Slides
Page 1
SENS-601 Program Day Gene therapy for GJB2-related congenital hearing loss From the biology to the clinic 22 SEPTEMBER 2026 Euronext Growth: ALSEN Copyright by Sensorion 2026. All Rights Reserved
Page 2
I n t r o d u c t i o n SENS-601 enters the clinic Fred Chereau, Chief Executive Officer, Sensorion 2 Copyright by Sensorion 2026. All Rights Reserved
Page 3
Welcome to the SENS-601 Program Day Copyright by Sensorion 2026. All Rights Reserved 3 Listen-only mode All participants are in listen- only mode for the presentations. Q&A Dial in on the conference number. Lines open for questions at the end. 90 minutes The presentation runs ~ 90 minutes, and will be followed by a Q&A. Slides and replay Slides and a replay will be available at www.sensorion.com.
Page 4
Forward-looking statements This presentation contains certain forward-looking statements concerning the “Company” and its business, including statements regarding the Company’s clinical development plans, its regulatory submissions and the expected timing of clinical data. Such forward- looking statements are based on assumptions that Sensorion considers to be reasonable. However, there can be no assurance that such forward-looking statements will be verified, which statements are subject to numerous risks, including the risks set forth in the 2025 full year report published on March 18, 2026, and available on the Company’s website, and to the development of economic conditions, financial markets and the markets in which Sensorion operates. The forward-looking statements contained in this presentation are also subject to risks not yet known to Sensorion or not currently considered material by Sensorion. The occurrence of all or part of such risks could cause actual results, financial condition, performance or achievements of Sensorion to be materially different from such forward-looking statements. Copyright by Sensorion 2026. All Rights Reserved 4 Investigational product SENS-601 is an investigational product candidate. It has not been approved by any regulatory authority, and its safety and efficacy have not been established. Data presented today are derived from preclinical studies and from natural history research. Preclinical results are not necessarily predictive of clinical outcomes. No offer This presentation and the information it contains do not constitute an offer to sell or subscribe for, or a solicitation of an offer to purchase or subscribe for, Sensorion shares in any country. The communication of this presentation in certain countries may constitute a violation of local laws and regulations. Any recipient of this presentation must inform themselves of any such local restrictions and comply therewith. Unpublished data Certain natural history data presented today are unpublished. They are presented for scientific discussion and are subject to further analysis.
Page 5
Today’s agenda Where we stand • French ANSM authorized HearConnex on 31 August 2026 • Health Canada clinical trial application under review • First patient dosing targeted by early 2027 Why this event, and why now • Turning point for the program as SENS-601 enter the clinic • Full review of GJB2-GT: the disease, the standard of care, the biology, the preclinical package and the trial design What today covers • Five sessions, from the patient journey to the design of the first-in-human study • Panel discussion with all speakers to close Session 1 The patient journey Session 2 The science Session 3 The preclinical package Session 4 The clinical strategy Session 5 Panel and Q&A Copyright by Sensorion 2026. All Rights Reserved 5GJB2-GT: GJB2 Gene Therapy; targeted genetic treatment designed to restore hearing loss caused by mutations in the GJB2 gene (en coding Connexin 26).
Page 6
Copyright by Sensorion 2026. All Rights Reserved 6 Fred Chereau Chief Executive Officer, Sensorion Laurent Désiré Head of Preclinical, Sensorion Valérie Salentey Head of Regulatory Affairs and Quality Assurance, Sensorion Dr. Sharon Cushing The Hospital for Sick Children (SickKids), Toronto Prof. Christine Petit Institut Pasteur, Institut de l’Audition / Institut reConnect Pediatric otolaryngologist, Director of the Cochlear Implant Program. Coordinating Investigator, HearConnex. Professor Emeritus at the Collège de France, laureate of several prizes incl. the Kavli Prize in Neuroscience. Today’s speakers
Page 7
SENS-601 is built on work that started long before the construct did 1 Foundation • Inner-ear expertise across gene therapy and small molecules: biology, cochlear delivery, translational models and audiology endpoints • Know-how that transfers across inner-ear modalities • Strategic collaboration with Institut Pasteur on the genetics of hearing 2 Technology • AAV vector design for cochlear targeting • miR-based silencing to keep transgene expression out of hair cells • Injection system built for intracochlear administration • In-house process development and analytics 3 Clinical execution • First-in-human experience of the surgical procedure in the pediatric population, through the Audiogene trial • OTOCONEX, a natural history study feeding patient identification and recruitment 4 Where that leads • HearConnex approved in France, Canadian CTA review ongoing. US IND and Australian CTA filings expected by year-end 2026 • First patient dosing by early 2027 • Clinical data to be generated throughout 2027 7 Copyright by Sensorion 2026. All Rights Reserved
Page 8
SENS-601 is entering the clinic in congenital GJB2-related HL, with plans to extend across the broader population Sensorion aims to build a leading global hearing loss franchise by expanding across indications Congenital onset In the clinic Preclinical Phase I/II Phase III CTA approved in France CTA under review in Canada U.S. IND and Australia CTA submission expected by year-end 2026 Pediatric progressive Extension Preclinical Phase I/II Phase III Preclinical studies and population identification ongoing Adult onset (presbycusis) Extension Preclinical Phase I/II Phase III Preclinical studies and population identification ongoing Throughout this presentation, GJB2-related HL refers to DFNB1A: autosomal recessive, non-syndromic hearing loss caused by biallelic GJB2 variants. EU4: the four largest European Union markets: France, Germany, Italy and Spain. UK: United Kingdom, US: United-states. 8 ~2,000 newborns per year across the US, EU4 and UK Copyright by Sensorion 2026. All Rights Reserved
