Targeting underlying causes of neurodegenerative diseases

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential L E G A L D I S C L O S U R E S Forward-Looking Statements & Disclaimers This slide deck may contain certain forward-looking information and “forward-looking statements” that are made pursuant to the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. Such information involves known and unknown risks, uncertainties and other factors that may cause actual results, performance or achievements to be materially different from those implied by statements herein, and therefore these statements should not be read as guarantees of future performance or results. Such forward-looking statements include, among others, statements pertaining to ProMIS Neurosciences Inc.'s (the “Company”) PRECISE-AD Phase 1b clinical trial, target engagement and biomarker findings, the results and nature of the blinded interim clinical data of PMN310 and anticipated topline clinical data of PMN310, its mechanism of action and potential benefits, the Company's development plans and anticipated milestone timing, market opportunity and patient populations or related projections or forecasted revenue for third parties, the Company's planned operations and cash runway, preclinical and discovery pipeline and activities related thereto. Forward-looking information is based on a number of opinions, assumptions and estimates that, while considered reasonable by the Company as of the date of this slide deck, are subject to known and unknown risks, uncertainties and assumptions and other factors that may cause the actual results, level of activity, performance or achievements to be materially different from those expressed or implied by such forward-looking information, including, but not limited to, the risk that the results of early clinical trials are not necessarily predictive of future results with PMN310, unforeseen delays in conducting and monitoring the Company's pre-clinical and clinical trials, inability to raise any additional required capital and the Company’s ability to fund its operations. All forward-looking statements are based on the Company's current beliefs as well as assumptions made by and information currently available to it, as well as other factors. Readers are cautioned not to place undue reliance on these forward-looking statements, which speak only as of the date of this slide deck. Important factors that could cause actual results to differ materially from those indicated in the forward-looking information include, among others, the factors discussed throughout the “Risk Factors” section of the Company's most recently filed Annual Report on Form 10-K for the year ended December 31, 2025 and in its subsequent filings filed with the United States Securities and Exchange Commission. Due to risks and uncertainties, including the risks and uncertainties identified by the Company in its public securities filings available online at sec.gov, actual events may differ materially from current expectations. The Company disclaims any intention or obligation to update or revise any forward-looking statements, whether as a result of new information, future events, or otherwise. Except as required by applicable securities laws, the Company undertakes no obligation to publicly update any forward-looking information, whether written or oral, that may be made from time to time, whether as a result of new information, future developments or otherwise. This presentation contains estimates and other information concerning our industry, our business and the markets for our product candidate. Unless otherwise expressly stated, this information is based on our internal estimates and research and on reports, surveys, studies and other materials prepared by market research firms and other third parties, industry, medical and general publications, government data and similar sources. Information based on estimates, market research or similar methodologies, including prevalence studies extrapolated to broader populations, is inherently subject to uncertainty. Actual events and circumstances may differ materially from those assumed. Industry publications and surveys generally state that their information comes from sources believed to be reliable. Although we believe this information is reliable as of the date of this presentation, it may prove inaccurate due to data collection methodologies, limitations on the availability and reliability of underlying data, the voluntary nature of certain data collection processes and other uncertainties. We are responsible for the accuracy of the information presented. We believe our internal research is reliable and our market definitions are appropriate; however, neither has been independently verified. Data for interim analysis presented as of July 22, 2026 2

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential W H O W E A R E 3 A clinical-stage biotech targeting the misfolded proteins that drive neurodegenerative disease Our pipeline of antibodies and vaccines are designed to selectively target the specific, disease-causing misfolded forms of proteins. BAL ANCE SHEET O UR FO CUS Protein misfolding Targeting only the toxic, misfolded forms of disease-causing proteins O UR PL AT FORM EpiSelect Computational tool that predicts the misfold and identifies the exposed epitope O UR L EAD ASSET Phase 1b PRECISE-AD trial in early Alzheimer’s Topline data expected in Q1 2027 Cash runway through 2027, well beyond Q1 2027 PRECISE-AD topline data NASDAQ: PMN · Cambridge, MA

