IMS 2026 Slide Library

CERVINO treatment-emergent adverse events (TEAEs), Etentamig (N=195) vs SAT (N=193). Any TEAE 97.9% vs 95.3%; Grade 3/4 70.8% vs 59.6%; Grade 5 (fatal) 2.6% vs 5.7% - roughly half the fatal AE rate with Etentamig. Neutropenia 46.7% vs 23.3%. CRS (among SUD recipients, N=113): 28.3%, no grade 3 or higher. ICANS 0.9%, no grade 2+. Infections any grade 69.7% (grade 3/4 27.7%, grade 5 1.5%) vs 59.1% (grade 3/4 19.2%, grade 5 3.1%). Serious TEAEs 51.8% vs 40.4%. Dose discontinuation due to AEs 3.6% vs 14.0%. Etentamig redefines anti-BCMA BsAb safety with low CRS and severe infection rates and a >=50% reduction of fatal AEs and fatal infections vs SAT. CERVINO CRS incidence and management by step-up dosing strategy. All patients (N=195): CRS 39.5% (Grade 1 27%, Grade 2 11%, Grade 3 1%). Single step-up dose (SUD) subgroup (n=113): CRS 28.3% (Grade 1 24%, Grade 2 4%, Grade 3 0%). Single SUD plus prophylactic tocilizumab (n=22): CRS 0%. Median time to CRS onset 17.6 hours (single SUD 20.9 hrs); median time to CRS resolution 8.5 hours (single SUD 6.0 hrs). CRS recurrence after Cycle 1: 0.5%. ICANS 3.6% overall (single SUD 0.9%). Etentamig's unique design with low CD3 affinity enables a predictable CRS profile, potentially supporting broad implementation across diverse treatment settings.

CERVINO PFS forest plot by subgroup - benefit favors Etentamig across all subgroups evaluated: sex (male HR 0.53, female HR 0.29), age <65y HR 0.59 / >=65y HR 0.32, age <75y HR 0.46 / >=75y HR 0.27, race (White HR 0.36, Black/African American HR 0.23, Asian HR 0.58), region (North America US-only HR 0.29, Europe HR 0.39), baseline ISS I/II HR 0.37 vs III HR 0.57, ECOG PS 0 HR 0.44 / >=1 HR 0.37, academic HR 0.42 / non-academic site HR 0.39, high cytogenetic risk HR 0.35 / standard risk HR 0.43, prior lines <=2 HR 0.36 / 3 lines HR 0.41 / >=4 lines HR 0.45, double-refractory HR 0.43 / triple-refractory HR 0.39. CERVINO duration of response (DoR) and MRD negativity. DoR: HR 0.31 (95% CI 0.18-0.53); median DoR not reached with Etentamig vs 10.2 months with SAT; 12-month DoR rate 79.6% vs 48.1%. Median time to response: 1.12 months Etentamig vs 1.58 months SAT. MRD negativity among evaluable >=CR patients: at MRD<10^-5, 89.1% (Etentamig, n=64) vs 25.0% (SAT, n=8); at MRD<10^-6, 82.8% vs 12.5%. Etentamig achieved deeper, more durable responses than SAT, supporting its potential for sustained remission. CERVINO secondary endpoint overall survival (early signal, immature). HR 0.48 (95% CI 0.29-0.77), P=0.0012 (prespecified efficacy boundary for OS, P=0.000017, was NOT crossed at this data cutoff). 12-month OS 87.9% Etentamig vs 72.0% SAT; median OS not reached in either arm. Non-relapse mortality 5.1% vs 9.8%; infection-related non-relapse mortality 1.5% vs 3.1%; relapse mortality 9.2% vs 15.0%. 75% of SAT-treated patients who received post-study therapy went on to a T-cell engager or CAR-T therapy. At 12 months, almost 90% of patients on Etentamig were still alive.

CERVINO patient disposition (CONSORT). Screened N=498, failed screening n=105, randomized (ITT) n=393. Etentamig n=196 (received 195); SAT n=197 (received 193). Discontinued Etentamig n=79; remain on Etentamig n=117. Discontinued SAT n=142; remain on SAT n=55 - twice as many patients remained on Etentamig vs SAT. Primary reason for discontinuation: progressive disease 61 (31.1%) Etentamig vs 100 (50.8%) SAT; AEs 7 (3.6%) vs 19 (9.6%). AE-related discontinuation under 5% reinforces Etentamig's favorable tolerability. Data cutoff: June 10, 2026. CERVINO primary endpoint ORR (International Myeloma Working Group criteria, independent review). Etentamig ORR 74.0% (>=CR 40%, VGPR 23%, PR 11%; >=VGPR 63%) vs SAT 45.7% (>=CR 7%, VGPR 14%, PR 25%; >=VGPR 20%). Difference 28.3% (95% CI 18.48-37.46), P<0.0001. Median follow-up 11.4 (0.5-24.3) months. Etentamig demonstrated substantially higher ORR and >=CR. CERVINO primary endpoint PFS (Kaplan-Meier). HR 0.40 (95% CI 0.29-0.54), P<0.0001. Median PFS: Etentamig not reached (NR); SAT 6.2 months. 12-month PFS rate: 62.1% Etentamig vs 28.4% SAT. At 11.4 months median follow-up, Etentamig reduced the risk of disease progression or death by 60%, with twice as many patients progression-free at 12 months vs SAT. Data cutoff: June 10, 2026.

CERVINO study design summary. Etentamig: IV, given less often, with one single step-up dose (SUD) given prior to full dose. Patients can stop therapy after 2 years. Comparator arm (SAT, standard available therapy): Kd (carfilzomib + dexamethasone), SVd (selinexor + bortezomib + dexamethasone), and EloPd (elotuzumab + pomalidomide + dexamethasone) - EloPd not commonly used in the US currently. Data cutoff: June 10, 2026. CERVINO baseline characteristics table. Etentamig (N=196) vs SAT (N=197): median age 69 vs 70 years; Triple-class exposed 196 (100%) vs 197 (100%); Triple-class refractory 101 (51.5%) vs 103 (52.3%); median prior lines of therapy 3 (range 2-10) vs 3 (range 1-8); North America (US only) 24 (12.2%) vs 33 (16.8%); high cytogenetic risk 41 (20.9%) vs 51 (25.9%). CERVINO enrolled a diverse population with a substantial proportion treated in non-academic settings, reflecting today's clinical practice.

Other challenges When to start treatment? - The earlier the use of these therapies -> The better - Should we use them at biochemical PD or clinical PD -> Biochemical PD is better but... - Would it be possible to start in case of relapse from MRD-ve? Schedule of treatment - Should we move to treatments of fixed duration, especially with Bispecific mAbs? - Fixed duration or duration adapted to the MRD? - Would it be possible to implement maintenance after CAR-T for maintaining the MRD-ve - Is it possible to extend the interval between doses to improve the safety profile? Summary: A future decision framework 01 RIGHT PATIENT — Disease tempo; Frailty and immune fitness; Prior exposure and refractoriness 02 RIGHT PLATFORM — CAR-T - bispecific - ADC; CELMoD-based combination; Conventional therapy 03 RIGHT BIOLOGY AND TARGET — BCMA - GPRC5D - CD38; Cereblon-driven activity; Preserve alternative targets 04 RIGHT TIME AND SEQUENCE — Treat before attrition; Match therapy to disease biology; Preserve the next effective option CELMoDs expand the choice beyond immune targeting and create new sequencing opportunities in early relapse The future of early relapse will rely on individualized sequencing across immune therapies and CELMoD-based combinations, maximizing survival and quality of life while preserving effective options for later relapse Will the future resolve these challenges? 2nd line BCMA as target: CAR-T cells: Cilta-cel; Anito-cel; BCMA-CD19 CAR-T cells BsAbs: Teclistamab-Dara; Teclistamab; Etentamig/Elranatamab/Linvo ADCs: Bela-Vd; Bela-Pd GPRC5D as target: CAR-T cells: Arlo-cel BsAbs: Talquetamab-Dara; Talquetamab-Dara-Pom BCMA-GPRC5D: CAR-T cells: BCMA-GPRC5d-CART BsAbs: Talquetamab-Teclistamab CelMoD's will help to restore the immune system and to sequence T-cell redirecting therapies or they will have a role instead of and their use may preserve BCMA or GPRC5D targeting for later relapse Other options in the future will include: in vivo CAR-T cells, trispecific monoclonal antibodies, CELMoD's-based combos