Page 9
S e s s i o n 1 Hearing loss and DFNB1A: the patient journey and limitations of standard of care Sharon Cushing, M.D., The Hospital for Sick Children (SickKids),Toronto Coordinating Investigator, HearConnex 9 Copyright by Sensorion 2026. All Rights Reserved
Page 10
E s t i m a t e d i n c i d e n c e ≈ 2,000 newborns each year across the USA, EU4 and UK 32 / 100,000 incidence in newborns in the US, EU4 and UK 10 Copyright by Sensorion 2026. All Rights Reserved
Page 11
≥ 71 dB Severe to profound deafness at birth in most patients Diagnosis Newborn hearing screening followed by GJB2 genotyping Before 3 years Management matters during cerebral plasticity, today with a cochlear implant Consequences Impact on language acquisition and development Care pathway Complex, involving screening, audiology, imaging, genetics and surgery Population 32 per 100,000 newborns in the US, EU4 and UK: approximately 2,000 patients per year 11 G J B 2 d e f i c i e n c y, c o n g e n i t a l f o r m Severe to profound deafness at birth, with a narrow window for intervention Copyright by Sensorion 2026. All Rights Reserved
Page 12
How children are identified today 12 Copyright by Sensorion 2026. All Rights Reserved
Page 13
Permanent hearing loss screening in Ontario runs on two parallel tracks Track 1 Physiologic hearing screening Track 2 Risk factor screening Congenital CMV Genetic Source: Newborn Screening Ontario 13 Copyright by Sensorion 2026. All Rights Reserved
Page 14
Genetic screening is the track that identifies DFNB1A Track 2 Risk factor screening Genetic Source: Newborn Screening Ontario 14 Copyright by Sensorion 2026. All Rights Reserved
Page 15
The genetic risk factor panel screens three genes Twenty variants identified in Ontario in total, half of them in GJB2. GJB2 10 variants screened c.35delG c.235delC c.167delT c.71G>A c.310_323del c.139G>T c.-23+1G>A c.231G>A c.427C>T c.269T>C GJB6 1 variant screened D13S1830 (342 kb deletion) SLC26A4 9 variants screened c.707T>C c.1001+1G>A c.1246A>C c.919-2A>G c.2168A>G c.1003T>C c.1229C>T c.1614+1G>A c.1541A>G Hearing loss risk factor screening, genetic panel 15 Copyright by Sensorion 2026. All Rights Reserved
Page 16
422,943 babies born in Ontario 412,424 received genetic screens (98.8%) 93 positive genetic screens 72 sensorineural hearing loss confirmed 98.8% consent rate 79 referred 3 passed 11 missed Ontario Population ~15 M Area 1.076 M km² Kernohan K, Genetics in Medicine 2024 16 Copyright by Sensorion 2026. All Rights Reserved O N T AR I O , J U L Y 2 0 1 9 – M AR C H 2 0 2 2 Universal newborn genetic risk factor screening works at population scale
Page 17
Genetic screening identifies the full severity spectrum, not only profound hearing loss Hearing phenotypes identified through universal genetic screening (n=93) Kernohan K, Genetics in Medicine 2024 17 Copyright by Sensorion 2026. All Rights Reserved 50% 28% 12% 7% 6% 3% 10% 21% 16% 40% 6% 21% 8% 3% 43% 6% 22% Panel/p.(Val37Ile) (n=40) Panel/Panel (n=52) All (n=92) Profound Severe Moderately severe Moderate Mild Minimal None • Three quarter of screened infants fall outside profound loss; only 28% fall into the profound category typically eligible for cochlear implantation • Biallelic panel variants carry the severe end: 50% profound and 67% severe or worse, 35 of 52 infants • p.(Val37Ile) sits at the other end: no profound or severe loss, 43% hearing normally at screening and 40% mild
Page 18
Reaching implantation takes many steps, each adding delay Age at each category in the cochlear implant candidacy process (days), n=113 Timeline of implant candidacy process CI: Cochlear implant 18 Copyright by Sensorion 2026. All Rights Reserved • Eighteen distinct steps separate newborn screening from implantation • Each dot is one child; 113 children followed through the full pathway • The median age rises at every step, and the spread widens as the pathway lengthens
Page 19
The cochlear implant journey, and where it stops Two implants, two sound processors, and a lifetime of device management. Copyright by Sensorion 2026. All Rights Reserved 19
Page 20
Cochlear implantation commits a child to lifelong device management Birth Newborn hearing screening Age 1 – 30 days Referral GP or pediatrician referral to ENT Age 1 – 3 months Diagnosis Hearing loss diagnosis, early intervention Work -up Hearing aid fitting, SLP / AVT evaluation, imaging, genetic diagnosis Age 4 – 6 months Surgery Cochlear implant surgery Age 7 – 14 months Activation Activation and switch-on 1, 3, 6 and 12 months post activation, annually thereafter Lifelong implications Ongoing CI programming Plus speech and language, and auditory verbal therapy appointments Monitoring and repairs Daily device management for the family Processor upgrade Every 5 years Explant and re-implant May occur once per ear during a lifetime Hearing loss diagnosis, cochlear implantation and CI activation may occur months or years later, depending on the underlying etiology, newborn screening methodology, completion of diagnostic follow-up and local care pathways. Genetic diagnosis may occur at any point in the journey, or not at all. 20 Copyright by Sensorion 2026. All Rights Reserved Note: time periods shown reflect age at each step of implantation process; does not reflect incremental time between stages
Page 21
An implant delivers a low-resolution version of what the cochlea does Cochlear implantation is the current standard of care The signal provided by cochlear implants can be compared to a pixelated, low-resolution image. A cochlea with restored transmission of the signal can be compared to a well-focused and clear image. The lateral spread of electrical current from each electrode contact stimulates too many auditory neurons at the same time, rather than the physiological one-to-one connectivity between hair cell and auditory neuron Copyright by Sensorion 2026. All Rights Reserved 21
Page 22
Listening with an implant stays effortful, even when it works Reaction time is longer for implant users than for normal-hearing peers when discriminating emotion in music, lateralizing bilateral cues and localizing sound. Hopyan et al., Child Neurophys 2014; Fung et al., Clinical Neurophys 2025; Alemu et al., Trends in Hearing 2025 22 Copyright by Sensorion 2026. All Rights Reserved
