Phase 2 study of zervimesine (CT1812) in participants with mild‐to‐moderate dementia with Lewy bodies (DLB)

Abstract

INTRODUCTION

Zervimesine is in clinical development for Alzheimer's disease and dementia with Lewy bodies (DLB). This Phase 2a study evaluated zervimesine in participants with mild‐to‐moderate DLB.

METHODS

This was a multi‐center, double‐blind, placebo‐controlled, parallel‐design study. Participants aged 50 to 85 years, with probable DLB, a Mini‐Mental State Examination score of 18 to 27, and without neurological co‐morbidities were randomized to zervimesine 100 mg, zervimesine 300 mg, or placebo for 26 weeks to assess safety and efficacy.

RESULTS

Of 293 participants screened, 130 were randomized, and 109 (83.8%) completed the study. Baseline characteristics were similar between groups. Discontinuation rates for adverse events were 4.5% for zervimesine 100 mg, 16.3% for zervimesine 300 mg, and 4.8% for placebo.

DISCUSSION

This study demonstrated that zervimesine was safe and well tolerated. The favorable efficacy profile from exploratory outcomes provides a rationale for testing zervimesine in larger studies. This study was registered at ClinicalTrials.gov: NCT05225415.

Highlights

• This study evaluated the safety and efficacy of zervimesine in Lewy body dementia.

• The incidence of adverse events was comparable to zervimesine 100 mg and placebo.

• Favorable, consistent trends were observed with zervimesine for exploratory endpoints.

1. BACKGROUND

Dementia with Lewy bodies (DLB) is the second most common type of neurodegenerative dementia after Alzheimer's disease (AD), affecting up to 5% of the general population. ¹ The incidence is estimated at ≈ 0.1% per year, but DLB may account for > 3% of new cases of dementia each year ¹ and up to 15% of all cases of dementia. ² In addition, AD and DLB share a co‐pathology of amyloid beta (Aβ) plaques in ≈ 50% of cases and up to two thirds with amyloid pathology at autopsy. ³ DLB is one form of Lewy body dementia, and is often underdiagnosed. DLB is distinguished by the early onset of cognitive, motor, sleep, autonomic, and behavioral symptoms; a more rapid progression than other dementias such as AD; and a poorer prognosis with an average duration of 5 to 8 years from symptom onset to death. ³ The symptom domains that contribute to the burden of DLB encompass changes in behavior, cognition, function, and movement. The core clinical features of DLB include visual hallucinations, parkinsonism, cognitive fluctuations, and rapid eye movement sleep behavior disorder (RBD), with supportive features of autonomic dysfunction, mood disturbance, falls, and neuroleptic sensitivity. ³ , ⁴ , ⁵ The economic burden of DLB is greater compared to other dementias as a result of higher rates of hospitalization and increased use of nursing home care. ² , ⁶ , ⁷ , ⁸ No pharmacological treatments are available to effectively treat DLB, highlighting the unmet need for novel therapies to treat across the symptom domains of DLB.

RESEARCH IN CONTEXT

1. Systematic review: PubMed searches were conducted using terms for dementia with Lewy bodies, drug treatment, any language, and randomized clinical trial that comprised 37 articles published between November 2000 and March 2025. A second PubMed search was conducted using terms for dementia with Lewy bodies, drug treatment, any language, and meta‐analysis, which retrieved 11 publications from June 2007 to July 2024. Drugs evaluated included acetylcholinesterase inhibitors, memantine, zonisamide, nilotinib, and neflamapimod. High‐level evidence for effective pharmacological strategies for treating dementia with Lewy bodies (DLB) is sparse. Individual agents demonstrate some improvement in cognition or motor function, but no drug has effects across the symptom burden experienced by those affected. Currently, no pharmacological agents are approved for treating DLB, and few drugs are in late‐stage clinical development. The paucity of high‐level clinical evidence from randomized controlled trials and evidence of efficacy and safety in patients with Alzheimer's disease provided the impetus to undertake clinical development of zervimesine for treating patients with DLB.

2. Interpretation: This exploratory study of zervimesine in patients with DLB demonstrated a safety and tolerability profile consistent with previous trials with no new safety signals. Zervimesine exhibited meaningful improvement in behavioral, functional, cognitive, and movement measures compared to placebo. These results are encouraging for those with DLB considering the paucity of evidence for other pharmacological agents.

3. Future directions: This exploratory study showed improvement with zervimesine in behavioral, cognitive, functional, and movement domains that encompass the clinical features of DLB including visual hallucinations, parkinsonism, cognitive fluctuations, rapid eye movement sleep behavior disorder, dysautonomia, and neuroleptic sensitivity. Results from this study will provide valuable insights for designing additional studies with zervimesine for treating DLB.

The key pathologic feature of DLB is the presence of cortical, subcortical, and brainstem Lewy bodies and Lewy neurites composed of α‐synuclein (α‐syn) fibrils and oligomers. Accumulation of α‐syn is linked to synaptic dysfunction, neurodegeneration, and neuronal death. In addition to Lewy bodies, at least 50% of DLB patients also have neuropathologic changes of AD, including accumulation of Aβ protein as senile plaques and neurofibrillary tangles composed of hyperphosphorylated tau protein. ⁹

Zervimesine (CT1812) is a sigma‐2 receptor modulator ¹⁰ that blocks binding and displaces Aβ and α‐syn oligomers from receptor sites to preserve neuronal synapses and normalize neuronal functioning. ¹¹ , ¹² , ¹³ In preclinical studies, zervimesine antagonized binding of Aβ oligomers to neuronal synapses, protected neuronal synapses, and restored membrane trafficking deficits induced by Aβ oligomers. ¹¹ Findings from a preclinical in vivo model of AD showing displacement of Aβ42 oligomers ¹¹ were confirmed in a clinical trial of patients with AD given a single dose of zervimesine ¹⁴ in which a time‐dependent increase in Aβ oligomers into the cerebrospinal fluid (CSF) was observed. Chronic treatment with zervimesine restored cognitive performance in a transgenic mouse model of AD. ¹¹ Further, preclinical studies demonstrated that sigma‐2 receptor antagonists can rescue neuronal dysfunction induced with human‐derived α‐syn oligomers. ¹³ Based on preclinical data and clinical trials in participants with AD, this study was undertaken to determine the safety and tolerability of exploratory analyses on the efficacy and pharmacodynamic (PD) effects of zervimesine in participants with probable mild‐to‐moderate DLB.