Phase 3 evidence has moved inmune therapy into early relapse [Table: STUDY | COMPARISON | forest plot (FAVOURS EXPERIMENTAL / FAVOURS CONTROL) | PFS HR | OS HR] CARTITUDE-4 | Cilta-cel vs DPd/PVd | PFS HR 0.29 | OS HR 0.55 DREAMM-7 | BVd vs DVd | PFS HR 0.41 | OS HR 0.58 DREAMM-8 | BPd vs PVd | PFS HR 0.52 | OS HR 0.77 MajesTEC-3 | Tec-Dara vs DPd/DVd | PFS HR 0.17 | OS HR 0.46 MonumenTAL-3 | Tal-DP vs DPd | PFS HR 0.28 | OS HR 0.47 (MonumenTAL-3) | Tal-D vs DPd | PFS HR 0.33 | OS HR 0.51 MajesTEC-9 | Teclistamab vs PVd/Kd | PFS HR 0.29 | OS HR 0.60 MonumenTAL-6 | Tal-Tec vs EPd/PVd | PFS HR 0.11 | OS HR NR (MonumenTAL-6) | Tal-Pom vs EPd/PVd | PFS HR 0.27 | OS HR NR CERVINO | Etentamig vs investigator's choice | PFS HR 0.40 | OS HR NR RANDOMIZED PHASE 3 TRIALS WITHOUT PUBLIC NUMERICAL HRs MagnetisMM-5 | Elranatamab +/- Dara vs DPd | Positive top-line result; no numerical data reported [Forest plot x-axis tick labels largely illegible]

DON'T GET LOST AT IMS 2026 GLASGOW - SEPTEMBER 23-26 THE MUST-SEE ROADMAP PEOPLE SCIENCE PROGRESS TOGETHER | BIGGER SCIENCE A BRIGHTER TOMORROW | GLASGOW WELCOMES A BRIGHTER TOMORROW THURSDAY - SEPT 24 5:00-6:00 PM - HALL 5 MRD & MAINTENANCE (MRD: DEEPER ANSWERS BRIGHTER TOMORROWS) - MIDAS: Isa-iberdomide - CASSIOPEIA: 5-year sustained MRD - GEM2014MAIN: MRD-guided stopping - IBEX: Iber-Dara post-ASCT FRIDAY - SEPT 25 3:00-5:00 PM - HALL 5 DO NOT MISS THE PLENARY - DO NOT MISS (BOLD SCIENCE CHANGES LIVES) - CERVINO: Etentamig vs standard therapy - PERSEUS: Does long-term PFS become OS? - EXCALIBER-RRMM: IberDd vs DVd - First-in-human CD19/BCMA in-vivo CAR-T SAME QUESTIONS. A BRIGHTER TOMORROW. SATURDAY - SEPT 26 8:30-9:30 AM - HALL 5 THE NEXT CELLULAR-THERAPY ERA (CELLULAR THERAPIES: REAL POTENTIAL FOR REAL PEOPLE) - Updated CEPHEUS MRD - BCMAxGPRC5D dual-target CAR-T - IASO206: IV in-vivo BCMA CAR-T 11:00 AM-12:00 PM - HALL 5 - Post-CAR-T elranatamab - Delayed cilta-cel neurotoxicity - MRD-guided T-cell redirection 11:00 AM-12:00 PM - LOMOND - CAR-T enterocolitis biology 1:00-2:00 PM - LOMOND - LV169: lentiviral in-vivo BCMA CAR delivery Same goal. Brighter lives. WHAT TO WATCH (DIFFERENT PERSPECTIVES A CLEARER FUTURE) - Durability - OS - sustained MRD - infection burden - Access - treatment-free time - late toxicity - In-vivo CAR-T: insertion sites, off-target transduction, vector shedding & persistence Go for the randomized data and the biology--not every new response rate. Single malt only AFTER the final session. SAME CURIOSITY FURTHER TOGETHER | GREAT SCIENCE DESERVES A DRAM

[Slide 1 - Incidence and severity of infections across different approved BsAbs and CAR T trials, cross-trial compilation table] BsAbs late-line relapse (triple-class exposed) - trial/drug/All grade/Grade 3-4/Grade 5: MajesTEC-1/Teclistamab: 78.8% / 55.2% / 12.7%. MagnetisMM-3/Elranatamab: 69.9% / 39.8% / 6.5%. LinkerMM-1/Linvoseltamab: 75% / 48% / NR. MonumenTAL-1/Talquetamab: 66.2% / 14.5% / 1.4%. RedirecTT-1/Tec-Tal: 89% / 64% / 12%. BsAbs early-line relapse (1-3 prior lines): MajesTEC-9/Teclistamab: 82.8% / 41.6% / 5.5%. MajesTEC-3/Tec-Dara: 96.5% / 41.6% / 5.5%. MonumenTAL-3/Tal-Dara: 83.0% / 29.2% / 1.5%. MonumenTAL-3/Tal-Dara-Pom: 84.3% / 37.7% / 0.7%. CAR T late-line relapse: CARTITUDE-1/Cilta-Cel: All grade 58%, Gr3-4 20%, Gr5 NR. KarMMa/Ide-Cel: 69%, 22%, NR (#, deaths from AEs not otherwise specified 7% total). CAR T early-line relapse: CARTITUDE-4/Cilta-Cel: 59%, 24%, 4%. KarMMa-3/Ide-Cel: 56%, 22%, NR (*, death from AEs 6% total). "Infections remain a major safety concern with T-cell-redirecting therapies, particularly with continuous bispecific antibody treatment." Table provided for ease of viewing, no comparisons intended. Sources: Moreau P et al NEJM 2022; Lesokhin AM et al Nat Med 2023;29(9):2259-2267; Lee HC et al Clin Lymphoma Myeloma Leuk 2026 Feb, 26(2):e201-e212,e8; Chari A et al NEJM 2022; Cohen Y et al Engl J Med 2025;392:138-49; Touzeau C et al N Engl J Med 2026;395:440-53; Costa LJ et al N Eng J Med 2026;394:739-52; Mina R et al N Engl J Med 2026;395:671-83; Einsele H et al, Lancet Oncol 2026;27:254-68; Ailawadhi S et al, Blood 2024 (144)23; Berdeja J et al, Lancet 2021;398:314-24; Munshi NC et al, N Engl J Med 2021 Feb 25;384(8):705-716. [Slide 2 - Pillars of infection prevention in patients receiving T-cell redirecting therapies] Table: Immune defect/risk -> Type of infection -> Preventive strategy. Lymphopenia/Neutropenia -> Bacterial infections/PJP, approx first 3-6 months -> Antiviral + PJP prophylaxis, G-CSF support. Immunosuppressive therapy (steroids, IL-6 blockage, cyclophosphamide) -> Bacterial infections/Sepsis/fungal infections/CMV -> Consider antibacterial/antifungal. B-cell aplasia/Hypogammaglobulinemia -> Bacterial/Viral (Respiratory virus, VZV/HSV)/CMV -> Immunoglobulin replacement, Vaccination. T-cell dysfunction -> Other opportunistic infections, Viral infections, fungal (in selected high risk patients) -> Fixed treatment duration/Schedule optimization. Sources: Rodriguez-Otero P et al, Lancet Oncology May 2024; O'Neill C, et al. eJHaem. 2023;4:811-822; Friedrich MJ, et al. Cancer Cell. 2023;41:711-725.e6; Frerichs KA et al, Blood Advances 2024; Mohan M et al, Blood Cancer Journal (2025) 15:74; Tix T et al, Molecular Therapy (2025)33:7; Richardson T et al, Blood Advances (2025) 9:6; Nath K et al, Blood Cancer Journal (2024) 14:88; Reynolds G et al, Blood advances (2023) 7:19; Banerjee R et al, Blood Advances (2025) 9:18; Dr Rodriguez Otero personal opinion.