Page 23
I n t h e i r o w n w o r d s What families say about living with an implant A short excerpt from the Hear Here podcast. Scan to listen Copyright by Sensorion 2026. All Rights Reserved 23
Page 24
I n t h e i r o w n w o r d s Hearing loss costs energy, every day Two adults describe listening fatigue in the Hear Here podcast. Copyright by Sensorion 2026. All Rights Reserved “I consider myself to be a very extroverted person, but if I’m at a restaurant with my family or my friends, after the first hour, I’m absolutely exhausted.” Sofia “The hearing fatigue is always kind of in the background. It’s hard to identify if the fatigue I’m experiencing is just from a normal activity like exercise, or if it is something additional that’s being added because of my hearing loss.” May 24
Page 25
A one-time administration changes what the rest of childhood looks like Birth Newborn hearing screening Age 1 – 30 days Referral GP or paediatrician referral to ENT Age 1 – 3 months Diagnosis Hearing loss diagnosis, genetic diagnosis, early intervention Work -up SLP / AVT evaluation, imaging 4 – 6 months One -time Gene therapy surgery One-time administration Post-operative speech therapy, assessment and monitoring at 1, 3 and 6 months, semi- annually until age 6, annually thereafter Potential long -term benefits Improvement in hearing without the need of a device Improvement in social skill development Ease of communication in social situations Increased social engagement with peers Improvement in ability to understand in noise Reduction of listening fatigue and listening effort Development of independence and autonomy Participation in sports and activities without fear of damage to the sound processor Eliminate Daily maintenance of a device Possible surgery to replace an implant Device stigma Use of assistive devices or technology Costs to replace devices or accessories Specific educational needs Vona B 2020; Wong CL 2017; Antoni M 2016; Awad R 2019. Note: time periods shown reflect age at each step of gene therapy treatment process; does not reflect incremental time between stages Copyright by Sensorion 2026. All Rights Reserved 25
Page 26
Two ears are not the same thing as hearing with two ears Bilateral hearing Sound is delivered to both ears. Each side is stimulated independently. Binaural hearing The two ears are integrated centrally. Gene therapy is what allows localization, lateralization and listening in noise. Part 2 of HearConnex: plan for bilateral administration Copyright by Sensorion 2026. All Rights Reserved 26
Page 27
S e s s i o n 1 , k e y m e s s a g e s 27 1 GJB2-related hearing loss is identified at birth, and the window for intervention is measured in months 2 Cochlear implantation restores access to sound, not physiological hearing 3 Binaural integration, listening in noise and listening effort are where the significant unmet need remains The population is identified at birth and the gap left by implants is well-defined: gene therapy addresses the cause of hearing loss instead of bypassing it Copyright by Sensorion 2026. All Rights Reserved
Page 28
S e s s i o n 2 GJB2 mouse models: decoding DFNB1A to drive therapeutic innovation Prof. Christine Petit, Institut Pasteur, Institut de l’Audition / Institut reConnect Professor Emeritus at the Collège de France Laureate of the Kavli Prize in Neuroscience Copyright by Sensorion 2026. All Rights Reserved 28
Page 29
“Les Causeuses” (The Conversationalists) Camille Claudel, 1897 Immanuel Kant "not seeing separates us from things; not hearing, from our fellow man." Copyright by Sensorion 2026. All Rights Reserved
Page 30
1992 Discovery of the genes responsible for deafness A strategy based on the study of affected blood-related families living in geographic isolates. This approach, applied across isolated populations mainly around the Mediterranean sea, became the foundation for uncovering the genes responsible for deafness over the following three decades. 30 Copyright by Sensorion 2026. All Rights Reserved
Page 31
Causal genes for monogenic early-onset forms of non-syndromic deafness 31 Weil D et al. 1995 Nature, Gibson et al, 1995 Nature The number of known deafness genes has climbed steadily since the first DFNB1 locus was mapped in 1994, accelerating after the 2001 first published human genome sequence and now reaching more than 150 genes. 201994: the first locus for human isolated deafness (DFNB1) # deafness genes 202001: first draft of the human genome sequence 150 100 50 0 2020201520102005200019951990 The first human deafness gene identified as responsible for isolated deafness Copyright by Sensorion 2026. All Rights Reserved
Page 32
Genetic architecture of hearing loss 157 Genes for monogenic congenital / prelingual / childhood / pre-adult / age-related hearing loss altogether 88 Genes for autosomal recessive (DFNB) forms — generally the most severe and with the earliest onset (congenital prelingual) 65 Genes for autosomal dominant (DFNA) forms, generally postlingual progressive and less severe Genes cause for both DFNA and DFNB forms 2 Modifier genes 9 Mitochondrial genes, incl. 2 genes predisposing to aminoglycoside ototoxicity ~300 Genes for monogenic syndromic forms of hearing loss ~100 / ~135 Genes associated with age-related hearing impairment / noise- induced hearing loss susceptibility 32 15 7 Chromosome X-linked genes 1 Chromosome Y-linked genes Copyright by Sensorion 2026. All Rights Reserved
Page 33
A single recessive, non syndromic form, DFNB1 caused by defect in GJB2 accounts for up to half congenital deafness cases in some geographic areas 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 2% 2% 3% 3% 4% 4% 4% 6% 6% 6% 48% EPS8 COL9A1 PCDH15 CABP2 USH1C TRIOBP TMPRSS3 STRC SOX10 PTPRQ PNPT1 OTOA ILDR PAX3 MITF SLC26A4 LHFPL5 CDH23 TMC1 MYO7A OTOF MYO15A LRTOMT GJB2 GJB2 variants of children affected by DFNB1 born with profound deafness retain residual hearing at birth; most of them will lose it in early childhood. Algeria 2/3 50% of congenital deafness in some regions, especially the Maghreb and Middle East (except Ghana and sub-Saharan Africa) Clinical course Copyright by Sensorion 2026. All Rights Reserved 33
Page 34