T H E P R O B L E M 7.4 million Americans are living with Alzheimer’s today, and that number could nearly double to 13.8M by 2060 1 in 9 Americans age 65 and older has Alzheimer’s Each one is a parent, a spouse or a friend losing memory, independence and the ability to recognize the people they love. Much of their care falls on family. T H E S C A L E O F T H E P R O B L E M 13.8M Americans 65+ projected by 2060 Absent treatments that slow or prevent progression 1 in 3 Older Americans die with dementia More than breast and prostate cancer combined $409B U.S. care costs in 2026 Excluding the unpaid care given by families Our pursuit: design precision therapies against the toxic proteins that drive Alzheimer's Disease (AD) PMN310 is designed to target only toxic Aβ oligomers, an early driver of synaptic loss, neuroinflammation, and memory loss in AD. Source: Alzheimer’s Association, 2026 Alzheimer’s Disease Facts and Figures (Americans 65+). 1. Brookmeyer R et al., Alzheimer’s Dement 2018;14:121–129 (2017 estimate; amyloid and/or neurodegeneration; many will not progress to dementia). 4 ~47M Americans with preclinical Alzheimer’s¹ No symptoms yet; not addressed by current labels Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential Anti-amyloid plaque therapies already sell ~$1.4B a year despite significant safety risks and a boxed warning 5 Clinical Benefit (CDR-SB) Boxed Warning: ARIA Annual Sales Ests. (1) Leqembi (lecanemab) Approved 2023 Adjusted Mean Change from Baseline in CDR-SB in Study 2 Kisunla (donanemab) Approved 2024 CDR-SB Change From Baseline in Combined Population Through 76 Weeks in Study 1 FDA BOXED WARNING ~13% ARIA-E (brain swelling)(2) ~17% ARIA-H (microbleeds)(2) FDA BOXED WARNING ~16% ARIA-E (brain swelling)(3) ~25% ARIA-H (microbleeds)(3) 1. 2026: annualized Q2 2026 in-market sales (Leqembi $184M, Biogen; Kisunla $167M, Lilly). 2033E: GlobalData, October 2026. CAGR 2026–2033E calculated by ProMIS. 2. Leqembi Prescribing Information. 3. Kisunla Prescribing Information, July 2025; Eli Lilly press release, July 9, 2025; PharmExec. M A R K E T Demand is proven Uptake keeps growing despite the safety risk. A safer option could widen use ~$0.7B ~$2.8B 2026 2033E ~$0.7B ~$2.9B 2026 2033E 21% CAGR(1) 23% CAGR(1)

An estimated ~95% of eligible U.S. patients remain untreated, facing safety, access and cost barriers M A R K E T O P P O R T U N I T Y 6 Market opportunity ~1.26M are early-stage and label-eligible today(2) ~95% untreated (~1.2M patients) ~5% treated (~63K patients)(2) How a Top obstacles to current anti-amyloid therapies 1 ARIA (boxed warning) ARIA-E in ~13% (lecanemab) and ~16% (donanemab); 40–50% of APOE4 homozygotes experience ARIA(5) 2 Other side effects Infusion reactions in 26% (lecanemab) and 9% (donanemab)(6) 3 MRI monitoring and IV infusion Access to MRI and infusion-center infrastructure required(7) 4 Contraindications Patients on blood thinners face higher risk of brain hemorrhage; APOE4 homozygotes face the highest ARIA risk and are excluded from EU approvals(5) 5 Out-of-pocket costs Estimated ~$5–7K/yr for a Medicare patient without supplemental coverage, plus infusion and MRI fees(8) Demand is proven ~$1.4B in annualized sales despite these barriers.(1) PMN310 designed to target the biggest barriers 1. Annualized Q2 2026 global in-market sales (Leqembi $184M per Biogen; Kisunla $167M per Lilly). 2. Monfared AAT et al., Neurol Ther 2025 (Eisai-sponsored U.S. model): 1,259,263 eligible, 62,963 treated. 3. Untreated patients × Leqembi U.S. first-year list price (~$26,500/yr); illustrative, ProMIS calculation. 4. TD Cowen via Axios, Sept 28, 2026: $60B if 75% of ~3.8M U.S. MCI patients are treated at $20K/yr, $70B by 2035. 5. Leqembi and Kisunla Prescribing Information (Kisunla rev. July 2025, modified titration); Wang H et al., J Prev Alzheimers Dis 2025 (TRAILBLAZER-ALZ 6, 18-month ARIA-E 15.6%); Smith EE et al., J Prev Alzheimers Dis 2025 (APOE4 homozygote ARIA, EU exclusion). 6. van Dyck CH et al., NEJM 2023; Sims JR et al., JAMA 2023. 7. Cummings J et al., Lecanemab appropriate use recommendations, J Prev Alzheimers Dis 2023. 8. Medicare Part B 20% coinsurance × CMS Part B ASP payment limits, Q4 2026 (Leqembi ~$26.5K/yr; Kisunla ~$33K/yr, modified titration, year 1); ProMIS calculation. ~$32B/yr U.S. drug value of untreated eligible patients at list price(3) ~$60–70B analyst sizing of the next-gen U.S. MCI market(4) Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