2. METHODS

2.1. Study design

This was a multi‐center, double‐blind, placebo‐controlled, parallel‐design clinical trial (SHIMMER) to evaluate zervimesine versus placebo in participants meeting diagnostic criteria for probable mild‐to‐moderate DLB (Figure S1 in supporting information). The study was conducted in compliance with the principles of the International Council for Harmonisation Good Clinical Practice, the Declaration of Helsinki, and all applicable regulations governing clinical trials. The study protocol, informed consent, and other relevant supporting information were submitted to an appropriate institutional review board/independent ethics committee for review and approval prior to study initiation. All participants provided written informed consent prior to any study procedures. The trial was registered at ClinicalTrials.gov: NCT05225415.

2.2. Participants

Men or women 50 to 85 years of age (inclusive) with a Mini‐Mental State Examination (MMSE) ¹⁵ score of 18 to 27 (inclusive) and meeting criteria for probable DLB ¹⁶ were eligible. All participants had a magnetic resonance imaging (MRI) or computed tomography (CT) scan (if contraindication to MRI) that was consistent with the clinical diagnosis of probable DLB and without findings of significant exclusionary abnormalities. An historical MRI or CT scan obtained up to 1 year prior to screening was allowed if there was no history of intervening neurologic disease or clinical events (e.g., stroke and head trauma) and the participant had no clinical symptoms or signs suggestive of such events. Participants agreed to undergo apolipoprotein E (APOE) genotyping at screening. Participants receiving acetylcholinesterase inhibitors (AChEI), memantine, or a combination had to be on a stable dose for at least 12 weeks before the screening visit, with no plans for dose adjustment during the study. Treatment with AChEIs or memantine from screening to the end of the study was not permitted in participants not previously taking these medications.

Exclusion criteria included any neurological condition that could contribute to cognitive impairment above and beyond those caused by DLB, including any comorbidities detected by clinical assessment or MRI or CT; screening MRI or CT scan or historical MRI/CT scan of the brain indicative of significant abnormality, including prior hemorrhage or infarct > 1 cm³, > 3 lacunar infarcts, cerebral contusion, encephalomalacia, aneurysm, vascular malformation, subdural hematoma, hydrocephalus, or space‐occupying lesion (e.g., abscess or brain tumor such as meningioma). Detailed inclusion and exclusion criteria are found in the Appendix in supporting information.

2.3. Randomization and masking

Eligible participants were randomized 1:1:1 to placebo, zervimesine 100 mg, or zervimesine 300 mg administered orally once daily. Study drug was provided as capsules of identical appearance containing either placebo or zervimesine. A non‐blinded statistician assigned to the trial generated a list with the individual study identifications for each arm so that participants were randomly assigned to either active or placebo treatment. Study participants, site personnel responsible for study drug administration and day‐to‐day study conduct, and sponsor personnel involved in day‐to‐day study conduct were blinded to individual treatment assignments and participant data.

2.4. Procedures

The study screening period was up to 42 days, followed by a double‐blind treatment period of 182 (± 2) days and a follow‐up visit at Day 210 (± 2) for a maximum of 254 days. Safety and tolerability were assessed from adverse events, physical and neurological examinations, vital signs (temperature, blood pressure, heart rate, and respiratory rate), and clinical laboratory testing (hematology, biochemistry, and urinalysis) obtained at each study visit. A 12‐lead electrocardiogram (ECG) was obtained and the Columbia Suicide Severity Rating Scale (C‐SSRS) ¹⁷ was administered at screening, baseline, and at days 28, 56, 98, 154, 182, and follow‐up during the study. Coagulation testing, serology, and a lipid panel were obtained at screening.

Exploratory efficacy outcomes were assessed at baseline and at days 28, 98, and 182 and included the Montreal Cognitive Assessment Scale (MoCA), ¹⁸ , ¹⁹ Cognitive Drug Research battery (CDRB), ²⁰ Clinician Assessment of Fluctuation (CAF), ¹⁶ Epworth Sleepiness Scale (ESS), ²¹ Movement Disorder Society—Unified Parkinson's Disease Rating Scale Part III (MDS‐UPDRS Part III), ²² , ²³ Alzheimer's Disease Cooperative Study–Clinical Global Impression of Change (ADCS‐CGIC), ²⁴ ADCS Activities of Daily Living (ADCS‐ADL), ²⁵ and the Neuropsychiatric Inventory (NPI). ²⁶

Amyloid status at screening was determined from the APS2 score (C2N Diagnostics). The APS2 score, which considers the ratio of plasma ptau217total/total tau217 and Aβ42/40 to predict the presence of brain amyloid > 25 Centiloids using positron emission tomography (PET), was used to indicate Aβ positivity. Plasma biomarkers were measured for amyloid probability score (APS2) at screening, or at baseline and day 182 for Aβ peptide 1₋42; Aβ peptide 1₋40, dopa decarboxylase, glial fibrillary acidic protein (GFAP), neurofilament light chain (NFL), phosphorylated tau181 (p‐tau181), and phosphorylated tau217 (p‐tau217; VuMC), and phospho‐α‐syn, total α‐syn (MagQu).

Phosphorylated α‐syn within cutaneous nerves was assessed by skin biopsy (Syn‐One, CND Life Sciences, Inc.). Three millimeter punch biopsies were taken from the distal leg, the distal thigh, and the posterior cervical region after local anesthesia with 1% lidocaine at baseline and week 26. Samples were placed in labeled tubes of Zamboni fixative, shipped to CND and immunostained for protein gene product 9.5 and phosphorylated α‐syn (P‐syn). A total composite Score was derived across the three biopsy sites calculated as the sum of the posterior cervical P‐syn composite score, the distal thigh P‐syn composite score, and the distal leg P‐syn composite score.