Connexin 26 (CX26) from connexons to gap junctions 34 conserved variable E1 E2 M1 M2 M3 M4 Topology of a generic connexin heteromeric heterotypic homotypic Intercellular channelsconnexins connexon heteromeric homomeric Gap junction cytoplasm of cell 1 cytoplasm of cell 2 intercellular gap Copyright by Sensorion 2026. All Rights Reserved
Page 35
35 Copyright by Sensorion 2026. All Rights Reserved CX26 gap junction networks in the cochlea
Page 36
“Evidence for a medial K+ recycling pathway from inner hair cells.” Spicer SS, Schulte BA. Hear Res. 1998 Apr;118(1–2):1–12. Copyright by Sensorion 2026. All Rights Reserved 36 Connective tissue CX26 gap junction network in the cochlea
Page 37
Physiology of CX26 gap junction networks in the cochlea 1. K+ buffering and recycling Mechanoelectrical transduction leads to extrusion of K+ ions from hair cells which enter supporting cells and move from cell to cell via CX26/(CX30) channels. K+ ions also enter the connective gap junction network generating the endocochlear potential and resulting in K+ secretion into the endolymph. 2. Development Cx26/(Cx30) channels are essential for cochlea development, notably for the tunnel of Corti formation. They are also essential for propagation of IP3 / Ca2+ waves and glucose transport during embryonic development. 37 Tunnel de Corti 3. Protection of cochlear cells in adulthood K+: potassium ion; CX26/CX30: connexins 26 and 30; IP3: inositol trisphosphate; Ca2+: calcium ion Copyright by Sensorion 2026. All Rights Reserved
Page 38
Pathophysiology of CX26 gap junction networks in the cochlea Pathophysiological studies in animal models reveal different defects in different models. Absence of tunnel of Corti opening Absence of, or decrease in, endocochlear potential Functional defects of cochlear cell types, including hair cells Elevation of the hearing threshold: mild to profound deafness 38 Copyright by Sensorion 2026. All Rights Reserved
Page 39
39 Osteocytes P8 Osteocytes P12 Osteocytes P20 chondrocytes Tympanic border cells New cell type Reissner’s membrane Endothelial cells Basal stria Fibrocytes Marginal stria Spindle cells Inner border/ Hensen’s cells Claudius/ Inner sulcus/ Inter dental cells Deiter’s cells Root cells Neurons Glial cells Hair cells Jean P, Petit C*, Michalski N*. PNAS 2023. Copyright by Sensorion 2026. All Rights Reserved When and where is CX26 essential for cochlear cell survival and proper auditory function? I n s i g h t s f r o m t a i l o r e d m o u s e m o d e l s
Page 40
The first transcriptomic atlas the mouse cochlea from postnatal to adult stages maps deafness gene expression across every cochlear cell type: hair cells, neurons, fibrocytes and stria vascularis cells… This atlas identifies precisely which cell types express Gjb2 and suggests the genes it may directly or indirectly interact with, providing the reference map currently used to design our targeted mouse models. Jean P, Petit C*, Michalski N*. PNAS 2023. 40 Copyright by Sensorion 2026. All Rights Reserved K+: potassium ion; CX26/CX30: connexins 26 and 30; IP3: inositol trisphosphate, an intracellular signaling molecule. Ca2+: calcium ion When and where is CX26 essential for cochlear cell survival and proper auditory function? I n s i g h t s f r o m t a i l o r e d m o u s e m o d e l s
Page 41
Three tailored Gjb2-deficient mouse models, built on the Cre/Lox system Connective tissue network — a) fibrocytes Cre driven by a promoter expressed exclusively in cochlear fibrocytes. Connective tissue network — b) stria vascularis intermediate cells Cre driven by a promoter expressed exclusively in the intermediate cells of the stria vascularis. Epithelial gap junction network - cKO1 model used to build the preclinical package 41 Gjb2-deficient mouse models Cre/Lox SYSTEM DRIVER Cre line FLOXED Gjb2 line Cre driven by a promoter expressed exclusively in the neuroepithelium Copyright by Sensorion 2026. All Rights Reserved cKO1:conditional knockout model 1
Page 42
Loss of CX26 expression in the spiral ligament and spiral limbus fibrocytes. While the expression in the epithelial gap junction network and the stria vascularis is persisting. No morphological defects observed (P30 cochlea cryosection). Adapted from doi.org/10.3389/fneur.2020.580639 42 Spiral limbus Spiral limbus Spiral ligament Spiral ligament CX26+ Fibrocyte type I CX26 +Fibrocyte type II CX26+ Fibrocyte type V Fibrocyte type IV Fibrocyte Type III CX26 Tubulin Fibrocytes of the spiral limbus (SL) and the spiral ligament (SLi) Control Mutant P30 cochlea cryosection; CX26 in white SV: stria vascularis FiIRESCre-/- Gjb2flox/flox FiIRESCre+/- Gjb2flox/flox SLi SL SV SLi SV SL Control Mutant Copyright by Sensorion 2026. All Rights Reserved C o n n e c t i ve t i s s u e n e t w o r k : A- t h e f i b r o c yt e s Loss of CX26 in the spiral ligament and spiral limbus, with no morphological defects
Page 43
Loss of CX26 expression in the intermediate cells of the stria vascularis (SV). Persisting expression in the epithelial gap junction network and in the fibrocytes of the spiral ligament and spiral limbus. 43 The basal and intermediate cells of the stria vascularis (SV) CX26 Phalloidin SVbSVin SVbSVin CX26 Phalloidin Control Mutant Mutant Control SVin: stria vascularis intermediate SVb: stria vascularis basal Copyright by Sensorion 2026. All Rights Reserved C o n n e c t i ve t i s s u e n e t w o r k : B- t h e s t r i a va s c u l a r i s Loss of CX26 in the intermediate cells of the stria vascularis
Page 44
Three phenotypes (1 to 3) are classified according to their ABRs. From P15 and at 1 month: ▪ phenotype 1: mice are profoundly deaf ▪ phenotype 2: mice are severely deaf ▪ phenotype 3: mice have normal hearing and become progressively deaf 44 CX26 Phalloidin Tubulin MutantControl 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) Phenotype 1 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) Phenotype 2 Phenotype 3 Control (normal hearing) 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 1 month 2 months 6 months3 months ABR: Auditory Brainstem Response Copyright by Sensorion 2026. All Rights Reserved E p i t h e l i a l g a p j u n c t i o n n e t w o r k One model - three phenotypes based on the ABR profile
Page 45