Lead Clinical Candidate PMN310 in Alzheimer’s Disease Selectivity for Toxic Aβ Oligomers

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential Neutralizing toxic AβOs upstream can reduce downstream damage 8 1.The Amyloid Cascade Hypothesis. Castellani, Rudy J. et al. The American Journal of Pathology, Volume 195, Issue 11, Nov 2025 2.Cline EN, et al. The Amyloid-β Oligomer Hypothesis: Beginning of the Third Decade. J Alzheimers Dis. 2018;64(s1):S567–S610 D I S E A S E P R O C E S S Aβ Monomers Abundant; physiologic role Toxic Aβ Oligomers (AβO) Main driver of disease2 Supported by numerous publications misfold Tau pathology p-tau and tangles Neuroinflammation and oxidative stress Synaptic toxicity direct neuron damage Cognitive decline PMN310 acts upstream of the pathological cascade Binds toxic AβOs selectively, avoiding monomers and plaque IgG1 effector function engages microglia to clear the bound oligomers CSF oligomer reduction demonstrated in Phase 1a (sFIDA assay) Potential to inhibit the downstream processes that lead to cognitive decline aggregate

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential Toxic Aβ oligomers: low abundance, high toxicity, the target of PMN310 in Alzheimer’s disease 9 ILLUSTRATIVE RELATIVE ABUNDANCE OF Aβ SPECIES1 1. 1. Goure WF, Krafft GA, Jerecic J, Hefti F. Targeting the proper amyloid-beta neuronal toxins: a path forward for Alzheimer's disease immunotherapeutics. Alzheimers Res Ther. 2014 Jul 9 PMN310 achieved CSF concentrations 100–600x greater than the target in Ph1a T A R G E T B I O L O G Y Toxic Aβ Oligomers Small, soluble clusters of Aβ Main driver of disease Aβ Plaque Large, insoluble aggregates of Aβ Aβ Monomers Abundant, produced over lifetime Illustrative; not to scale Target: Toxic Aβ Oligomers, the primary driver of disease Present in picomolar quantities; low abundance means less antibody is needed for target saturation Avoid: sink effect and ARIA risk Antibody binding to plaque soaks up dose and can result in ARIA Avoid: sink effect Present in nanomolar quantities; monomer binding adds no efficacy and can lead to a sub-therapeutic dose PMN310

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential EpiSelectTM computational platform allows for the generation of antibodies selective for toxic misfolded proteins 10 D I S C O V E R Y P L A T F O R M Normal protein (monomers) Misfolds Toxic oligomer Conformational epitope exposed EpiSelect PMN310 binds only the toxic-form epitope Illustrations are schematic. 1. Gibbs E et al., Sci Rep 2019;9:9870. Selectivity: Kaplan JM et al., Alzheimer’s Dement TRCI 2025;11:e70184. Epitope FcγR Microglia binding FcγR captures the Fc portion Fc engages EpiSelect computational platform Predicts the misfold and identifies the exposed epitope IgG1 effector function Fc recruits microglia for clearance The most toxic AβOs are low molecular weight and unstable: their shape keeps changing Aggregates Engulfs Complex removed Microglia engulf and clear oligomers