MajesTEC-3: Time-Stratified Cox Analysis of OS Curve (Surviving %, Tec-Dara vs DPd/DVd, 0-45 months): >10 months: HR, 0.22 (95% CI, 0.13-0.38) ≤10 months: HR, 1.08 (95% CI, 0.64-1.81) Deaths within first 10 months were comparable: 29 Tec-Dara vs 29 DPd/DVd Overall HR, 0.46 (95% CI, 0.32-0.65); P<0.0001 Cause of death within first 10 months (curve cross): Tec-Dara: AEs (n=18), PD (n=5), 'Other' (n=6) DPd/DVd: AEs (n=8), PD (n=20), 'Other' (n=1) Tec-Dara substantially reduced the risk of death by 78% versus DPd/DVd after 10 months Of 18 AE deaths, 12 were infections. 3 patients died due to unknown reasons; 3 occurred outside the TEAE reporting window (n=1, unrelated to study treatment [COVID-19 occurring >30 days after last dose]; n=1, death occurred after initiation of subsequent therapy [septic shock]; n=1, death following diagnosis of secondary amyloidosis). 1 patient died due to unknown reasons. TEAE, treatment-emergent adverse event. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland (slide 13)

Why Competing-Risk Analyses Matter in MajesTEC-3 - NRM and disease progression are mutually exclusive components of PFS - NRM: Death without prior progression, irrespective of cause - NRM and disease progression are also "competing events," requiring separation - A patient who experiences 1 event can no longer experience the other event - Competing-risk analyses (cumulative incidence of progression [CIP] and NRM) are needed to accurately assess what is driving patient survival in MajesTEC-3 Flowchart: "2 possible routes" -> "Alive/on study" splits into: (left) Death after disease progression -> Disease progression (counted toward CIP) -> Subsequent therapy -> Death (right) No prior progression (NRM) -> Death without prior progression (counted toward NRM) Banner: Competing-risk analyses help to separate "deaths after progression" from "deaths without prior progression" Footnotes: aPFS event (defined as the time from the date of randomization to the date of first documented disease progression or death due to any cause, whichever comes first). PFS combines progression and death into a single endpoint but cannot distinguish between them. bOS event (defined as the time from the date of randomization to the date of the patient's death due to any cause). OS captures all deaths but cannot distinguish deaths after progression from deaths without prior progression (NRM). Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 5 MajesTEC-3: Causes of Non-Relapse Mortality NRM was primarily due to AEs in both treatment groups - AEs were primarily infections and infestations, occurring most commonly within 6 months of treatment initiation - Tec-Dara: 12; 11 ≤6 months(a) - DPd/DVd: 4; 2 ≤6 months - Infection prophylaxis was recommended per protocol,(1) with IgRT and antimicrobial use reinforced as experience and guidelines evolved - Supportive RWE and IMWG guidelines recommend IgRT prophylaxis for the reduction of high-grade infections with BCMA-targeting BsAbs(2-5) Table - NRM, n | Tec-Dara (n=28) | DPd/DVd (n=26): AE-related(b): 19 | 16 Infections and infestations: 12 | 4 General disorders and administration-site conditions: 2 | 4 Respiratory, thoracic, and mediastinal disorders: 2 | 3 Nervous system disorders: 1 | 2 Hepatobiliary disorders: 1 | 1 Cardiac disorders: 1 | 0 Immune system disorders: 1 | 0 Neoplasms benign, malignant, and unspecified (including cysts and polyps): 0 | 1 Gastrointestinal disorders: 0 | 1 PD without IRC-determined progression: 2 | 8 Other: 7 | 2 Banner: Infection-related NRM may be reduced through diligent use of established IgRT and antimicrobial prophylaxis protocols Footnotes: aOnly 3 out of 11 Tec-Dara patients received ≥1 dose of IgRT. bListed per System Organ Class. Of note, patients may have been listed as having more than 1 cause of death, which may have been due to 1 or more system organ class. AE, adverse event; BsAb, bispecific antibody; IgRT, immunoglobulin replacement therapy; PD, progressive disease; RWE, real-world evidence. 1. Costa LJ, et al. N Engl J Med. 2026;394(8):739-752. 2. Smits F, et al. Blood Cancer J. 2026;16(1):26. 3. Lancman G, et al. Blood Cancer Discov. 2023;4(6):440-451. 4. Cheruvalath H, et al. Blood. 2024;144(Supp 1):256. 5. Rodriguez-Otero P, et al. Lancet Oncol. 2024;25(5):e205-e216. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 9 MajesTEC-3: Time-Stratified Cox Analysis of OS Kaplan-Meier style survival curve (Surviving, % vs Time, months 0-45), two arms: Tec-Dara (red) and DPd/DVd (black) >10 months: HR, 0.22 (95% CI, 0.13-0.38) ≤10 months: HR, 1.08 (95% CI, 0.64-1.81) Deaths within first 10 months were comparable: 29 Tec-Dara vs 29 DPd/DVd Overall: HR, 0.46 (95% CI, 0.32-0.65); P<0.0001 Cause of death within first 10 months (curve cross): - Tec-Dara: AEs (n=18),(a) PD (n=5), "Other" (n=6)(b) - DPd/DVd: AEs (n=8), PD (n=20), "Other" (n=1)(c) Banner: Tec-Dara substantially reduced the risk of death by 78% versus DPd/DVd after 10 months Footnotes: aOf 18 AE deaths, 12 were infections. b3 patients died due to unknown reasons; 3 occurred outside the TEAE reporting window (n=1, unrelated to study treatment [COVID-19 occurring >30 days after last dose]; n=1, death occurred after initiation of subsequent therapy [septic shock]; n=1, death following diagnosis of secondary amyloidosis). c1 patient died due to unknown reasons. TEAE, treatment-emergent adverse event. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 13 Conclusions - Tec-Dara significantly improved PFS (HR, 0.17) and OS (HR, 0.46) over DPd/DVd in patients with RRMM and 1-3 prior LOTs,(1,2) with a sustained plateau in survival curves - With Tec-Dara, progressions became uncommon after 6 months, with an unparalleled reduction in disease progression (sHR, 0.10), which is the leading cause of mortality in RRMM(3) - There was no meaningful difference in NRM between treatments over time; infection-related NRM occurred with both treatments - Infections can be better managed now with guideline-based IgRT and antimicrobial prophylaxis(4) - OS benefit with Tec-Dara versus DPd/DVd was driven by durable disease control without an increase in NRM - Superior OS with Tec-Dara was supported by RMST(1) and time-stratified Cox analysis (>10 months) Banner: Tec-Dara's OS benefit is driven by profound suppression of disease progression, while NRM remains comparable over time 1. Costa LJ, et al. N Engl J Med. 2026;394(8):739-752. 2. Mateos MV, et al. Presented at: 67th ASH Annual Meeting and Exposition; December 6-9, 2025; Orlando, FL. Oral LBA-6. 3. Tan CR, et al. Blood Adv. 2026;10(9):2937-2946. 4. Rodriguez-Otero P, et al. Lancet Oncol. 2024;25(5):e205-e216. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 14