45 Hearing mice Hearing impaired mice Control Hearing mice Phenotype 1 Phenotype 2 Phenotype 3 CX26↓↓↓, TC collapsed CX26↓↓, TC open CX26↓, TC open Cryosections at P16TC: Tunnel of Corti Copyright by Sensorion 2026. All Rights Reserved E p i t h e l i a l g a p j u n c t i o n n e t w o r k Three phenotypes - different levels of CX26 expression
Page 46
46 Phenotype 1 Phenotype 2 Phenotype 3 cKO1 Phenotype 3 Hearing cKO1 Phenotype 2 Hearing loss cKO1 Phenotype 1 Profound hearing loss Control Hearing No Inner Hair Cell loss in any of the 3 phenotypes up to 6 months base apex Loss of Outer Hair Cell limited to phenotype 1 in the basal and middle-base of the cochlea at 6 months Inner and outer hair cell persistence allows for therapeutic intervention Copyright by Sensorion 2026. All Rights Reserved E p i t h e l i a l g a p j u n c t i o n n e t w o r k Three phenotypes – hair cells viability
Page 47
Outer hair cells (OHC) Function analyzed by DPOAEs from 1 to 6 months. In Phenotypes 1 & 2, absence of DPOAE associated with OHC persistence, indicating OHC dysfunction. In Phenotype 3, Outer Hair Cells develop functional defects between 3 and 5 months. 47 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 1 month 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 3 months 6 months 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) Phenotypes 1 & 2 Phenotype 3 Control 5 10 15 20 32 40 0 30 60 90 120 1 month Frequencies (kHz) ABR thresholds (dB SPL) Otogl IRESCre/IRESCre Gjb2 loxP/ loxP HI mice CTRL Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP hearing mice Otogl IRESCre/+ Gjb2 loxP/ loxP HI mice 5 10 15 20 32 40 2 months Frequencies (kHz) 5 10 15 20 32 40 3 months Frequencies (kHz) 5 10 15 20 32 40 4 months Frequencies (kHz) 5 10 15 20 32 40 5 months Frequencies (kHz) 6 8 12 16 24 32 0 20 40 60 80 Frequencies (kHz) DPOAE thresholds (dB SPL) 5 10 15 20 32 40 6 months Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) 6 8 12 16 24 32 Frequencies (kHz) Inner hair cells (IHC) Function analyzed by suprathreshold ABR response (wave 1 amplitude), demonstrating no functional defects in any of the three models. Together, this GJB2 model with three phenotypes provides a unique experimental platform to assess the efficacy of gene therapy across distinct cochlear cell types within the epithelial gap junction network Copyright by Sensorion 2026. All Rights Reserved DPOAEs: distortion product otoacoustic emissions E p i t h e l i a l g a p j u n c t i o n n e t w o r k Three phenotypes – hair cells functionality
Page 48
S e s s i o n 2 , k e y m e s s a g e s 48 1 2 3 Unique, highly relevant, exclusive mouse models developed Our bi-allelic GJB2 deletion model with congenital severe to profound HL, mimics the most common form of DFNB1A By decoding DFNB1A, we aim to drive therapeutic innovation and pave the way for gene therapy development for the benefit of patients Unraveling the pathogenic mechanisms underlying DFNB1A Identification of the cochlear cells in which CX26 expression is essential for normal hearing A new era of innovative therapeutic solutions for patients enabled the discovery of the genes responsible for hearing loss Copyright by Sensorion 2026. All Rights Reserved
Page 49
AC K N O W L E D G M E N T S With thanks to the team behind this work Amrit Estivalet Anne-Valérie Héritier Solène Roux Andrea Lelli Nawel Mekdad Jeanne Rakotopare Lauralee Robichon Muriel Sudres Christine Petit Nicolas Michalski 49 Copyright by Sensorion 2026. All Rights Reserved
Page 50
S e s s i o n 3 SENS-601: the preclinical package Laurent Désiré, Head of Preclinical, Sensorion Valérie Salentey, Head of Regulatory Affairs and Quality Assurance, Sensorion 50 Copyright by Sensorion 2026. All Rights Reserved
Page 51
The Institut Pasteur collaboration gives SENS-601 a strong scientific foundation based on unique expertise in biology and genetics of hearing An exclusive collaboration Exclusive license option Co-development model on the GJB2 gene therapy program, developed with the laboratory of Prof. Christine Petit The two teams have successfully advanced the program from target identification into a clinical candidate What the collaboration provides Mechanistic understanding. Cell-level mapping of CX26 expression across the cochlea, used to define which cells the vector has to reach to be able to build the most relevant construct Disease-relevant models. Exclusive conditional knockout models, characterized for hearing phenotype and for the cochlear cell types involved Why it matters for SENS -601 Target cell selection, the silencing strategy and the choice of efficacy readouts all come out of that work. Partner institutions Prof. Christine Petit, Institut Pasteur, Institut de l’Audition / Institut reConnect 51 Copyright by Sensorion 2026. All Rights Reserved
Page 52
The construct and the design rationale Laurent Désiré, Head of Preclinical, Sensorion Copyright by Sensorion 2026. All Rights Reserved 52
Page 53
Objective. Enabling targeted expression of the CX26 protein to restore gap junctions in supporting cells. AAV-DJ capsid, providing broad tropism across cochlear supporting cell types Promoter smCBA, well characterized, ensuring strong and durable expression of the human GJB2 gene Regulatory sequence miR-183 target sequence, silencing CX26 expression in hair cells for high selectivity and safety Patents on SENS-601 and the miR-183 target sequence SENS-601 vector and cassette 53 Copyright by Sensorion 2026. All Rights Reserved P r e c l i n i c a l , s p e c i f i c t a r g e t i n g The SENS-601 vector is designed to restore Connexin 26 expression and the gap-junction network in cochlear supporting cells
Page 54
Efficient. Transgene silencing efficacy: 85-100 % Sustained. miR-183 sequence remains expressed in aged cochlea, including human Durable. Long-term miR-183-mediated transgene silencing in hair cells, at least 12 months after treatment miR-183-mediated transgene expression control is well tolerated We validated by in situ hybridization that miR183 is mostly detected in cochlear sensory hair cells and maintained for at least 1 year in mice Copyright by Sensorion 2026. All Rights Reserved 54 Selective. Off in sensory hair cells, on in target cells P r e c l i n i c a l , s p e c i f i c t a r g e t i n g A miR-183 target sequence keeps expression out of the sensory hair cells
Page 55