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential PMN310 is the only IgG1 mAb designed to solely target toxic amyloid-beta (Aβ) oligomers PMN310 (ProMIS) Phase 1b ongoing Lecanemab (Eisai) Approved Donanemab (Eli Lilly) Approved Sabirnetug (Acumen) Phase 2 Target binding Plaque binding is linked to ARIA-E. Cross-reactivity reduces oligomer targeting. Toxic Aβ oligomers only Aβ Oligomers and Plaques Aβ Oligomers and Plaques Aβ Oligomers and Plaques Isotype Determines effector function; IgG1 – neutralizes, actively clears IgG2 – neutralizes only IgG1 · active clearance IgG1 · active clearance IgG1 · active clearance IgG2 · passive clearance Weak effector function ARIA-E rate Main safety risk of the class 0% ARIA-E1 4.4% ARIA-H, all mild 12.6% ARIA-E2 17.3% ARIA-H 15.6% ARIA-E3 20.3% ARIA-H ~10% ARIA-E4 Details not disclosed Sources: 1. ProMIS PRECISE-AD interim, Jul 28, 2026; 2. van Dyck et al., NEJM 2023 (Clarity AD); 3. Wang et al., Alzheimer's & Dementia 2025 and JPAD 2025 (TRAILBLAZER-ALZ 6); 4. Acumen ALTITUDE-AD topline, Oct 30, 2026; company disclosures (isotype). Note: These data are derived from different clinical trials at different points in time, with differences in trial design, including endpoints, and patient populations. As a result, cross-trial or product comparisons cannot be made, it is only provided for illustrative purposes, and no head-to-head clinical trials have been conducted. 11 D I F F E R E N T I A T I O N

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential Plaque detection (no mAb) Amylo-Glo Staining with mAb Staining with mAb Isotype negative controls Test antibodies (fluorescence) (chromogenic) PMN310 was the only mAb tested with undetectable plaque binding through fluorescent and chromogenic staining 12 Staining conducted using 1.0 μg/mL concentration of each antibody; similar results achieved at 0.1 and 0.3 μg/mL of each antibody and quantification of plaque stain fluorescent intensity across concentrations can be provided upon request ARIA-E Rates ~25% ~21% Ph1 top dose ~13% ~35% P L A Q U E B I N D I N G PMN310

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential Healthy volunteers Randomized 3:1, PMN310:Placebo (N=8 per cohort) First subject dosed Nov. 2023 Dose 2.5 mg/kg (N = 6:2) Dose 5 mg/kg (N = 6:2) Dose 10 mg/kg (N = 6:2) Dose 20 mg/kg (N = 6:2) Dose 40 mg/kg (N = 6:2) PMN310 Placebo-Controlled Study (2.5–40 mg/kg) Phase 1a first-in-human single ascending dose (SAD) study met all objectives and informed Phase 1b design and dosing 13 SAD, Single Ascending Dose; MAD, Multiple Ascending Dose; AD, Alzheimer’s Disease Per protocol, doses fixed per cohort with a 70 kg body weight assumption Phase 1a SAD Trial Design (40 healthy volunteers, 5 dose levels) Key Findings • PMN310 was generally well tolerated in healthy volunteers across all five dose levels tested • Linear dose-proportionality of PMN310 CSF levels exhibited • CSF half-life was ~27 days, supporting once-monthly dosing, and potential subcutaneous formulation • PMN310 crossed the BBB, reaching CSF concentrations of 100–600 times the estimated CSF AβO molar concentration • Achieved CSF levels believed to be required for target saturation without a brain shuttle T R I A L D E S I G N

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential T A R G E T E N G A G E M E N T PMN310 reduced Aβ oligomers in CSF in Phase 1a, measured directly by the sFIDA assay What it measures Aβ oligomer particles in CSF, the target of PMN310 How it works • Captures oligomers and detects them with a fluorescence-tagged antibody • Signal scales with particle size Why it matters • Femtomolar sensitivity • A direct pharmacodynamic readout of target engagement sFIDA = surface-based fluorescence intensity distribution analysis (attyloid GmbH, Düsseldorf); exploratory assay. 1. Kaplan et al. Poster, Alzheimer’s Association International Conference (AAIC), July 2026. Phase 1a SAD: NCT06105528, 40 healthy volunteers. 14 Phase 1a: PMN310 reaches its target in CSF1 • Oligomer reduction detected by sFIDA at Days 3 and 29, in healthy volunteers with low baseline levels First evidence consistent with PMN310’s mechanism of action Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential A A I C 2 0 2 6 · P H A S E 1 a T A R G E T E N G A G E M E N T PMN310 produced a dose-dependent reduction in CSF Aβ oligomer particles Day 3 post-dose Day 29 post-dose First human evidence of Aβ oligomer target engagement by PMN310 • Relative to placebo, the number of detectable oligomer particles declined at doses of 5–40 mg/kg across all particle sizes, at both Day 3 and Day 29 post-dose. • Exploratory assay, but the first direct pharmacodynamic evidence that PMN310 engages and removes the oligomer species it was designed to target. 15 sFIDA analysis of CSF (attyloid GmbH); number of oligomer particles by size vs. placebo. Phase 1a SAD, NCT06105528, N = 40 healthy volunteers. Exploratory endpoint; presented at AAIC 2026 (Kaplan et al.). Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