2.5. Outcomes

The primary outcome was safety and tolerability, assessed from adverse events, clinical laboratory testing, ECG, vital signs, and the C‐SSRS obtained at regular intervals. Exploratory efficacy endpoints included the MoCA, NPI‐12, CDRB, CAF, MDS‐UPDRS Part III, ESS, ADCS‐CGIC, and ADCS‐ADL. Change from baseline and observed values of plasma biomarker concentrations were assessed as an exploratory endpoint (Table S1 in supporting information).

2.6. Statistical analysis

The sample size of the study was not based on statistical considerations but was chosen to provide preliminary information regarding the safety and efficacy of zervimesine when administered according to the protocol. The MoCA was considered a key exploratory endpoint. For comparison of combined zervimesine groups versus placebo, a total sample size of 105 participants had an 80% power to detect a mean difference of 0.78 points in the change from baseline in MoCA score assuming a standard deviation (SD) of 1.32 points (PASS 2020: two‐sample t test, α = 0.05). To account for potential dropouts, a sample size of 120 participants (40 participants per treatment arm) was planned. If the treatment group comparisons versus placebo on the MoCA were non‐significant, then group comparisons on the other endpoints would be reported with nominal p values with no additional adjustments for multiplicity.

Baseline and both observed and mean change from baseline values for day 28 (week 4), day 98 (week 14), and day 182 (week 26), were summarized by treatment group using descriptive statistics. The difference between treatment groups (combined zervimesine vs. placebo) was assessed using a mixed model for repeated measures (MMRM) analysis, with change from baseline values at day 28, day 98, and day 182 as the response values, treatment group (zervimesine 100 mg, zervimesine 300 mg, placebo), time point, and a treatment group‐by‐time point interaction as fixed effects, participant as a random effect, and baseline value as a covariate. An unstructured covariance structure was applied, and degrees of freedom for the denominator were computed using the Kenward–Roger method. Least square (LS) means, standard errors, LS mean differences between combined zervimesine and placebo, 95% confidence intervals (CIs) for the LS mean differences, and p values at each post‐baseline time point were determined. Pairwise comparisons of each zervimesine group and placebo were performed from the same model in a similar manner. Pre‐planned analyses included subgroups by age, sex, baseline MMSE, baseline MDS‐UPDRS, baseline Hoehn and Yahr scale, ²⁷ AChEI/memantine use, baseline APOE status, baseline skin biopsy total composite score, baseline plasma p‐tau217/tau217 ratio, and amyloid positivity at screening (Table S2 in supporting information). Percent slowing was calculated as:

[1‐ (LS mean change for zervimesine + LS mean change for placebo)] × 100.

The intent‐to‐treat (ITT) population included all participants who were randomly assigned to the study drug. The safety population included all participants in the ITT population who received at least one dose of the study drug. The completers/compliers population included all randomized participants who completed the study and had between 80% and 120% compliance with the study drug. The pharmacokinetic (PK) population included all participants in the safety population who had at least one quantifiable plasma concentration of zervimesine. The PD population included all participants in the safety population who had an exploratory biomarker evaluation. The skin biopsy population included all participants in the safety population with Syn‐One Test skin biopsy results. The change from baseline to the end of the study is reported, and each dose group versus placebo is compared using an analysis of covariance with the baseline sum total composite score as a covariate.

2.7. Role of the funding source

The sponsor of this study in collaboration with authors had a role in study design, data collection, data analysis, data interpretation, and writing of the report.

3. RESULTS

Between May 19, 2022 and November 25, 2024, 293 participants were screened at 31 sites in the United States, 130 participants were randomized and included in the ITT population, and 109 (83.8%) participants completed the study (Figure 1). Study discontinuations occurred for adverse events in 15 (11.5%) participants, physician decision in 1 (0.8%) participant, and withdrawal by participant or caregiver in 5 (3.8%) participants. At baseline, mean (SD) age was 72.8 (6.7) years, 106 (81.5%) participants were male, 119 (91.5%) were White, and mean (SD) time since probable diagnosis of DLB was 2.0 (1.9) years. A positive skin biopsy was found in 80% of participants, and the mean skin biopsy composite score ranged from 6.3 to 8.9. Treatment groups were comparable for baseline demographic and clinical characteristics (Table 1).

FIGURE 1.

Participant disposition. AChE, acetylcholinesterase; MMSE, Mini‐Mental State Examination

TABLE 1.

Baseline characteristics (intent‐to‐treat [ITT] population).