Model Estimates of Mortality Risk With Tec-Dara Bullet: Across tested models,[a] estimated statistical cure fractions were consistently high with Tec-Dara,[b] and low with DPd/DVd Table - Statistical cure fraction, % | Tec-Dara | DPd/DVd Best-fit model (95% CI)[c,d] | 86.6 (81-91) | 0 (0-53) Range across tested models | 84.4 to 86.8 | 0 to 51.3 Right chart "Relative survival[e]": line chart, y-axis "Relative survival, %" 0-100, x-axis "Time (years)" 0-25. Tec-Dara line (blue) plateaus near 87% out to 25 years, annotated "Statistical cure fraction 86.6%" (green triangle marker ~86.6%). DPd/DVd line (gray) declines steadily from 100% to about 5% by 25 years. Black banner: ~87% of patients treated with Tec-Dara are estimated to have no excess long-term mortality risk on top of the general population by best-fit models Footnote: aTested models analyzed survival related to excess risk in the uncured fraction using 7 classical parametric models including exponential, Weibull, lognormal, loglogistic, Gompertz, gamma, and generalized gamma. bResults from the Gompertz function were excluded for Tec-Dara because it resulted in a non-zero ceiling in excess mortality due to the shape parameter being negative. cBest-fit models were selected according to statistical goodness of fit criteria; MCMs applied the exponential function for Tec-Dara OS and DPd/DVd OS. dProfile-likelihood 95% CIs shown here; these replace the Wald CIs reported in the abstract and typically provide more accurate coverage. eFrom best-fit model. Presented by NWCJ van de Donk at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland [Slide 11]

Why Competing-Risk Analyses Matter in MajesTEC-3 - NRM and disease progression are mutually exclusive components of PFS - NRM: Death without prior progression, irrespective of cause - NRM and disease progression are also 'competing events,' requiring separation - A patient who experiences 1 event can no longer experience the other event - Competing-risk analyses (cumulative incidence of progression [CIP] and NRM) are needed to accurately assess what is driving patient survival in MajesTEC-3 Flow diagram: '2 possible routes' -> Alive/on study -> [left] Death after disease progression -> Disease progression (counted toward CIP) -> Subsequent therapy -> Death; [right] No prior progression (NRM) -> Death without prior progression (counted toward NRM) Banner: 'Competing-risk analyses help to separate "deaths after progression" from "deaths without prior progression"' MajesTEC-3: Causes of Non-Relapse Mortality NRM was primarily due to AEs in both treatment groups - AEs were primarily infections and infestations, occurring most commonly within 6 months of treatment initiation - Tec-Dara: 12; 11 <=6 months(a) - DPd/DVd: 4; 2 <=6 months - Infection prophylaxis was recommended per protocol, with IgRT and antimicrobial use reinforced as experience and guidelines evolved - Supportive RWE and IMWG guidelines recommend IgRT prophylaxis for the reduction of high-grade infections with BCMA-targeting BsAbs Table - NRM, n | Tec-Dara (n=28) | DPd/DVd (n=26): AE-related(b): 19 | 16 Infections and infestations: 12 | 4 General disorders and administration-site conditions: 2 | 4 Respiratory, thoracic, and mediastinal disorders: 2 | 3 Nervous system disorders: 1 | 2 Hepatobiliary disorders: 1 | 1 Cardiac disorders: 1 | 0 Immune system disorders: 1 | 0 Neoplasms benign, malignant, and unspecified (including cysts and polyps): 0 | 1 Gastrointestinal disorders: 0 | 1 PD without IRC-determined progression: 2 | 8 Other: 7 | 2 Banner: 'Infection-related NRM may be reduced through diligent use of established IgRT and antimicrobial prophylaxis protocols' Conclusions - Tec-Dara significantly improved PFS (HR, 0.17) and OS (HR, 0.46) over DPd/DVd in patients with RRMM and 1-3 prior LOTs, with a sustained plateau in survival curves - With Tec-Dara, progressions became uncommon after 6 months, with an unparalleled reduction in disease progression (sHR, 0.10), which is the leading cause of mortality in RRMM - There was no meaningful difference in NRM between treatments over time; infection-related NRM occurred [text cut off at right edge of photo] - Infections can be better managed now with guideline-based IgRT and antimicrobial prophylaxis - OS benefit with Tec-Dara versus DPd/DVd was driven by durable disease control without an increase in N[RM, cut off] - Superior OS with Tec-Dara was supported by RMST and time-stratified Cox analysis (>10 months) Banner: 'Tec-Dara's OS benefit is driven by profound suppression of disease progression, while NRM remains comparable over time' Citations (partially legible): 1. Costa LJ, et al. N Engl J Med...394(8):739-752. 2. Mateos MV, et al. Presented at 67th ASH Annual Meeting and Exposition; December 6-9, 2025; Orlando, FL. Oral LBA-6. 3. Tan CR, et al. Blood Adv. 2026;10(5):2937-2946. 4. Rodriguez-Otero P, et al. ... 2026;25(5):e255-e218 Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2025; Glasgow, Scotland [year on footer reads 2025, likely a typo for 2026]