16 PRECISE-AD Interim Analysis Results Presented July 28, 2026 1 Demographics and Study Design 2 Safety Results 3 Biomarker Trends

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential PRECISE-AD: Phase 1b with FDA Fast Track Randomized, double-blind, placebo-controlled multiple ascending dose study of PMN310 in early AD 17 T R I A L D E S I G N STUDY DESIGN · STAGGERED COHORTS · 12 MONTHS EACH 6-month interim blinded 12-month topline unblinded · Q1 2027 COHORT 1 350 mg (5 mg/kg) 3:1 active : placebo · monthly IV n = 16 COHORT 2 700 mg (10 mg/kg) 3:1 active : placebo · monthly IV n = 56 COHORT 3 1400 mg (20 mg/kg) 3:1 active : placebo · monthly IV n = 72* First dose Interim cut, July 22, 2026 Topline Study timeline KEY ASSESSMENTS Powered for 6-month biomarker change; 95% confidence to detect ≥1 ARIA case Fluid biomarkers CSF & plasma · baseline, 3, 6, 9, 12 tau pathology, synaptic damage, neuroinflammation MRI for ARIA Baseline, 2, 4, 6, 9, 12 mo Amyloid PET Baseline, 12 mo Clinical outcomes Baseline, 6, 12 mo iADRS, CDR-SB, ADAS-Cog and other cognition, function and composite scales *56 planned. 128 planned overall to achieve 100 completers; 144 enrolled across 22 sites. Interim data blinded.

T R I A L D E M O G R A P H I C S PRECISE-AD Enrolled a Representative AD Population, Including the Highest ARIA-Risk Patients WHY APOE4 MATTERS ✓ Representative population, including highest ARIA-risk patients Treatment options may open for ~15% of AD patients who are APOE4 homozygotes, the group current therapies largely exclude. Source: AD population genotype percentage and allele frequencies; Yamazaki et al. Nat Rev Neurol 2019 KEY DEMOGRAPHICS · Early AD (MCI or mild dementia due to AD) · 144 enrolled across 22 sites ✓ 18 144 enrolled 136 safety-evaluable 73.3 mean age (yrs) 58% female APOE4 STATUS 61% APOE4 carriers 39% non-carrier 50% heterozygous 11% ε4/ε4 ε4/ε4 homozygotes are the highest ARIA-risk group and are largely excluded from current anti-amyloid therapies 69% White 24% Hispanic 6% Black 1% Asian Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

I N T E R I M S A F E T Y · 6 - M O N T H B L I N D E D Favorable Safety Data, Observed Across Key Measures 6-month blinded interim · N = 144 dosed, 136 safety-evaluable · a representative population 0.0% ARIA-E The most severe treatment-driven form of ARIA No treatment-related serious AEs Across all genotypes No treatment-related discontinuations Low overall dropout rate 4.4% ARIA-H In line with placebo-range ARIA-H, all mild, asymptomatic, non-serious Minimal infusion reactions A recognized liability of approved anti-amyloids and brain-shuttle candidates. PRECISE-AD: one non-serious event, non-systemic, dosing continued. PMN310 <1% 1 event, non-serious Donanemab ~9% Lecanemab ~26% APOE4 carriers included 61% carry ≥1 APOE4 allele 11% ε4/ε4 homozygotes — the highest ARIA-risk group ✓ No ARIA-E in any genotype ARIA detection using Susceptibility-Weighted Imaging (SWI) for increased sensitivity. 6-month blinded. Infusion-reaction comparators: lecanemab 26.4% (CLARITY AD, van Dyck NEJM 2023); donanemab 8.7% (TRAILBLAZER-ALZ 2, Sims JAMA 2023). Descriptive comparison only. No head-to-head studies have been conducted to compare PMN310 with approved products or other product candidates in development. Copyright 2026, ProMIS Neurosciences Inc., Non-confidential 19