Full hair cell transgene silencing in mouse cochlea Negative control, AAV-GFP without miR-TS, AAV-GFP with miR-183 TS. Arrows indicate transduced inner hair cells. Full hair cell transgene silencing in NHP cochlea Same pattern reproduced in non-human primate cochlea. Hair cells. +++ without miR-183 TS, no expression with miR-183 TS Supporting cells. +++ in both conditions miR-183 allows efficient transgene expression in the target cells with virtually no expression in hair cells Copyright by Sensorion 2026. All Rights Reserved 55 AAV-DJ GFP (no miR183-TS) AAV-DJ GFP (+ miR183-TS)Non-injected ear AAV-DJ GFP (no miR183-TS) AAV-DJ GFP (+ miR183-TS) NHP: non-human-primate P r e c l i n i c a l , s p e c i f i c t a r g e t i n g SENS-601 combines broad cochlear reach with built-in hair cell transgene silencing via miRNA
Page 56
Yellow staining indicates the characteristic CX26 expression pattern at the cell membrane, overlapping with the transgene Flag No expression of either endogenous CX26 or transgenic CX26 Flag is detected in hair cells Copyright by Sensorion 2026. All Rights Reserved 56 No transgene expression in the sensory hair cells GJB2-Flag-GT: SENS-601 surrogate vector with CX26 fused to Flag epitope P r e c l i n i c a l , s p e c i f i c t a r g e t i n g Connexin 26 is observed at the cell membrane of the appropriate target cells in NHP cochlea
Page 57
Copyright by Sensorion 2026. All Rights Reserved 57 Protein expression remains highly expressed for at least 6 months in mice No loss of mRNA expression detected by RT-qPCR up to 6 months in mice and NHP cochlea Long term transgene mRNA and protein expression demonstrated in cochlea CX26 or Flag immunostaining in WT mice 6 months after injection CX26 mRNA quantification in WT mice 4 months after SENS-601 injection P r e c l i n i c a l , d u r a b l e t r a n s g e n e e x p r e s s i o n Long-term transgene expression in mice and NHP
Page 58
Efficacy and functional restoration Laurent Désiré, Head of Preclinical, Sensorion Copyright by Sensorion 2026. All Rights Reserved 58
Page 59
P r e c l i n i c a l , d e m o n s t r a t e d r e s t o r a t i o n SENS-601 restores functional hemichannels in Connexin-deficient HeLa cells Confirmatory dye transfer assays further demonstrate that GJB2 transduction leads to functional gap junctions and associated intercellular communication. 0 20 40 60 80 100 HPA % PI + cells no Ca2+ Ca2+ CBX Hela DH Hela DH GJB2 Hela DH GJB2-GT 35delG: truncating mutation, severe-to-profound HL M34T / V37I: missense variants, mild-to-moderate HL Propidium iodide assay monitors functional hemichannels SENS-601 restores functional hemichannels Rescue of functionality from CX26 human pathological mutants cell line transduced Copyright by Sensorion 2026. All Rights Reserved 59
Page 60
ABR threshold, 7 weeks post-injection Pooled data from multiple studies using several batches **** p<0.0001; two-way ANOVA multiple comparisons followed by Tukey test ABR threshold by dose, 7 weeks post-injection *** p<0.001; two-way ANOVA multiple comparisons followed by Tukey test SENS-601 1X dose (n=10) SENS-601 0.3X dose (n=19) SENS-601 0.1X dose (n=8) Control mice + SENS-601 1X dose (n=17) Vehicle (n=12) 8 16 24 32 0 20 40 60 80 100 Frequency (kHz) ABR Threshold (dB SPL) **** **** Control mice + GJB2-GT 6.3E9 VG/ear (n=42) Mutant mice + Vehicle (n=27) Responders Mutant mice + GJB2-GT 6.3E9 VG/ear (n=21 out of 35) Mutant mice + GJB2-GT 6.3E9 VG/ear (n=35) Mutant mice + SENS-601 1X dose (n=35) Control mice + SENS-601 1X dose (n=42) Responder mutant mice + SENS-601 1X dose (n=21/35) Mutant mice + vehicle (n=27) Copyright by Sensorion 2026. All Rights Reserved 8 16 24 32 0 20 40 60 80 100 Frequency (kHz) ABR Threshold (dB SPL) **** **** Control mice + GJB2-GT 6.3E9 VG/ear (n=42) Mutant mice + Vehicle (n=27) Responders Mutant mice + GJB2-GT 6.3E9 VG/ear (n=21 out of 35) Mutant mice + GJB2-GT 6.3E9 VG/ear (n=35) ▪ Efficacy within 3 weeks of injection at P0/P2, across all tested frequencies ▪ Reproduced across multiple batches, from R&D through to final manufacturing ▪ Restored ABR correlates with DPOAE and EP; effective dose confirmed in dose-ranging studies ABR: Auditory Brainstem Response 60 P r e c l i n i c a l , d e m o n s t r a t e d r e s t o r a t i o n SENS-601 restores ABR thresholds across multiple frequencies in the cKO1 model with severe to profound HL
Page 61
Tonotopic and dose-dependent improvement of CX26 levels, evaluated by AI-based image analysis of the cochlear CX26 network. Analyses performed 18 weeks post-administration at P0/P2. 61 CX26 levels quantification CX26 signal m ean intensity level (A.U.) Copyright by Sensorion 2026. All Rights Reserved Vehicle injected controls SENS-601 – 1X dose injected mutantsVehicle injected mutants SENS-601 – 0.3X dose injected mutants P r e c l i n i c a l , d e m o n s t r a t e d r e s t o r a t i o n Cochlear CX26 levels are normalized by SENS-601 in the cKO1 model – outer sulcus
Page 62
The C26 network itself is rescued, not only the level of expression CX26 network metrics are improved in multiple cochlear structures and cell type CX26 network tracks auditory recovery Analyses performed 18 weeks post-administration at P0/P2. 62 CX26 network metrics Association with ABR outcome Associations are significant: p<0.0001; loop density p=0.0135 Branch-point density r=0.4826 Branch density r=0.4718 Network length r=0.4683 CX26 intensity r=0.4505 Loop density r=0.2540 150 CX26 network features measured per sample Main CX26 network metrics Copyright by Sensorion 2026. All Rights Reserved P r e c l i n i c a l , d e m o n s t r a t e d r e s t o r a t i o n …
Page 63
Preclinical Observations & Tolerability Laurent Désiré, Head of Preclinical, Sensorion Copyright by Sensorion 2026. All Rights Reserved 63
Page 64
Surgical delivery approach Dual-fenestration technique combines two familiar procedures, cochlear implantation and stapedotomy, and is designed to minimize overpressure and no backflow. Designed to maximize vector distribution and target cell exposure along the full length of the cochlea, allowing widespread transduction. Well tolerated in our previous gene therapy program with no serious adverse events reported; reproducibility confirmed across multiple administrations performed to date. Proprietary injection system combining cochlear implantation and stapedotomy approaches Copyright by Sensorion 2026. All Rights Reserved 64 Adapted from Yoshimura et al (2018)