MajesTEC-3: Causes of Non-Relapse Mortality NRM was primarily due to AEs in both treatment groups AEs were primarily infections and infestations, occurring most commonly within 6 months of treatment initiation - Tec-Dara: 12; 11 ≤6 months - DPd/DVd: 4; 2 ≤6 months Infection prophylaxis was recommended per protocol, with IgRT and antimicrobial use reinforced as experience and guidelines evolved Supportive RWE and IMWG guidelines recommend IgRT prophylaxis for the reduction of high-grade infections with BCMA-targeting BsAbs Table - NRM, n | Tec-Dara (n=28) | DPd/DVd (n=26) AE-related: 19 | 16 Infections and infestations: 12 | 4 General disorders and administration-site conditions: 2 | 4 Respiratory, thoracic, and mediastinal disorders: 2 | 3 Nervous system disorders: 1 | 2 Hepatobiliary disorders: 1 | 1 Cardiac disorders: 1 | 0 Immune system disorders: 1 | 0 Neoplasms benign, malignant, and unspecified (including cysts and polyps): 0 | 1 Gastrointestinal disorders: 0 | 1 PD without IRC-determined progression: 2 | 8 Other: 7 | 2 Infection-related NRM may be reduced through diligent use of established IgRT and antimicrobial prophylaxis protocols Only 3 out of 11 Tec-Dara patients received ≥1 dose of IgRT. Listed per System Organ Class. Of note, patients may have been listed as having more than 1 cause of death, which may have been due to 1 or more system organ class. AE, adverse event; BsAb, bispecific antibody; IgRT, immunoglobulin replacement therapy; PD, progressive disease; RWE, real-world evidence. 1. Costa LJ, et al. N Engl J Med. 2026;394(8):739-752. 2. Smits F, et al. Blood Cancer J. 2026;16(1):26. 3. Lancman G, et al. Blood Cancer Discov. 2023;4(6):440-451. 4. Cheruvalath H, et al. Blood. 2024;144(Supp 1):256. 5. Rodriguez-Otero P, et al. Lancet Oncol. 2024;25(5):e205-e216. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland (slide 9) MajesTEC-3: Non-Relapse Mortality With CIP as a competing risk, NRM was similar between treatments Deaths without progression - Tec-Dara: 28/291; 19 due to AEs - DPd/DVd: 26/296; 16 due to AEs NRM was similar between groups regardless of early AE-related deaths Cumulative incidence of non-relapse mortality curve (Tec-Dara vs DPd/DVd, 0-48 months): sHR, 1.16 (95% CI, 0.69-1.98); nominal P=0.5668 36-mo NRM rate: DPd/DVd 10.2%, Tec-Dara 9.0% No. at risk - Tec-Dara: 291 256 245 238 231 221 99 9 0 (at 0,6,12,18,24,30,36,42,48 months); DPd/DVd: 295 228 175 145 123 98 32 4 0 Censored - Tec-Dara: 0 9 10 10 13 19 140 230 239; DPd/DVd: 0 5 5 5 5 12 70 95 99 No meaningful difference between Tec-Dara and DPd/DVd in NRM Rate of dying from non-relapse causes when taking into consideration that patients who experience disease progression first are no longer 'at risk' of NRM. In the Tec-Dara group, 12 out of 19 AEs were infections; in the DPd/DVd group, 4 out of 16 were infections. Other NRM deaths included deaths attributed to progressive disease without IRC-determined progression (Tec-Dara, n=2; DPd/DVd, n=8) and 'Other' (Tec-Dara, n=7; DPd/DVd, n=2). Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland (slide 10) MajesTEC-3: Restricted Mean Survival Time (RMST) OS Pre-planned RMST analysis supported an OS benefit with Tec-Dara versus DPd/DVd Tec-Dara increased average survival time through 34.1 months post-randomization (nominal P=0.0088) RMST is an alternative, clinically interpretable measure of treatment effect, particularly when proportional hazards are not met Calculated as area under the KM curve up to a prespecified time point (tau) Curve (Probability of patients being alive, 0-48 months): tau= 34.1 mo; Tec-Dara additional average survival time with Tec-Dara vs DPd/DVd HR, 0.46 (95% CI, 0.32-0.65); P<0.0001 RMST: P=0.0088 RMST demonstrated consistent OS benefit with Tec-Dara versus DPd/DVd 34.1 months represents the smaller value of the longest OS event time observed from either arm at the time of clinical cut-off (Tec-Dara, 34.1 months; DPd/DVd, 41.4 months). The estimated average survival time through 34.1 months was 30.3 months for Tec-Dara and 28.2 months for DPd/DVd; RMST difference was 2.15 months, thus favoring Tec-Dara. KM, Kaplan-Meier. 1. Costa LJ, et al. N Engl J Med. 2026;394(8):739-752. 2. Mateos MV, et al. Presented at: 67th ASH Annual Meeting and Exposition; December 6-9, 2025; Orlando, FL. Oral LBA-6. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland (slide 12) MajesTEC-3: Time-Stratified Cox Analysis of OS Curve (Surviving %, Tec-Dara vs DPd/DVd, 0-45 months): >10 months: HR, 0.22 (95% CI, 0.13-0.38) ≤10 months: HR, 1.08 (95% CI, 0.64-1.81) Deaths within first 10 months were comparable: 29 Tec-Dara vs 29 DPd/DVd Overall HR, 0.46 (95% CI, 0.32-0.65); P<0.0001 Cause of death within first 10 months (curve cross): Tec-Dara: AEs (n=18), PD (n=5), 'Other' (n=6) DPd/DVd: AEs (n=8), PD (n=20), 'Other' (n=1) Tec-Dara substantially reduced the risk of death by 78% versus DPd/DVd after 10 months Of 18 AE deaths, 12 were infections. 3 patients died due to unknown reasons; 3 occurred outside the TEAE reporting window (n=1, unrelated to study treatment [COVID-19 occurring >30 days after last dose]; n=1, death occurred after initiation of subsequent therapy [septic shock]; n=1, death following diagnosis of secondary amyloidosis). 1 patient died due to unknown reasons. TEAE, treatment-emergent adverse event. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland (slide 13)

MajesTEC-3: Study Design Phase 3, multicenter, open-label trial conducted in 150 sites across 20 countries. Key inclusion criteria: - RRMM - 1-3 prior LOTs including a PI and lenalidomide - Patients with only 1 prior LOT must have been lenalidomide refractory per IMWG criteria - ECOG PS score of 0-2 Key exclusion criteria: - Prior BCMA-directed therapy - Refractory to anti-CD38 mAbs 1:1 randomization N=587 (22 Oct 2021 to 29 Sept 2023) -> Tec-Dara N=291, SC dosing following Dara schedule -> DPd/DVd N=296, by investigator's choice Primary endpoint: PFS per IRC Key secondary endpoints: >=CR and ORR; MRD negativity (10^-5); OS; MySIm-Q Total Symptom score Other secondary endpoints: Safety; PK and immunogenicity Fully immunotherapy-based Tec-Dara regimen with a steroid-sparing schedule after Cycle 1 Day 8 and monthly dosing after 6 cycles Footnotes: Prior exposure to anti-CD38 mAbs was permitted. During the COVID-19 pandemic. DPd/DVd were administered per the approved schedules. Response and disease progression were assessed by a blinded IRC per IMWG criteria. Dexamethasone, acetaminophen, and diphenhydramine pre-medication was required for the first 2 weeks; subsequent dexamethasone was not required thereafter. BCMA, B-cell maturation antigen; CR, complete response; ECOG PS, Eastern Cooperative Oncology Group performance status; IMWG, International Myeloma Working Group; IRC, independent review committee; mAbs, monoclonal antibodies; MRD, minimal residual disease; MySIm-Q, Multiple Myeloma Symptom and Impact Questionnaire; ORR, overall response rate; PI, proteasome inhibitor; PK, pharmacokinetics; SC, subcutaneous. 1. Costa LJ, et al. N Engl J Med. 2026;394(8):739-752. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 4 Why Competing-Risk Analyses Matter in MajesTEC-3 - NRM and disease progression are mutually exclusive components of PFS - NRM: Death without prior progression, irrespective of cause - NRM and disease progression are also "competing events," requiring separation - A patient who experiences 1 event can no longer experience the other event - Competing-risk analyses (cumulative incidence of progression [CIP] and NRM) are needed to accurately assess what is driving patient survival in MajesTEC-3 [Flow diagram] Alive/on study -> 2 possible routes: "Death after disease progression" (purple person icon) or "No prior progression (NRM)" (red person icon) Death after disease progression -> Disease progression (counted toward CIP) -> Subsequent therapy -> Death No prior progression (NRM) -> Death without prior progression (counted toward NRM) Competing-risk analyses help to separate "deaths after progression" from "deaths without prior progression" Footnotes: PFS event (defined as the time from the date of randomization to the date of first documented disease progression or death due to any cause, whichever comes first). PFS combines progression and death into a single endpoint but cannot distinguish between them. OS event (defined as the time from the date of randomization to the date of the patient's death due to any cause). OS captures all deaths but cannot distinguish deaths after progression from deaths without prior progression (NRM). Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 5 Statistical Analysis Table with 3 columns: Primary analysis: PFS and OS HR with 95% CIs from stratified Cox regression model Stratified log-rank test Competing-risk analyses: CIP and NRM Cumulative incidence of progression (CIP): Progression with NRM as competing risk Non-relapse mortality (NRM): Death with progression as competing risk Fine-Gray sub-distribution HRs (sHR) and Gray's test Expanded analyses: OS Restricted mean survival time: Applied when proportional hazards were not met, per SAP Piecewise analysis: Time-stratified Cox analysis (<=10 months vs >10 months) Complementary competing-risk analyses allow for a more accurate interpretation of long-term survival Footnotes: Per IRC. CI, confidence interval; HR, hazard ratio; SAP, statistical analysis plan. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 6 MajesTEC-3: Cumulative Incidence of Progression With NRM as a competing risk, CIP was substantially lower with Tec-Dara - Progression: - Tec-Dara: 24/291; 18 with subsequent death - DPd/DVd: 171/296; 72 with subsequent death Cumulative Incidence of Progression curve, x-axis Time (months) 0-48, y-axis Cumulative incidence of disease progression 0.00-1.00: sHR, 0.10 (95% CI, 0.07-0.16); nominal P<0.0001 36-mo CIP rate: DPd/DVd 62.1%; Tec-Dara 8.7% Number at risk table: Tec-Dara: 291, 253, 241, 231, 223, 214, 89, 9, 0 (months 0,6,12,18,24,30,36,42,48) DPd/DVd: 295, 224, 169, 136, 112, 87, 27, 3, 0 Censored: Tec-Dara: 0,12,14,17,21,26,150,230,239 DPd/DVd: 0,9,11,14,16,23,75,96,99 Tec-Dara substantially reduced CIP versus DPd/DVd; progression is uncommon with 90% reduction versus DPd/DVd Footnotes: Rate of disease progression when accounting for the fact that patients who die prior to progression cannot go on to experience disease progression. Based on sHR. Presented by LJ Costa at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 8