	Zervimesine 100 mg (N = 44)	Zervimesine 300 mg (N = 44)	Placebo (N = 42)
Age, years	72.6 ± 7.8	72.1 ± 5.9	73.7 ± 6.3
Range, years	53–85	58–85	55–83
Male, n (%)	35 (79.5)	38 (86.4)	33 (78.6)
Not Hispanic or Latino, n (%)	43 (97.7)	44 (100)	39 (92.9)
Race, n (%)
White	42 (95.5)	39 (88.6)	38 (90.5)
Asian	0	3 (6.8)	0
Black or African American	1 (2.3)	0	1 (2.4)
Not reported	1 (2.3)	2 (4.5)	3 (7.1)
Weight, kg	78.2 (16.5)	83.3 (14.2)	79.5 (18.0)
Body mass index, kg/m2	26.5 ± 4.5	26.8 ± 3.4	26.4 ± 4.4
Time since DLB diagnosis, years	2.0 ± 1.8	2.1 ± 2.1	1.9 ± 1.9
Geriatric depression scale score	3.0 ± 2.4	3.2 ± 2.7	3.2 ± 1.8
Mini‐Mental State Examination	24.6 ± 2.6	23.6 ± 2.6	23.8 ± 2.7
Hoehn and Yahr stage	2.1 ± 0.9	2.1 ± 0.8	2.0 ± 0.7
Stage > 2, n (%)	20 (45.5)	21 (47.7)	14 (33.3)
Apolipoprotein E4 genotype positive, n (%)	17 (38.6)	21 (47.7)	16 (38.1)
Amyloid probability score 2 (APS2)	38.4 ± 28.6; n = 30	41.2 ± 32.1; n = 32	41.3 ± 30.8; n = 32
APS2 positive, n (%)	12 (27.3)	11 (25.0)	15 (35.7)
Plasma p‐tau217/tau217 ratio	3.3 ± 2.4; n = 34	4.2 ± 3.6; n = 33	3.6 ± 2.6; n = 33
Skin biopsy total composite score	8.9 ± 6.3; n = 42	6.7 ± 5.7; n = 42	6.3 ± 5.5; n = 40
Skin biopsy positive, n (%)	38 (86.4)	35 (79.5)	31 (73.8)
AChE inhibitor and/or memantine use, n (%)	36 (81.8)	36 (81.8)	35 (83.3)
Montreal Cognitive Assessment scale	19.5 ± 4.3	17.8 ± 5.4; n = 43	17.9 ± 4.6
Epworth Sleepiness Scale total score	7.9 ± 4.1	9.2 ± 4.4; n = 43	8.2 ± 5.2
Clinician Assessment of Fluctuation severity	4.8 ± 3.8	5.9 ± 3.4; n = 43	4.2 ± 3.4
ADCS‐ADL total score	62.7 ± 10.3	60.7 ± 12.9; n = 43	63.3 ± 9.8
MDS‐UPDRS3 total score	29.2 ± 13.9	25.4 ± 13.0; n = 43	28.1 ± 13.4
Power of attention composite score	1817 ± 316; n = 43	1878 ± 543; n = 43	1829 ± 470; n = 41
Quality of episodic secondary memory composite score	71.2 ± 45.2	68.6 ± 59.0; n = 43	61.3 ± 45.6
Quality of memory composite score	178.2 ± 85.3	180.6 ± 104.9; n = 43	173.9 ± 79.2
Quality of working memory composite score	1.1 ± 0.6	1.2 ± 0.6; n = 43	1.2 ± 0.5
Speed of memory composite score	7066 ± 2331	7796 ± 4129; n = 43	6962 ± 2431
Neuropsychiatric Inventory 12	12.0 ± 11.7	13.8 ± 12.3; n = 43	10.0 ± 11.0
Neuropsychiatric Inventory 10	8.5 ± 8.0	10.7 ± 10.0; n = 43	6.6 ± 10.2
Neuropsychiatric Inventory 12 Distress score	6.9 ± 6.1	6.7 ± 6.4; n = 43	5.0 ± 5.5
Neuropsychiatric Inventory 10 Distress score	5.5 ± 5.1	5.7 ± 5.5; n = 43	3.8 ± 5.3

Note: Values are mean ± standard deviation unless noted.

Abbreviations: AChE, acetylcholinesterase; ADCS‐ADL, Alzheimer's Disease Cooperative Study—Activities of Daily Living; DLB, dementia with Lewy bodies; MDS‐UPDRS3, Movement Disorder Society—United Parkinson's Disease Rating Scale, Part III; p‐tau, phosphorylated tau.

3.1. Safety endpoints

At least one treatment‐emergent adverse event (TEAE) occurred in 42 (95.5%) participants with zervimesine 100 mg, 40 (93.0%) with zervimesine 300 mg, and 37 (88.1%) with placebo and were of mild or moderate severity in 109 (84.5%) participants (Table 2). The most common TEAEs with zervimesine were fall (24.1%), headache (12.6%), lipase increased (13.8%), diarrhea (10.3%), and dizziness (8.0%). TEAEs were designated as related or possibly related to study drug in 51 (39.5%) participants. TEAEs were classified as serious in 4 (9.1%) participants with zervimesine 100 mg, 5 (11.6%) participants with zervimesine 300 mg, and 8 (19.0%) with placebo. Metabolic encephalopathy in one participant with zervimesine 300 mg was the only serious adverse event that was considered related to therapy. Asymptomatic elevations in hepatic transaminases greater than three times the upper limit of normal (ULN) occurred in 3 (6.8%) participants with zervimesine 100 mg and 6 (14.0%) participants with zervimesine 300 mg. All events resolved with discontinuation of zervimesine and without evidence for serious liver injury. Discontinuation for treatment‐related TEAEs occurred with 2 (4.5%) participants with zervimesine 100 mg, 7 (16.3%) participants with zervimesine 300 mg, and 2 (4.8%) participants with placebo. Three (2.3%) deaths occurred, all unrelated to treatment. The most common TEAEs were fall in 31 (24.0%) participants, headache in 19 (14.7%), lipase increased in 18 (14.0%), and urinary tract infection in 14 (10.9%). No other remarkable changes or shifts in chemistry, hematology, or urinalysis were observed and there were no clinically significant changes in vital signs. Physical examination results for most of the body systems were normal or not clinically significant. In cases that were abnormal and clinically significant, they were noted most commonly in neurological and psychiatric systems.

TABLE 2.

Treatment‐emergent adverse events (safety population).

	Number (%) of subjects/number of events
	Zervimesine 100 mg (N = 44)	Zervimesine 300 mg (N = 43)	Placebo (N = 42)
Any TEAE	42 (95.5)	40 (93.0)	37 (88.1)
Any treatment‐related TEAE	14 (31.8)	21 (48.8)	16 (38.1)
Any serious TEAE	4 (9.1)	5 (11.6)	8 (19.0)
Any treatment‐related serious TEAE	0	1 (2.3)	0
Any deaths from TEAE	0	2 (4.7)	0
Any severe TEAE	1 (2.3)	4 (9.3)	5 (11.9)
Any treatment‐related severe TEAE	0	1 (2.3)	0
TEAE causing study drug discontinuation	4 (9.1)	9 (20.9)	5 (11.9)
Treatment‐related TEAE causing discontinuation	2 (4.5)	7 (16.3)	2 (4.8)
Incidence of individual TEAEs occurring in at least 5 patients
Fall	7 (15.9)/15	14 (32.6)21	10 (23.8)/19
Headache	4 (9.1)/6	7 (16.3)/12	8 (19.0/9
Lipase increased	5 (11.4)/6	7 (16.3)/10	6 (14.3)/6
Urinary tract infection	3 (6.8)/4	3 (7.0)/6	8 (19.0)/12
Dizziness	3 (6.8)/3	4 (9.3)/6	5 (11.9)/5
COVID‐19	3 (6.8)/3	5 (11.6)/5	3 (7.1)/3
Diarrhea	4 (9.1)/4	5 (11.6)/5	2 (4.9)/2
Fatigue	4 (9.1)/5	4 (9.3)/4	3 (7.1)/3
Alanine aminotransferase increased	3 (6.8)/3	7 (16.3)/7	0
Constipation	2 (4.5)/4	4 (9.3)/4	4 (9.5)/4
Anxiety	3 (6.8)/3	3 (7.0)/3	3 (7.1)/3
Aspartate aminotransferase increased	4 (9.1)/4	5 (11.6)/5	0
Confusional state	1 (2.3)/1	5 (11.6)/5	3 (7.1)/3
Abdominal discomfort	1 (2.3)/1	5 (11.6)/5	0
Back pain	3 (6.8)/3	2 (4.7)/2	1 (2.4)/1
Agitation	1 (2.3)/1	2 (4.7)/2	2 (4.8)/3
Creatine phosphokinase increased	2 (4.5)/2	3 (7.0)/4	0
Hallucination	1 (2.3)/1	1 (2.3)/1	3 (7.1)/3
Hypertension	2 (4.5)/3	1 (2.3)/1	2 (4.8)/2
Orthostatic hypotension	1 (2.3)/2	2 (4.7)/2	2 (4.8)/2
Somnolence	3 (6.0)/3	2 (4.7)/3	0
Upper respiratory tract infection	3 (6.8)/3	1 (2.3)/1	1 (2.4)/1