Page 65
Copyright by Sensorion 2026. All Rights Reserved 65 Injection of SENS-601 (Flag surrogate), at the NHP equivalent effective dose, leads to strong transgene detection in the cochlea of NHP from base to apex Flag signal within the CX26 network was quantified across the different compartments of interest and tonotopic axis Transgene expression pattern in NHP is consistent to the one observed in mice SENS-601 Efficiently Transduces NHP Cochlea CLAUDIN11 SOX2 GFP Actin SENS-601 injected with intended surgical delivery approach efficiently transduces NHP cochlea
Page 66
Mouse studies Design. Intravenous administration. Three groups: vehicle, low dose, high dose. Three cohorts: 3 months, 6 months and biodistribution. c Tolerability.No mortality. No SENS-601-related clinical or histology findings; one minor non-adverse creatinine change in females only. Biodistribution and shedding. Dose-proportional; still detectable at 6 months, mainly at the injection site (tail) and in the liver (highest level); DNA in blood up to Week 5. Copyright by Sensorion 2026. All Rights Reserved 66 Non-human primate studies Design. Unilateral intracochlear administration. Three groups: vehicle, low dose, high dose. Two cohorts: 3 months and 6 months. Tolerability.No mortality. Mild and transient anti-AAV-DJ response only in low-titer animals; no anti-CX26 antibodies. Weak cellular immune response against AAV-DJ, no cellular response to hGJB2. Biodistribution and shedding. Dose-related, confined to the injected ear and draining lymph nodes. No blood or CSF detection, rapid clearance in urine and feces (undetectable by Day 15); transient positivity in nasal swabs to Day 29 and ear swabs to Week 26. P r e c l i n i c a l , s a f e t y a n d t o l e r a b i l i t y No safety signal was observed in SENS-601 combined 3 and 6-month GLP toxicology and biodistribution studies
Page 67
S e s s i o n 3 , k e y m e s s a g e s 67 1 The target cells are identified, and the construct is designed to reach them Transgene is expressed in target cells but not in hair cells 2 SENS-601 studies showed functional restoration of hearing and normalization of the CX26 network 3 No safety signal was observed in GLP toxicology and biodistribution studies Surgery was well tolerated in previous gene therapy program With data and models translating well to humans, we are ready to take SENS-601 into the clinic Copyright by Sensorion 2026. All Rights Reserved
Page 68
Regulatory path Valérie Salentey, Head of Regulatory Affairs and Quality Assurance, Sensorion Copyright by Sensorion 2026. All Rights Reserved 68
Page 69
First-in-human study with SENS-601 is designed as a multi-country trial across four geographies Enrollment opens geography by geography as each regulatory clearance lands Site activation follows regulatory clearance and site contracting. 69 France CTA approved Hôpital Necker Enfants Malades, Paris Dr. Natalie Loundon Canada Application under review Hospital for Sick Children, Toronto Dr. Sharon Cushing Australia CTA Filing expected by YE 2026 Children’s Hospital Westmead, Sydney Pr. Catherine Birman United States IND Filing expected by YE 2026 Site to be disclosed Copyright by Sensorion 2026. All Rights Reserved
Page 70
S e s s i o n 4 Clinical development strategy and the HearConnex design Sharon Cushing, M.D., The Hospital for Sick Children (SickKids), Toronto Coordinating Investigator, HearConnex Copyright by Sensorion 2026. All Rights Reserved 70
Page 71
OTOCONEX: the natural history foundation Copyright by Sensorion 2026. All Rights Reserved 71
Page 72
Longitudinal clinical, audiological and genetic data support endpoint selection and stratification Data collected Genetic GJB2 genotype, molecular diagnosis Audiological Pure tone audiometry and objective auditory assessments over time (at least 2 years) Functional Functional hearing assessments and quality of life questionnaires Population Aged 16 years or under without a cochlear implant, and 10 years or under with an implant. Non-syndromic, bilateral, mild to profound sensorineural hearing loss, pre-lingual or post-lingual. What it enables Endpoint selection, patient stratification and trial design decisions, taken on data rather than on assumption. Copyright by Sensorion 2026. All Rights Reserved 72
Page 73
OTOCONEX also built the patient, site and data infrastructure HearConnex runs on 1 Patient asset Genotyped patients Clinical characterization Longitudinal follow-up 2 Site asset Experienced investigators Referral pathways Standardized assessments 3 Data asset Natural history data Outcome measures Disease progression insights Patient identification, and reduced recruitment uncertainty A trained network of investigators and sites A natural history benchmark for treatment evaluation Copyright by Sensorion 2026. All Rights Reserved 73
Page 74
HearConnex: study design Copyright by Sensorion 2026. All Rights Reserved Grant from French State (Bpifrance) as part of the France 2030 investment plan 74
Page 75
The standard of care improves access to sound; the aim here is sustained physiological hearing The disease GJB2-related hearing loss is the most frequent type of autosomal recessive non-syndromic hearing loss. The limit of current care Cochlear implantation improves access to sound and speech stimuli but does not restore physiological hearing, and some limitations still need to be overcome. The intent of treatment SENS-601 aims to facilitate physiological hearing while overcoming the drawbacks of implantation and is intended to last a lifetime. Based on efficacy results from proof-of-concept studies in mouse models. Copyright by Sensorion 2026. All Rights Reserved 75
Page 76
Open-label, multicenter Phase I/II trial to assess the safety, tolerability and efficacy of intracochlear administration of SENS- 601 at escalating doses and the injection system. Trial objectives Primary Dose escalation. Safety and tolerability of SENS-601 and the injection system Dose expansion. Efficacy of SENS-601 on Auditory Brainstem Response (ABR) and Pure Tone Audiometry (PTA) Secondary Efficacy and safety of SENS-601; injection system performance, usability and clinical benefit Status Regulatory France: CTA approved 31 August 2026 Canada: CTA under review United States: IND submission expected by YE 2026 Australia: CTA submission expected by YE 2026 Enrollment First patient in anticipated by early 2027 Clinical data generation expected throughout 2027 CTA: Clinical Trial Application; CTN: Clinical Trial Notification; IND: Investigational New Drug. 76 Copyright by Sensorion 2026. All Rights Reserved HearConnex will evaluate SENS-601 safety and efficacy, with first patient dosing anticipated in early 2027
