2024;135:320-331. PMID 38832504
  19. Zhang R, et al. Effects of exercise and intensive vascular risk reduction on cognitive function in older adults. JAMA Neurol. 2026;83:424-434. PMID 41870419
  20. Baker LD, et al. Structured vs self-guided multidomain lifestyle interventions: the US POINTER randomized clinical trial. JAMA. 2025;334:681-691. PMID 40720610
  21. Wilcock D, et al. MarkVCID cerebral small vessel consortium: I. Enrollment, clinical, fluid protocols. Alzheimers Dement. 2021;17:704-715. PMID 33480172
  22. Luckey AM, et al. Biological validation of peak-width of skeletonized mean diffusivity as a VCID biomarker. Alzheimers Dement. 2024;20:8814-8824. PMID 39569745
  23. Bahrani AA, et al. Reliability and construct validity of the MarkVCID white matter hyperintensity growth and regression protocol. J Alzheimers Dis. 2023;96:683-693. PMID 37840499
  24. Maillard P, et al. MRI free water as a biomarker for cognitive performance: validation in the MarkVCID consortium. Alzheimers Dement (Amst). 2022;14:e12362. PMID 36523847
  25. Liu P, et al. Cerebrovascular reactivity MRI as a biomarker for SVD-related cognitive decline: multi-site validation in MarkVCID. Alzheimers Dement. 2024;20:5281-5289. PMID 38951718
  26. Pham SD, et al. Small vessel dysfunction at 7T MRI locally predicts white matter damage progression in CADASIL. J Cereb Blood Flow Metab. 2026;46:521-533. PMID 40832975
  27. Charisis S, et al. Intensive versus standard blood pressure control and overall brain small vessel disease burden: post-hoc analysis of SPRINT. EClinicalMedicine. 2026;99:104143. PMID 42614618
  28. Di Folco C, et al. CADA-PRO: a patient questionnaire measuring key outcomes in CADASIL. Stroke. 2024;55:2439-2448. PMID 39234671
  29. Walsh J, et al. Microglial activation and blood-brain barrier permeability in cerebral small vessel disease. Brain. 2021;144:1361-1371. PMID 34000009