Abbreviation: TEAE, treatment‐emergent adverse event.

Six ,   participants had ECG abnormalities reported during the study, three in the zervimesine 100 mg group, one in the zervimesine 300 mg group, and two in the placebo group. Most were considered unrelated to the study drug with the exception of an event of bundle branch block (zervimesine 100 mg) and an event of ECG T‐wave abnormality (zervimesine 300 mg) as possibly related to the zervimesine. One (2.3%) participant with zervimesine 100 mg, 2 (4.7%) with zervimesine 300 mg, and 2 (4.8%) with placebo reported suicidal ideation at any time during the study, but no suicidal behavior was reported.

3.2. Exploratory efficacy endpoints

Impact on cognition was assessed with the MoCA and CDRB (Table 3). Shown in Table 3 are LS mean treatment differences between zervimesine and placebo for the individual and combined doses (100 mg, 300 mg, and 100 + 300 mg). While not statistically significant, numerical treatment differences favoring zervimesine were observed on the MoCA and CDRB memory items (Table 3). Percent slowing of the pooled zervimesine dose groups versus placebo on the memory items of the CDRB were calculated and shown in Figure 2 along with the changes over time observed in the specific domains that were affected the most.

TABLE 3.

Change from baseline to week 26 for exploratory efficacy endpoints (intent‐to‐treat [ITT] population).

	Zervimesine 100 mg (N = 44)	Zervimesine 300 mg (N = 44)	Pooled zervimesine (N = 88)	Placebo (N = 42)
Montreal Cognitive Assessment
LS mean change from baseline, 95% CI	0.04 (–1.0, 1.1)	−0.55 (–1.6, 0.5)	−0.26 (–1.0, 0.5)	−0.65 (–1.7, 0.4)
LS mean difference from placebo, 95% CI	0.69 (–0.8, 2.2)	0.10 (–1.4, 1.6)	0.39 (–0.9, 1.7)
p value	0.361	0.897	0.550
CDR Composite Score Quality of Memory
LS mean change from baseline, 95% CI	−11.1 (–35.0, 12.8)	−7.7 (–32.9, 17.5)	−9.4 (–26.8, 7.9)	−33.7 (–59.0, –8.4)
LS mean difference from placebo, 95% CI	22.6 (–12.2, 57.3)	25.9 (–9.8, 18.0)	24.3 (–6.4, 54.9)
p value	0.201	0.153	0.120
Quality of Episodic Secondary Memory
LS mean change from baseline, 95% CI	−3.1 (–15.7, 9.5)	−1.1 (–14.5, 12.4)	−2.1 (–11.3, 7.1)	−13.9 (–27.2, –0.5)
LS mean difference from placebo, 95% CI	10.7 (–7.6, 29.1)	12.8 (–6.1, 31.7)	11.8 (–4.4, 28.0)
p value	0.248	0.183	0.153
Quality of Memory
LS mean change from baseline, 95% CI	−11.1 (–35.1, 12.8)	−7.7 (–32.9, 17.5)	−9.4 (–26.8, 7.9)	−33.7 (–59.0, –8.4)
LS mean difference from placebo, 95% CI	22.6 (–12.2, 57.3)	25.9 (–9.8, 61.6)	24.3 (–6.4, 54.9)
p value	0.201	0.153	0.120
Quality of Working Memory
LS mean change from baseline, 95% CI	−0.07 (–0.25, 0.11)	−0.07 (–0.26, 0.12)	−0.07 (–0.20, 0.06)	−0.07 (–0.36, 0.02)
LS mean difference from placebo, 95% CI	0.10 (–0.17, 0.36)	0.10 (–0.17, 0.37)	0.10 (–0.13, 0.33)
p value	0.471	0.464	0.403
Speed of Memory
LS mean change from baseline, 95% CI	522 (–609, 1652)	1984 (785, 3182)	1253 (429, 2076)	974 (–221, 2168)
LS mean difference from placebo, 95% CI	−452 (–2096, 1193)	1010 (–684, 2704)	279 (–1173, 1731)
p value	0.587	0.240	0.704
Power of Attention
LS mean change from baseline, 95% CI	120 (–182, 421)	621 (308, 934)	370 (153, 588)	280 (–31, 590)
LS mean difference from placebo, 95% CI	−160 (–592, 273)	341 (–100, 782)	90.7 (–288, 470)
p value	0.465	0.128	0.636
NPI‐12
LS mean change from baseline, 95% CI	0.9 (–3.1, 4.9)	0.7 (–3.6, 5.0)	0.8 (–2.1, 3.8)	5.8 (1.7, 10.0)
LS mean difference from placebo, 95% CI	−4.9 (–10.6, 0.9)	−5.1 (–11.1, 0.9)	−5.0 (–10.1, 0.1)
p value	0.097	0.093	0.055
NPI‐12 (Distress)
LS mean change from baseline, 95% CI	−0.4 (–2.5, 1.7)	−0.5 (–2.8, 1.8)	−0.4 (–2.0, 1.1)	2.7 (0.4, 4.9)
LS mean difference from placebo, 95% CI	−3.1 (–6.2, 0)	−3.2 (–6.3, 0)	−3.1 (–5.9, –0.4)
p value	0.051	0.053	0.025
Clinician Assessment of Fluctuation
LS mean change from baseline, 95% CI	0.1 (–1.1, 1.4)	0 (–1.4, 1.3)	0.1 (–0.9, 1.0)	1.1 (–0.2, 2.4)
LS mean difference from placebo, 95% CI	−0.9 (–2.8, 0.9)	−1.1 (–3.0, 0.8)	−1.0 (–2.6, 0.6)
p value	0.311	0.248	0.210
Epworth Sleepiness Scale
LS mean change from baseline, 95% CI	0.6 (–0.5, 1.7)	1.1 (–0.1, 2.3)	0.9 (0, 1.7)	0.2 (–1.0, 1.4)
LS mean difference from placebo, 95% CI	0.4 (–1.2, 2.0)	0.9 (–0.7, 2.6)	0.7 (–0.8, 2.1)
p value	0.604	0.277	0.351
MDS‐UPDRS Part III
LS mean change from baseline, 95% CI	1.6 (–1.9, 5.0)	2.1 (–1.6, 5.7)	1.8 (–0.7, 4.3)	4.7 (1.2, 8.3)
LS mean difference from placebo, 95% CI	−3.2 (–8.1, 1.8)	−2.6 (–7.7, 2.5)	−2.9 (–7.3, 1.5)
p value	0.211	0.307	0.191
ADCS—Activities of Daily Living
LS mean change from baseline, 95% CI	−4.5 (–8.0, –0.9)	−3.9 (–7.7, –0.1)	−4.2 (–6.8, –1.6)	−8.7 (–12.4, –5.0)
LS mean difference from placebo, 95% CI	4.3 (–0.9, 9.4)	4.8 (‐0.5, 10.2)	4.5 (0, 9.1)
p value	0.105	0.075	0.050
ADCS—Clinical Global Impression of Change
LS mean, 95% CI	4.6 (4.3, 5.0)	4.7 (4.4, 5.1)	4.7 (4.4, 4.9)	4.9 (4.5, 5.3)
LS mean difference from placebo, 95% CI	−0.3 (–0.8, 0.2)	−0.19 (–0.7, 0.3)	−0.24 (–0.7, 0.2)
p value	0.241	0.478	0.279