T I M E - M A T C H E D S A F E T Y Cumulative ARIA-E Over Time: Zero Edema Observed in First 6 Months ARIA-E (the most serious ARIA type) has been observed to occur almost entirely in the first six months of dosing. Comparator ARIA-E plateaus after the first few months; rates are time-matched to published data where available, with ~ indicating values estimated between published time points. A T 6 M O N T H S Time-matched to estimated or published 6 -month rates Donanemab 23.7% Lecanemab ~12% Historic, third-party placebo data (avg across both trials) ~0.6%* PRECISE-AD 0% Illustrative cross-trial comparison — not head-to-head; ~ denotes values estimated between published time points. * Placebo: 6-mo ARIA-E (~0.6%) is ProMIS-calculated by linear accrual — (6/18) × ~1.9%, the average of the two trials’ 18-mo placebo rates (1.7% and 2.1%). Lecanemab: 12.6% treated / 1.7% placebo ARIA-E at 18 mo; 6-mo (~12%) approximated from this front-loaded cumulative (van Dyck et al., NEJM 2023;388:9–21). Donanemab, TRAILBLAZER-ALZ 2: 24.0% treated / 2.1% placebo over 76 wk (Sims et al., JAMA 2023;330:512–527). Donanemab standard regimen, TRAILBLAZER-ALZ 6: 23.7% ARIA-E at wk 24 (6 mo), 24.2% at 76 wk (Wang et al., Alzheimer’s & Dementia 2025;21:e70062); pooled standard-dosing 18-mo ~24.4%. PRECISE-AD: blinded pooled cohort (active + placebo, 3:1); 0/136 ARIA-E at 6-mo interim — observed, not projected. Donanemab, 23.7% 24.4% Lecanemab, ~12% 12.6% Historic Placebo ~0.6% ~1.9% PRECISE-AD, 0% 0% 10% 20% 30% 0 6 12 18 Cumulative ARIA-E incidence (%) Months on study 6-months Copyright 2026, ProMIS Neurosciences Inc., Non-confidential 20

T I M E - M A T C H E D S A F E T Y Cumulative Total ARIA Over Time: Blinded PRECISE-AD Data Tracks the Placebo Range Reported in Other Trials Approved anti-amyloids accrue ARIA steeply in the first months of dosing, with 90%+ of ARIA-E occurring in the first 6 months. Comparator rates are time-matched to published data where available, with ~ indicating values estimated between published time points. Donanemab-trial placebo Lecanemab-trial placebo P LAC E B O R AN GE 14.2% 9.3% AT 6 M O N T H S * C a l c u l a t e d b a s e d o n p u b l i s h e d d a t a Donanemab *~31% Lecanemab *~16% Historic, third-party placebo data (both trials) *~3–5% PRECISE-AD 4.4% MODELED SHAPE, ACTUAL ENDPOINTS — comparator and placebo curves are reconstructed from published trial time-course; trial-end values are as reported. Illustrative, not a head-to-head comparison. Any ARIA (ARIA-E and/or ARIA-H): Lecanemab 21.5% / 9.3% placebo at 18 mo (van Dyck, NEJM 2023); Donanemab 37.0% / 14.2% placebo, TRAILBLAZER-ALZ 2 ~76 wk (Sims, JAMA 2023; Zimmer, JAMA Neurol 2025). 6-mo points calculated from the published time-course (>90% of ARIA-E within 6 mo): Donanemab ~31%, Lecanemab ~16%, placebo ~3–5%; placebo accrues linearly to reported endpoints. PRECISE-AD: blinded pooled cohort (active + placebo, 3:1); 4.4% total ARIA (6/136) at 6-mo interim — observed, not projected. 21 observed interim — no projection 0% 10% 20% 30% 40% 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Months on study · cumulative total ARIA incidence (%) Donanemab 37.0% Lecanemab 21.5% PRECISE-AD 4.4% Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