Model Estimates of Median OS for Tec-Dara Left: "Estimated all-cause survival" plot, x-axis Time (years) 0-25, y-axis All-cause survival % 0-100. Blue box annotation "MajesTEC-3: ~3yr follow-up" pointing to early portion of curves. Three curves: General population (black dashed), Tec-Dara (red/orange solid), DPd/DVd (grey solid, steepest decline). Right: "Estimated median survival" bar/arrow chart: General population: 21.1 years Tec-Dara: 18.5 years DPd/DVd: 4.9 years The model projects substantially longer median OS with Tec-Dara, approaching the matched general-population estimate Footnotes: Median follow-up: 34.5 months. Determined using the best-fit relative survival MCM. Determined using the best-fit relative survival model without cure. Presented by NWCJ van de Donk at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 10 Model Estimates of Mortality Risk With Tec-Dara - Across tested models, estimated statistical cure fractions were consistently high with Tec-Dara, and low with DPd/DVd Table - Statistical cure fraction, %: Tec-Dara DPd/DVd Best-fit model (95% CI): 86.6 (81-91) 0 (0-53) Range across tested models: 84.4 to 86.8 0 to 51.3 Right: "Relative survival" plot, x-axis Time (years) 0-25, y-axis Relative survival % 0-100. Two curves: Tec-Dara (blue, plateauing near 86-90%, annotated "Statistical cure fraction 86.6%"); DPd/DVd (grey, declining toward 0 by ~20-25 years). ~87% of patients treated with Tec-Dara are estimated to have no excess long-term mortality risk on top of the general population by best-fit models Footnotes: Tested models analyzed survival related to excess risk in the uncured fraction using 7 classical parametric models including exponential, Weibull, lognormal, loglogistic, Gompertz, gamma, and generalized gamma. Results from the Gompertz function were excluded for Tec-Dara because it resulted in a non-zero ceiling in excess mortality due to the shape parameter being negative. Best-fit models were selected according to statistical goodness of fit criteria; MCMs applied the exponential function for Tec-Dara OS and DPd/DVd OS. Profile-likelihood 95% CIs shown here; these replace the Wald CIs reported in the abstract and typically provide more accurate coverage. From best-fit model. Presented by NWCJ van de Donk at the 23rd International Myeloma Society (IMS) Annual Meeting and Exposition; September 23-26, 2026; Glasgow, Scotland 11

Phase 3 evidence has moved inmune therapy into early relapse [Table: STUDY | COMPARISON | forest plot (FAVOURS EXPERIMENTAL / FAVOURS CONTROL) | PFS HR | OS HR] CARTITUDE-4 | Cilta-cel vs DPd/PVd | PFS HR 0.29 | OS HR 0.55 DREAMM-7 | BVd vs DVd | PFS HR 0.41 | OS HR 0.58 DREAMM-8 | BPd vs PVd | PFS HR 0.52 | OS HR 0.77 MajesTEC-3 | Tec-Dara vs DPd/DVd | PFS HR 0.17 | OS HR 0.46 MonumenTAL-3 | Tal-DP vs DPd | PFS HR 0.28 | OS HR 0.47 (MonumenTAL-3) | Tal-D vs DPd | PFS HR 0.33 | OS HR 0.51 MajesTEC-9 | Teclistamab vs PVd/Kd | PFS HR 0.29 | OS HR 0.60 MonumenTAL-6 | Tal-Tec vs EPd/PVd | PFS HR 0.11 | OS HR NR (MonumenTAL-6) | Tal-Pom vs EPd/PVd | PFS HR 0.27 | OS HR NR CERVINO | Etentamig vs investigator's choice | PFS HR 0.40 | OS HR NR RANDOMIZED PHASE 3 TRIALS WITHOUT PUBLIC NUMERICAL HRs MagnetisMM-5 | Elranatamab +/- Dara vs DPd | Positive top-line result; no numerical data reported [Forest plot x-axis tick labels largely illegible]

International Myeloma Society New 2nd line treatments according to patient characteristics Columns: Tec +Ig (MajesTec 9) | Tec Dara +Ig (MajesTec 3) | Tal Dara (MonumenTal 3) | Tal Dara Pom (MonumenTal 3) | Mezi Kd (Successor 2) | [Press release only:] Tal Pom (MonumenTal 6) | Tec Tal +Ig (MonumenTal 6) | Elra +Ig (MagnetisMM 5) CD38 refractory: Yes | Possibly* | Possibly* | Possibly* | Yes | Yes | Yes | ? High risk: Possibly | Yes | Possibly | Possibly | Yes | Yes | Yes | ? EMD: Possibly | Possibly | Possibly | Possibly | Possibly | Possibly | Yes | ? Frail: Yes | Caution | Possibly | Caution | Yes | Possibly | Yes | ? High Infection risk: Possibly | Caution | Yes | Caution | Possibly | Possibly | Caution | ? Tolerability: Yes | Yes | Possibly | Possibly | Yes | Possibly | Caution | ? [Note: row-to-cell alignment approximate due to photo angle] Yes = Preferred / data supports; Possibly = Conditional / Limited data; Caution = possible but caution required Tec = Teclistamab | Tal = Talquetamab | Elra = Elranatamab | Dara = Daratumumab | Pom = Pomalidomide | Ig = Immunoglobulin | Mezi = mezigdomide | K = carilzomib | *based on TRIMM-2 data Based on trial eligibility + published efficacy/safety signals for BCMA & GPRC5D bispecifics. Table represents personal opinion and based on limited evidence! Costa LJ, et al. N Engl J Med. 2026;394(8):739-752, Touzeau C, et al. N Engl J Med. 2026, Mina R, et al. N Engl J Med. 2026, Dimopoulos et al., The Lancet 2026, Johnson & Johnson. press release, 23 July 2026, Pfizer. press release, 29 April 2026.