Note: LS means, 95% CIs, and p values are from a mixed model for repeated measures with treatment group, time‐point, and a treatment group by time‐point interaction as fixed effects, subject as a random effect, and baseline value as a covariate. ADCS‐ADL is a 23‐item informant‐administered assessment of functional impairment for activities of daily living. Outcome measure is the total score (0–78) with lower scores indicating greater functional impairment. An increase from baseline indicates improvement. ADCS‐CGIC measures whether the effects of active treatment are substantial enough to be detected by a skilled and experienced clinician on the basis of a clinical interview and examination. An increase from baseline indicates worsening. CAF is a screening questionnaire asked of an informant to capture frequency and duration of episodes. Outcome measure is the score (0–16) with higher scores indicating more severe fluctuations. An increase from baseline indicates worsening. For CDR, an increase from baseline indicates improvement for Continuity of Attention, Quality of Episodic Secondary Memory, Quality of Memory, Quality of Working Memory. A decrease from baseline indicates improvement for Cognitive Reaction Time, Power of Attention, Speed or Memory, and Response Variability. ESS assesses subjective sleepiness over the prior 2 weeks. An increase from baseline indicates increased sleepiness. MDS‐UPDRS covers 18 motor signs associated with parkinsonism covering bradykinesia, rigidity, tremor, and gait. Outcome measure is the score (0–136) with higher scores indicating more severe symptoms. An increase from baseline indicates worsening. MoCA is a brief dementia screening assessment that covers nine domains. Outcome measure is the score (0–30) with lower scores meaning more impairment. An increase from baseline indicates improvement. NPI assesses common behaviors associated with dementia. Outcome measure is the total score (0–12) with higher score indicating worsen symptoms. An increase from baseline indicates worsening.

Abbreviations: ADCS‐ADL, Alzheimer's Disease Cooperative Study—Activities of Daily Living; ADCS‐CGIC, Alzheimer's Disease Cooperative Study—Clinical Global Impression of Change; CAF, Clinician Assessment of Fluctuation; CDR, Cognitive Drug Research Battery; CI, confidence interval; ESS, Epworth Sleepiness Scale; LS, least squares; MDS‐UPDRS, Movement Disorder Society—United Parkinson's Disease Rating Scale; MoCA, Montreal Cognitive Assessment; NPI, Neuropsychiatric Inventory.

FIGURE 2.

Percent slowing for zervimesine 100 + 300 mg versus placebo for Cognitive Drug Research battery (CDRB) composite score (A) and change in CDRB domains (B, C, and D; intent‐to‐treat [ITT] population). CI, confidence interval

The NPI, CAF, and ESS were used to assess behavior. Importantly, percent slowing from baseline to week 26 across the exploratory efficacy endpoints consistently favored pooled zervimesine versus placebo (Figure 3). At week 26, LS mean (95% CI) differences versus placebo trended favorably for the NPI‐12 with differences of –5.0 (–10.1, 0.1) with zervimesine 100 + 300 mg (p = 0.055; Table 3). LS mean changes also favored zervimesine versus placebo for most domains of the NPI‐12 (Figure 4). A similarly favorable result was observed for the NPI‐12 assessment of caregiver distress, with LS mean (95% CI) difference versus placebo of –3.1 (–5.9, –0.4) with zervimesine 100 + 300 mg (p = 0.025; Table 3). LS mean changes also favored zervimesine versus placebo for its domains (Figure 4). Cognitive fluctuations on the CAF (inconsistent/reduced responsiveness, variable attention, altered consciousness) were reduced with LS mean (95% CI) differences versus placebo of –1.0 (–2.6, 0.6) with zervimesine 100 + 300 mg, p = 0.210, at week 26 (Table 3, Figure 5). Minimal changes from baseline to week 26 were observed with zervimesine and placebo for ESS (Table 3, Figure S2 in supporting information).