Page 77
HearConnex Phase I/II trial design Cochlear implant-naive children Children already unilaterally implanted 77 Part 1 – Dose Escalation Part 2 – Dose Expansion Study population Children with non-syndromic, bilateral, severe sensorineural hearing loss eligible for cochlear implantation Patients 8 children 10 children Age ≥ 6 months to ≤ 31 months, cochlear implant-naïve ≥ 6 months to ≤ 6 years, already unilaterally implanted within 18 months ≥ 6 months to ≤ 31 months, cochlear implant-naïve GT procedure Unilateral intracochlear injection Bilateral intracochlear injection Dosing Dose level 1 / Dose level 2 At the optimal dose Primary objective Safety and tolerability Efficacy (co-primary endpoints: ABR, PTA) Copyright by Sensorion 2026. All Rights Reserved ABR: Auditory Brainstem Response, PTA: Pure-tone audiometry
Page 78
HearConnex is designed to allow fast transition to pivotal trial PX: patient number X. DMC: Data Monitoring Committee. Part 1 uses unilateral administration, Part 2 bilateral administration. 78 Copyright by Sensorion 2026. All Rights Reserved HearConnex Phase I/II Open-Label Trial Design
Page 79
HearConnex is planned at centers already trained on the procedure Sensorion has long-standing relationships with an extensive global network of sites with expertise in hearing loss. Hospital for Sick Children, Toronto Dr. Sharon Cushing Hôpital Necker Enfants Malades, Paris Dr. Natalie Loundon Children’s Hospital Westmead, Sydney Pr. Catherine Birman Planned investigator centres, each trained on the procedure. Site activation follows regulatory clearance and site contracting. France: CTA approved by ANSM. Canada: application under review. Australia and United States: filing targeted by YE 2026 ; outlined marker indicates site selection under way. 79 United States Site to be disclosed Copyright by Sensorion 2026. All Rights Reserved Centers trained on procedures
Page 80
Positive benefit / risk assessment The case rests on four elements, each shown earlier today. Efficacy grounded in mechanism • Proof-of-concept established in mouse models • Addresses the pathophysiology of GJB2-related hearing loss directly • Restores CX26 expression and the Cx26 network in disease- relevant models • Recovery of ABR thresholds and improved endocochlear potential GLP Toxicology / safety package • Combined three- and six-month studies in mice and non-human primates • Absence of safety signal A surgical procedure already performed in this population • Intracochlear administration and injection system already used in the Audiogene trial of SENS-501 • No safety signal observed for the procedure itself The cochlear implant option is preserved • Children who do not benefit can still receive a cochlear implant • The schedule keeps that option open within the recommended developmental window Copyright by Sensorion 2026. All Rights Reserved 80
Page 81
S e s s i o n 4 , k e y m e s s a g e s 81 1 OTOCONEX defined the endpoints, the disease trajectory and the patient survey before the trial was designed 2 HearConnex leverages our experience from previous trials in the hearing field as well as our relationships with already trained centers 3 The Phase I/II trial allows flexibility to yield early data on dose selection and confirmatory data on bilateral administration to move to next steps HearConnex was informed by learnings from previous trial and natural history study to carry a path rapidly to give access to gene therapy to the largest number of patients Copyright by Sensorion 2026. All Rights Reserved
Page 82
S e s s i o n 5 Panel discussion and Q&A Copyright by Sensorion 2026. All Rights Reserved 82
Page 83
Panel discussion To ask a question, dial in on the conference call number ir.contact@sensorion-pharma.com Copyright by Sensorion 2026. All Rights Reserved 83 Fred Chereau Chief Executive Officer, Sensorion Laurent Désiré Head of Preclinical, Sensorion Valérie Salentey Head of Regulatory Affairs and Quality Assurance, Sensorion Dr. Sharon Cushing The Hospital for Sick Children (SickKids), Toronto Prof. Christine Petit Institut Pasteur, Institut de l’Audition / Institut reConnect Pediatric otolaryngologist, Director of the Cochlear Implant Program. Coordinating Investigator, HearConnex. Professor Emeritus at the Collège de France, laureate of several prizes incl. the Kavli Prize in Neuroscience.
Page 84
S e s s i o n 6 Closing remarks Fred Chereau, Chief Executive Officer, Sensorion Copyright by Sensorion 2026. All Rights Reserved 84
Page 85
Three things to take from today Copyright by Sensorion 2026. All Rights Reserved 85 1 The biology is mapped and the target cells are identified. The cochlear cell types critical to hearing in DFNB1A have been characterized in disease-relevant models, and SENS-601 is designed to reach them and to stay out of hair cells. 2 The preclinical package shows functional restoration. CX26 expression and network restoration, ABR threshold recovery across frequencies, GLP toxicology in mice and non-human primates were well tolerated. 3 The trial is designed on Sensorion’s own experience. Prior trials and natural history studies set the endpoints, disease trajectory and patient identification pathway. The same investigator centers and practitioners, already trained on the delivery procedure, carry that experience into HearConnex. With validated biology and a unique collaboration, Sensorion aims to run a best-in-class trial with HearConnex
Page 86
What comes next Under review Health Canada CTA decision By Q4 2026 1 IND submission in the United States By Q4 2026 1 CTA submission in Australia Early 2027 First patient dosed in HearConnex Throughout 2027 Clinical data generation ir.contact@sensorion-pharma.com Copyright by Sensorion 2026. All Rights Reserved 1 Expected milestones. 86
Page 87
Thank you Copyright by Sensorion 2026. All Rights Reserved 87