23rd Annual MEETING & EXPOSITION - International Myeloma Society THURSDAY, SEPTEMBER 24, 2026 | 11:00-12:00 Abstract Session 3 Co-chairs: Meral Beksac and Martin Kortum - Double hit ultra-high risk myeloma treated with Isa-VRD plus cyclophosphamide induction and Isa-VRD consolidation: QoL analysis of the UKMRA RADAR trial in newly diagnosed transplant eligible patients - Karthik Ramasamy - Updated safety and efficacy of ramantamig (Ram; JNJ-5322) at the recommended phase 2 dose (RP2D), demonstrating feasibility of outpatient dosing (OPD) in relapsed/refractory multiple myeloma (RRMM) - Jeffrey Matous - DREAMM-9 Patient-Reported Vision-Related Function with Extended Belantamab Mafodotin Dosing Intervals with Bortezomib, Lenalidomide, Dexamethasone (BVRd) in Transplant Ineligible Multiple Myeloma - Hang Quach - A fixed-duration, modified cevostamab dosing regimen induces durable remissions that continue after completion in BCMA-naive pts with relapsed/refractory multiple myeloma (RRMM): CAMMA 1 Arm A results - Amrita Krishnan - Overall Survival, Progression and Non-Relapse Mortality With Teclistamab Plus Daratumumab (Tec-Dara) vs Dara-Based Triplets in Relapsed/Refractory Multiple Myeloma (RRMM): MajesTEC-3 Post Hoc Analysis - Luciano Costa Panelist photos shown: Dr. Beksac, Dr. Kortum, Dr. Matous, Dr. Ramasamy, Dr. Costa, Dr. Krishnan, Dr. Quach 23rd Annual MEETING & EXPOSITION - International Myeloma Society THURSDAY, SEPTEMBER 24, 2026 | 11:00-12:00 Abstract Session 4 Co-chairs: Lugui Qiu and Niels van de Donk - Augmenting TRAIL-dependent apoptosis enables BCMA CAR-T cells to overcome resistance in multiple myeloma - Mansour Poorebrahim - First-in-human phase Ia/Ib study of QLS32015 (GPRC5DxCD3) in patients (pts) with relapsed/refractory multiple myeloma (RRMM): updated safety and efficacy data - Jian Cui - An AI-Designed BCMA/FcRL5 Targeting TriTE Overcomes Antigen Escape and Outperforms BCMA-Directed Therapy in Multiple Myeloma - Jessica Encinas Mayoral - A CD4-Dominant, T-Helper-Skewed T Cell State Underlies Movement and Neurocognitive Toxicity After BCMA-Directed CAR-T Therapy in Multiple Myeloma - Theresia Akhlaghi - Single-cell antigen:TCR mapping reveals tumor-specific effector T-cell responses in multiple myeloma - David Avigan Panelist photos shown: Dr. Akhlaghi, Dr. Avigan, Dr. Cui, Dr. Encinas Mayoral, Dr. Poorebrahim, Dr. Qiu, Dr. van de Donk

DON'T GET LOST AT IMS 2026 GLASGOW - SEPTEMBER 23-26 THE MUST-SEE ROADMAP PEOPLE SCIENCE PROGRESS TOGETHER | BIGGER SCIENCE A BRIGHTER TOMORROW | GLASGOW WELCOMES A BRIGHTER TOMORROW THURSDAY - SEPT 24 5:00-6:00 PM - HALL 5 MRD & MAINTENANCE (MRD: DEEPER ANSWERS BRIGHTER TOMORROWS) - MIDAS: Isa-iberdomide - CASSIOPEIA: 5-year sustained MRD - GEM2014MAIN: MRD-guided stopping - IBEX: Iber-Dara post-ASCT FRIDAY - SEPT 25 3:00-5:00 PM - HALL 5 DO NOT MISS THE PLENARY - DO NOT MISS (BOLD SCIENCE CHANGES LIVES) - CERVINO: Etentamig vs standard therapy - PERSEUS: Does long-term PFS become OS? - EXCALIBER-RRMM: IberDd vs DVd - First-in-human CD19/BCMA in-vivo CAR-T SAME QUESTIONS. A BRIGHTER TOMORROW. SATURDAY - SEPT 26 8:30-9:30 AM - HALL 5 THE NEXT CELLULAR-THERAPY ERA (CELLULAR THERAPIES: REAL POTENTIAL FOR REAL PEOPLE) - Updated CEPHEUS MRD - BCMAxGPRC5D dual-target CAR-T - IASO206: IV in-vivo BCMA CAR-T 11:00 AM-12:00 PM - HALL 5 - Post-CAR-T elranatamab - Delayed cilta-cel neurotoxicity - MRD-guided T-cell redirection 11:00 AM-12:00 PM - LOMOND - CAR-T enterocolitis biology 1:00-2:00 PM - LOMOND - LV169: lentiviral in-vivo BCMA CAR delivery Same goal. Brighter lives. WHAT TO WATCH (DIFFERENT PERSPECTIVES A CLEARER FUTURE) - Durability - OS - sustained MRD - infection burden - Access - treatment-free time - late toxicity - In-vivo CAR-T: insertion sites, off-target transduction, vector shedding & persistence Go for the randomized data and the biology--not every new response rate. Single malt only AFTER the final session. SAME CURIOSITY FURTHER TOGETHER | GREAT SCIENCE DESERVES A DRAM

[Slide 1 - TRIlogy-1 Update: Durable PFS in the Expanded RP2D Cohort, presented by JV Matous] Expanded RP2D cohort - BCMA/GPRC5D naive (n=47), median follow-up 21.8 months. KM curve of PFS: 18-month PFS (95% CI) 85.5 (70.0-93.4); 24-month PFS (95% CI) 81.3 (63.3-91.0). Patients at risk: 47/44/37/23/16/6/0 at months 0/6/12/18/24/30/36. "Majority of patients progression free and alive at 2 years." Clinical cut-off: July 15, 2026. [Slide 2 - TRIlogy-1 Update: High Rate of Deep Responses and MRD Negativity] Expanded RP2D - BCMA/GPRC5D naive (n=47): ORR 93.6% (44/47) - sCR 63.8%, CR 14.9%, VGPR 12.8%, PR 2.1% (>=CR 78.7%, >=VGPR 91.5%). 19/20 (95%) evaluable BCMA/GPRC5D-naive patients MRD negative at 10^-5; 14/15 (93.3%) at 10^-6. Time to first response median 1.1 months (0.3-5.2); time to best response median 6.9 months (0.3-22.9). "Nearly 80% >=CR rate and >90% of evaluable patients MRD negative (10^-6)." Clinical cut-off July 15 2026, median follow-up 21.8 months.