FIGURE 3.

Percent slowing at week 26 for exploratory endpoints for pooled zervimesine versus placebo. ADCS‐ADL, Alzheimer's Disease Cooperative Study—Activities of Daily Living; ADCS‐CGIC, Alzheimer's Disease Cooperative Study—Clinical Global Impression of Change; CAF, Clinician Assessment of Fluctuation; CDR, Cognitive Drug Research; CI, confidence interval; MDS‐UPDRS3, Movement Disorder Society—Unified Parkinson's Disease Rating Scale Part III; MoCA, Montreal Cognitive Assessment; NPI, Neuropsychiatric Inventory

FIGURE 4.

Least squares (LS) mean change from baseline to week 26 for the NPI‐12 frequency × severity and subscales (top) and Distress and subscales (bottom; intent‐to‐treat [ITT] population). CI, confidence interval; NPI, Neuropsychiatric Inventory

FIGURE 5.

Least squares (LS) mean change from baseline to week 26 for Clinician's Assessment of Fluctuation (top), ADCS‐ADL (middle), and MDS‐UPDRS 3 (bottom; intent‐to‐treat [ITT] population). ADCS‐ADL, Alzheimer's Disease Cooperative Study—Activities of Daily Living; MDS‐UPDRS3, Movement Disorder Society—Unified Parkinson's Disease Rating Scale Part III; SE, standard error

The ADCS‐ADL and MDS‐UPDRS Part III, were used to assess function. For ADCS‐ADL, LS mean (95% CI) difference versus placebo on week 26 was 4.5 (0, 9.1) with zervimesine 100 + 300 mg, p = 0.050 (Table 3 and Figure 5). Subgroup/sensitivity analysis for ADCS‐ADL showed similar effects in most subgroups but should be interpreted with caution due to small sample sizes (Figure S3 in supporting information). On the MDS‐UPDRS Part III, LS mean (95% CI) difference versus placebo on week 26 was –2.9 (–7.3, 1.5) with zervimesine 100 + 300 mg, p = 0.191 (Table 3). Change in motor function measured by MDS‐UPDRS Part III with zervimesine versus placebo over time is shown graphically (Figure 5). The global measure of ADCS‐CGIC showed a numerical treatment difference at week 26 versus placebo for the two individual dose zervimesine groups and pooled dose groups that were modest but similarly favored zervimesine (Table 3).

3.3. Exploratory biomarker endpoints

No significant treatment differences were observed for the pre‐specified subgroups between zervimesine and placebo for plasma biomarkers. Change from baseline in the placebo, zervimesine 100 mg, zervimesine 300 mg, and combined dose groups (100 and 300 mg), and difference from placebo are shown in Table S3 in supporting information. Baseline scores are reported in Table 1. The LS mean (SD) P‐syn composite score (um²/mm²) from the skin biopsies increased from baseline to week 26 by 1.8 (0.64), 2.7 (0.61), 1.7 (0.65), and 2.2 (0.44) for the placebo, zervimesine 100 mg, zervimesine 300 mg, and combined dose groups, respectively. Each of these group changes from baseline was statistically significant with p < 0.01. There were no significant differences between groups or compared to placebo. CSF samples were not analyzed for biomarkers due to insufficient numbers to provide meaningful data; only 10 zervimesine‐treated participants (6 at 100 mg and 4 at 300 mg) provided samples at both baseline and week 26, as this was optional per the protocol.

4. DISCUSSION

The results from this exploratory Phase 2a trial confirmed the primary objective that zervimesine was safe and well tolerated in DLB participants, consistent with results from other clinical trials with zervimesine in AD participants. ²⁸ , ²⁹ , ³⁰ , ³¹ The overall incidence of adverse events was similar with the 100 mg zervimesine and placebo groups, with higher rates of discontinuation in the 300 mg dose group. While elevated alanine transaminase and aspartate transaminase levels occurred with zervimesine, these effects were dose related and reversible with study drug discontinuation without clinical sequelae or signals of serious liver injury (meaningful elevations in bilirubin or alkaline phosphatase or signs and symptoms such as right upper quadrant pain, fatigue, and flu‐like symptoms). Future studies with zervimesine will allow participants to continue dosing past 3 × ULN. Given the differences in the frequency of asymptomatic reversible liver function elevations with similar efficacy at the two doses tested, future studies should focus on 100 mg or lower doses. Zervimesine also was associated with a low risk of treatment‐related serious adverse events, and no treatment‐related deaths occurred. Discontinuation rates were similar for the 100 mg dose and placebo, and higher for the 300 mg group in which the most common reason for discontinuation was liver enzyme elevations.

COG1201 (SHIMMER) was the first study of zervimesine in individuals with DLB and was designed to explore a wide range of symptoms. Exploratory efficacy assessments favored treatment with zervimesine across all outcomes. This included favorable profiles in the NPI‐12 with reduced frequency and severity of hallucinations, anxiety, and delusions; the NPI‐12 Distress scale, indicating a reduced impact of behavioral manifestations on caregivers; the CAF reflecting a reduction in the frequency and duration of cognitive fluctuations; the MDS‐UPDRS Part III reflecting less parkinsonism; and the ADCS‐ADL scale supporting improved daily functioning. The signals in the cognitive outcome measures contained more variability but still favored treatment with zervimesine. The MoCA and ESS did not move notably from baseline in placebo or treatment groups and so provided limited additional data on the potential impact of zervimesine in this study.

This was a relatively small and first exploratory study of zervimesine in DLB and all results will require replication in larger controlled studies; however, it is notable that across all symptom domains surveyed, there was a clear trend toward slowed progression in the zervimesine treatment group. Of particular note were the NPI and NPI Distress scores, on which the zervimesine treatment groups were not changed from baseline compared to the consistent progression of the placebo‐treated individuals. These differences had nominal p values of 0.055 and 0.025, respectively, for the pooled zervimesine versus placebo. This is particularly interesting as treatment of neuropsychiatric symptoms can be challenging in the presence of hypodopaminergic symptoms. In COG1201, zervimesine slowed progression in the NPI without worsening motor symptoms as assessed by the MDS‐UPDRS Part III. The general slowing of progression in these wide‐ranging symptoms may allow people with DLB to live longer in their home and maintain a better quality of life. Overall, the safety and tolerability profile of zervimesine, along with a consistent pattern favoring treatment effects across all symptom domains studied support further study.