22 PRECISE-AD Interim Biomarker Results

B I O M A R K E R S R E P O R T E D Two tau biomarkers reported, both expected to rise without treatment 23 Plasma pTau217 BLOOD-BASED W HAT I T M EASURES Disease progression: a downstream marker of amyloid-positive AD pathology and tau pathway activation. I N UNTREATED PATI ENTS In untreated (placebo) patients, pTau217 rises ~6% over 18 months1 . CSF MTBR-tau243 CEREBROSPINAL FLUID W HAT I T M EASURES A CSF marker of tauopathy, the pathology most closely tied to symptoms. Correlates with tau-tangle burden and cognition comparable to tau-PET, and more strongly than other CSF markers. I N UNTREATED PATI ENTS MTBR-tau243 rises as tangles accumulate and moves in step with disease progression over time. In untreated patients, MTBR-tau243 is therefore expected to trend upward. PRECISE-AD interim results on the following slides are pooled and blinded. 1. Pontecorvo et al., JAMA Neurology 2022 (TRAILBLAZER-ALZ Ph2 donanemab trial, placebo-arm natural history). Horie et al., Nature Medicine 2023. 6% 0% 5% 10% Baseline 6 mo 12 mo 18 mo NATUR AL HI STO RY PRO G RESSION Donanemab Ph2 placebo arm1 Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

6 - M O N T H B L I N D E D A N A L Y S I S · P L A S M A PRECISE-AD INTERIM RESULTS: Plasma pTau217 declined steadily through Day 169 PRECISE-AD Mean % change from baseline over time -25 -20 -15 -10 -5 0 5 0 29 85 141 169 % change from baseline Study day Linear trend PRECISE-AD patients with a Day-169 decline 68.5% 31.5% 0 20 40 60 80 Declined (≤0) Increased (>0) (Improved) Why this is compelling at 6 months Untreated pTau217 rises as the disease progresses. The 15% decline shown in the PRECISE-AD data (including both PMN and placebo treated patients) indicates a possible early treatment-associated reversal. Although responder data remains blinded, the ~68% responder fraction closely aligns with the 75% active allocation in the 3:1 design based on a prior ProMIS analysis. A 6-month pTau217 change could be an early leading indicator of benefit and a predictor of 12-month clinical outcome (Pentara/ProMIS). Randomization 3:1 Aggregated drug + placebo (3:1) 24 (Worsened) Ref: “Leveraging recent advances in plasma biomarkers to optimize early proof of concept trials in Alzheimer’s disease.” Alzheimer’s & Dementia: Translational Research & Clinical Interventions (TRCI), 2025. DOI 10.1002/trc2.70183. -15% Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

6 - M O N T H B L I N D E D A N A L Y S I S · C S F PRECISE-AD INTERIM RESULTS: CSF MTBR-tau243 declined through Day 169 PRECISE-AD Mean % change, screen → Day 169 -25 -20 -15 -10 -5 0 5 Screen Day 169 % change from baseline Timepoint PRECISE-AD: Patients with a Day-169 decline 62.5% 37.5% 0 20 40 60 80 Declined Increased Why this matters at 6 months MTBR-tau243 is correlated with tau tangle burden and normally rises as aggregates accumulate, so a downward move is a favorable, disease-relevant signal. Across the trial participants, it declined on average (13.3%) and in the majority of patients. The effect is early, not yet powered for significance, but its direction, against a rising natural history, is the meaningful result and is directionally consistent with the upstream plasma pTau217 decline and tracks the 3:1 randomization pattern. Aggregated drug + placebo (3:1) 25 (Improved) (Worsened) -13.3% Randomization 3:1 Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

C O N C L U S I O N A Differentiated, Precision Approach to Alzheimer’s — With a Clear Path Forward PRECISE-AD’s blinded interim analysis shows favorable safety data and early, mechanism-consistent biomarker signals, supporting a differentiated approach for patients underserved by approved anti-amyloid therapies. 01 DIFFERENTIATED MECHANISM Oligomer-selective by design. PMN310 is engineered to bind toxic amyloid-β oligomers while avoiding plaque, decoupling efficacy from the ARIA liability that defines the approved class. 02 FAVORABLE SAFETY PROFILE Zero ARIA-E; 4.4% total ARIA — all mild, asymptomatic ARIA-H. Zero treatment-related SAEs and zero drug-related discontinuations, minimal (n=1) drug-related infusion reactions as of data cutoff date. 03 TARGET ENGAGEMENT WITH RELEVANT BIOMARKERS Coherent, biologically aligned signal. Upstream plasma pTau217 declined significantly (~69% of patients, tracking 3:1 randomization) and downstream CSF MTBR-tau243 moved favorably (~63% of patients declined), both opposite a rising natural history. 04 OPENING ACCESS TO APOE4 CARRIERS No ARIA-E in any genotype. With 61% APOE4 carriers and 11% homozygotes enrolled, a favorable safety profile could remove the class’s single biggest barrier, especially for the patients at highest risk. 05 POTENTIAL IN PRECLINICAL AD A safety profile suited to earlier intervention. A placebo-level ARIA profile makes PMN310 a strong candidate to move upstream into preclinical AD, where prevention has the greatest impact, but tolerability is paramount. Copyright 2026, ProMIS Neurosciences Inc., Non-confidential 26

27 Topline in Q1 2027 and the Path Forward

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential T O P L I N E · Q 1 2 0 2 7 Q1 2027 topline data aims to show a material shift in benefit-risk 28 Target scenario reflect management’s view on possible safety and directional efficacy signals and is not based on any review of clinical data beyond the interim data in July 2026. PRECISE-AD is a 12-month Phase 1b study designed to assess safety and directional efficacy signals; it is not powered for statistical significance on clinical endpoints. If successful, PMN310 could potentially offer a favorable and differentiated benefit-risk profile ▼ LOWER RISK TARGET SCENARIO • ARIA similar to placebo (background) rate • Minimal infusion reactions • ~15–20% slowing of progression vs placebo • Similar efficacy to marketed drugs (derived from 18-month data) using iADRS or CDR-SB • Coherent directional trends in biomarkers, supportive of oligomer-selective MOA = COMPARABLE BENEFIT Materially improved safety Similar efficacy

Copyright 2026, ProMIS Neurosciences Inc., Non-confidential Q 1 Q 2 Q 3 Q 4 PRECISE-AD Phase 1b PMN310 in early AD — execute + complete Phase 1b Execution Subcutaneous PMN310 Accelerated SC formulation development SC formulation + PK/PD Registrational study prep PMN310 pivotal trial readiness Site activation + start-up FDA interactions EOP1b/2 Type B meeting + regulatory strategy Briefing package + Type B meeting prep Pipeline advancement PMN267 (ALS) + PMN442 (PD / DLB) PMN267 IND-enabling tox + CMC PMN442 IND-enabling studies Subcutaneous formulation development Registrational study prep Type B meeting prep PMN442 Dev (PD/DLB) PMN267 Dev (ALS) 29 Execution of 2H'26 activities is subject to interim data outcomes, regulatory feedback, and resource allocation S T R A T E G I C P R I O R I T I E S 2026 Strategic Priorities Parallel workstreams converging on PMN310 Phase 3 readiness and broader pipeline advancement 2026

W H Y P R O M I S Early PRECISE-AD signals support PMN310’s differentiated profile S U M M A R Y A potentially favorable and differentiated benefit-risk profile 12-month PRECISE-AD topline in Q1 2027, with cash runway through 2027 30 01 Novel precision targeting EpiSelect discovery platform, with a pipeline of antibodies and vaccines against toxic misfolded proteins 02 Clean safety profile at 6 months No ARIA-E, 4.4% ARIA-H, no treatment-related SAEs Minimal infusion reactions 03 Target engagement demonstrated Toxic oligomers reduced in Phase 1a pTau217 and MTBR-tau243 declining in Phase 1b 04 Topline results ahead 12-month PRECISE-AD data in Q1 2027 Cash runway through 2027 05 FDA Fast Track Designation Granted for PMN310 Allows more frequent FDA interaction Copyright 2026, ProMIS Neurosciences Inc., Non-confidential

NASDAQ: PMN For further information, contact: info@promisneurosciences.com