There are no blood‐based biomarkers specific to DLB. In this study, we assessed biomarkers that are followed in neurodegenerative conditions such as AD. We also assessed baseline and changes in P‐syn within cutaneous nerve fibers. In the zervimesine‐treated groups, there were no significant changes compared to placebo in Aβ monomers, GFAP, NfL, p‐tau181, p‐tau217, dopa decarboxylase, total α‐syn, or P‐syn (Table S3 and Figure S4 in supporting information). While there may have been some trends within these data, longer and larger trials will likely be necessary to reliably measure changes related to degeneration. Unlike AD, in which amyloid and tau can be visualized in the living human brain, in synucleinopathies it is not currently possible to visualize or accurately measure α‐syn in the human brain.

P‐syn within cutaneous nerve fibers is increasingly viewed as a potential diagnostic biomarker for central synucleinopathies while its change over time and utility as a biomarker to predict clinical progression is less clear. ³² In this study, cutaneous P‐syn was explored using the Syn‐One test, which scores both area and intensity of immunostaining. Change in α‐syn content is not part of the purported mechanism of zervimesine and indeed there were no significant differences in change from baseline between zervimesine and placebo‐treated groups (Figure S3). There were significant increases from baseline in all groups ranging between 21% and 27% over the 6 month study period. Each group had a similar magnitude of change despite different starting values. Baseline absolute value (µm²/mm² [mean, SD]) were 6.3 (5.49), 8.9 (6.33), and 6.7 (5.67) for the placebo, 100 mg zervimesine, and 300 mg zervimesine groups, respectively.

A limitation of this Phase 2a study was a relatively small sample size across the three arms, with treatment follow‐up to 182 days. Designed as a study of the safety and tolerability of zervimesine, efficacy outcomes were exploratory and intended to gather information on the optimal measures to incorporate into future clinical trials. Designing clinical trials to evaluate therapy for DLB presents several challenges and obstacles. ³³ , ³⁴ The DLB population is heterogenous and overlaps with other dementias, notably AD. ³⁵ , ³⁶ , ³⁷ Thus, identifying the right target population to study is difficult. Additionally, validated outcome measures that are specific to DLB to assess clinical and functional changes are lacking. ³⁵ , ³⁸ , ³⁹ , ⁴⁰ In addition, the demographics of our study population were predominantly White males, which is consistent with other trials in patients with DLB. ⁴¹ , ⁴²

In summary, this exploratory study of zervimesine in participants with mild‐to‐moderate DLB demonstrated a safety and tolerability profile consistent with previous clinical trials, with no new safety signals. A trend in slowing of decline across all cognitive measures was observed with zervimesine compared to placebo, including fluctuations in attention for which zervimesine slowed decline by 91%. These results support further clinical trials designed to assess the clinical and functional efficacy of zervimesine.

AUTHOR CONTRIBUTIONS

Study hypothesis and design: Anthony O. Caggiano, James E. Galvin, Jennifer F. Iaci, and Michael Grundman. Analysis and interpretation: Anthony O. Caggiano, James E. Galvin, Michael Grundman, Jennifer F. Iaci, Mary E. Hamby, Douglas W. Scharre, and Magdalena I. Tolea. Study supervision: Anthony O. Caggiano and James E. Galvin. Drafting of manuscript: Anthony O. Caggiano, James E. Galvin, Jennifer F. Iaci, and Michael Grundman. Review and revision of manuscript for intellectual content: Anthony O. Caggiano, James E. Galvin, Michael Grundman, Jennifer F. Iaci, Mary E. Hamby, Douglas W. Scharre, and Magdalena I. Tolea.

CONFLICT OF INTEREST STATEMENT

James E. Galvin: grants from NIH; consultant to Beckman Coulter, Biogen, BMS, Cognition Therapeutics, Cognivue, Eisai, Eli Lilly, GE Healthcare, Lobe, Lundbeck, Roche, Thema Medical, Cervomed, CND Life Sciences. Douglas W. Scharre: grants from NIH; consultant to Otsuka, BrainTest; advisory board and speakers bureau for Eisai, Eli Lilly; grants or contracts with payments made to Ohio State University from Eli Lilly, Insightec, Roche, Precision Medicine, Avanir, Premier Research Group, Genetech, Cerevel Therapeutics, UCB Pharma, Janssen, Vivoryon Therapeutics, Cassava, BioVie, UniQure, Cognition Therapeutics, Cognitive Research Corporation, Biogen, EIP, Cognito, Alzheimer's Disease Cooperative Study, and Alzheimer's Therapeutic Research Institute. Mary E. Hamby is an employee of and shareholder in Cognition Therapeutics, Inc. Mary E. Hamby reports grants from NIH. Michael Grundman is an employee of Global R&D Partners, LLC, receives fees from Cognition Therapeutics, and owns stock in Cognition Therapeutics. Jennifer F. Iaci is an employee of, and stockholder in, Cognition Therapeutics. Anthony O. Caggiano is an employee of and shareholder in Cognition Therapeutics, Inc. Anthony O. Caggiano reports grants from NIH. Magdalena I. Tolea: co‐investigator for this study. Author disclosures are available in the Supporting Information.

CONSENT STATEMENT

All participants and their caregivers provided written informed consent prior to any study procedures.

Supporting information

Supporting information

Supporting information

Galvin JE, Tolea MI, Scharre DW, et al. Phase 2 study of zervimesine (CT1812) in participants with mild‐to‐moderate dementia with Lewy bodies (DLB). Alzheimer's Dement. 2025;21:e71004. 10.1002/alz.71004

DATA AVAILABILITY STATEMENT

All data for this study are reported in the manuscript and Data Supplement. Additional data may be available upon written request to the sponsor. The study protocol and statistical analysis plan are available as supplemental material.