[Footer from prior slide, partially visible at top: ...reached; PD, progressive disease; PR, partial response; Q8W, every 8 weeks; RP2D, recommended phase 2 dose; sCR, stringent complete response; SD, stable disease; VGPR, very good partial response; Q8W, every 8 weeks. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK (slide 11)] TRIlogy-1 Update: Durable PFS in the Expanded RP2D Cohort Expanded RP2D cohort - BCMA/GPRC5D naive (n=47) Median follow-up: 21.8 months PFS curve (Patients progression free and alive %, 0-36 months): 18-month PFS (95% CI): 85.5 (70.0-93.4) 24-month PFS (95% CI): 81.3 (63.3-91.0) Patients at risk: 47 44 37 23 16 6 0 (at 0,6,12,18,24,30,36 months) Majority of patients progression free and alive at 2 years

TRIlogy-1 Update: Durable PFS in the Expanded RP2D Cohort Kaplan-Meier curve: Expanded RP2D cohort - BCMA/GPRC5D naive (n=47), Median follow-up: 21.8 months Y-axis: Patients progression free and alive, %, 0-100 X-axis: PFS, months, 0-36 18-month PFS (95% CI): 85.5 (70.0-93.4) 24-month PFS (95% CI): 81.3 (63.3-91.0) Patients at risk (by month 0,6,12,18,24,30,36): 47, 44, 37, 23, 16, 6, 0 Banner: Majority of patients progression free and alive at 2 years

TRIlogy-1 Update: High Rate of Deep Responses and MRD Negativity Expanded RP2D - BCMA/GPRC5D naive Bar chart, y-axis ORR % 0-100. Single bar (n=47): total 93.6% (44/47), sCR 63.8%, CR 14.9%, VGPR 12.8%, PR 2.1%. Bracketed sub-totals: >=CR 78.7%, >=VGPR 91.5%. 19/20 (95%) evaluable BCMA/GPRC5D-naive patients MRD negative at 10^-5 - 14/15 (93.3%) at 10^-6 Table - BCMA/GPRC5D naive (n=44): Time to first response, months, median (range): 1.1 (0.3-5.2) Time to best response, months, median (range): 6.9 (0.3-22.9) Nearly 80% >=CR rate and >90% of evaluable patients MRD negative (10^-6) Footnotes: Clinical cut-off: July 15, 2026. Median follow-up: 21.8 months. Response was assessed by the investigator using IMWG 2016 criteria. BCMA, B-cell maturation antigen; CR, complete response; GPRC5D, G protein-coupled receptor class C group 5 member D; MRD, minimal residual disease; ORR, overall response rate; PR, partial response; RP2D, recommended phase 2 dose; sCR, stringent complete response; VGPR, very good partial response. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK 10 TRIlogy-1 Update: Durable, Deepening Responses Over Time Even With Switch to Q8W dosing Swimmer plot, Expanded RP2D BCMA/GPRC5D naive, x-axis Months 0-36. Each horizontal bar = one patient, colored by response depth over time (sCR dark red, CR red, VGPR light blue, PR grey/tan, MR blue, SD green, PD purple). Symbols: arrow = On Treatment as of July 15, 2026; diamond colors = D/C-PD (orange), D/C-AE (dark red), D/C-Patient withdrawal (navy), Death (red), D/C-Physician decision (grey); yellow circle = Switch to Q8W dosing; star = MRD negative (10^-5). Median DOR: NR (not reached) 61.4% of responders switched to Q8W dosing, with responses sustained or deepening post switch Footnotes: Clinical cut-off: July 15, 2026. Median follow-up: 21.8 months. Response was assessed by the investigator using IMWG 2016 criteria. AE, adverse event; BCMA, B-cell maturation antigen; CR, complete response; D/C, discontinued; GPRC5D, G protein-coupled receptor class C group 5 member D; MR, minimal response; MRD, minimal residual disease; NR, not reached; PD, progressive disease; PR, partial response; Q8W, every 8 weeks; RP2D, recommended phase 2 dose; sCR, stringent complete response; SD, stable disease; VGPR, very good partial response. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK 11 TRIlogy-1 Update: >80% 2-year PFS in the Initial RP2D Cohort (median 26.5 months follow-up) Initial RP2D cohort - BCMA/GPRC5D naive (n=27) Median follow up: 26.5 months Kaplan-Meier PFS curve, x-axis PFS months 0-36, y-axis Patients progression free and alive % 0-100. 18-month PFS (95% CI): 88.9 (69.4-96.3) 24-month PFS (95% CI): 84.4 (63.5-93.9) Patients at risk: 27, 27, 25, 23, 16, 6, 0 (at months 0,6,12,18,24,30,36) Majority of patients progression free and alive at 2 years Footnotes: Clinical cut-off: July 15, 2026. BCMA, B-cell maturation antigen; GPRC5D, G protein-coupled receptor class C group 5 member D; PFS, progression-free survival; RP2D, recommended phase 2 dose. 1. Krishnan A, et al. Blood 2025;146(Supplement 1):4042. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK 13

TRIlogy-1 Update: Severe infections plateaued by 12 months Table - Patients, n (%) | Expanded RP2D (n=56) Any Grade | Grade 3/4 Infections: 48 (85.7) | 14 (25.0) Upper respiratory tract infection: 26 (46.4) | 1 (1.8) Pneumonia: 11 (19.6) | 6 (10.7) Urinary tract infection: 7 (12.5) | 1 (1.8) Nasopharyngitis: 6 (10.7) | 0 Covid-19: 6 (10.7) | 0 Hypogammaglobulinemia: 30 (53.6) Received ≥1 dose IgRT: 51 (91.1) If serum IgG <400 mg/dL, IgRT used to maintain levels regardless of active or history of infection Grade ≥3 infections most common in first 6 months and plateaued after 1 year (cumulative incidence curve, 0-35 months, Expanded RP2D): Patients at risk: 56 48 42 40 36 35 34 27 19 17 17 14 11 9 6 4 2 1 0 0 0 (across 0-35 months) Cumulative events: 0 6 9 10 12 13 13 13 13 14 14 14 14 14 14 14 14 14 14 14 14 Incidence of grade ≥3 infections was 25% with proactive infection management Clinical cut-off date: July 15, 2026. Median follow-up: 17.0 months. TEAEs graded by CTCAE v5.0. Shown are infections occurring in ≥10% (any grade) or 5% (grade 3/4). TEAEs listed by descending order of frequency in the total population. Patients with ≥1 TEAE of hypogammaglobulinemia or postbaseline IgG value <400 mg/dL. CTCAE, Common Terminology Criteria for Adverse Events; IgG, immunoglobulin G; IgRT, immunoglobulin replacement therapy; RP2D, recommended phase 2 dose; TEAE, treatment-emergent adverse event. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK (slide 8) TRIlogy-1 Update: Taste Changes, Skin AEs, Nail AEs, and Weight Loss Were Low Grade Table - AE, n (%) | Expanded RP2D (n=56) Taste changes: 37 (66.1); Grade 1/2, %: 42.9 / 23.2; Leading to discontinuation of treatment: 0; Leading to dose modification: 2 (3.6) Non-rash skin: 38 (67.9); Grade 1/2, %: 50.0 / 17.9; Leading to discontinuation of treatment: 1 (1.8); Leading to dose modification: 1 (1.8) Nail related: 36 (64.3); Grade 1/2, %: 62.5 / 1.8; Leading to discontinuation of treatment: 0; Leading to dose modification: 0 Rash related: 15 (26.8); Grade 1/2, %: 16.1 / 8.9; Grade ≥3: 1 (1.8); Leading to discontinuation of treatment: 1 (1.8); Leading to dose modification: 0 Mean weight loss from baseline was <5%, independent of taste changes (line chart, Mean change in weight from baseline % (SE) by Dose SUD-21) Number of patients - Patients with taste changes: 0 37 31 34 34 34 34 31 29 30 27 26 22 18 16 18 16 15 15 14 11 12 11 (by dose SUD through 21); Patients with no taste changes: 19 16 18 16 13 13 13 11 11 9 9 10 7 7 7 6 6 6 6 6 6 5 Majority of taste changes, skin, and nail AEs were grade 1, with only 1 patient discontinuing treatment; mean weight loss from baseline was <5% Clinical cut-off: July 15, 2026. Median follow-up: 17.0 (0.4-34.1) months. Taste changes included dysgeusia, ageusia, hypogeusia, and taste disorder. Per the Common Terminology Criteria for Adverse Events, the maximum severity for taste changes is grade 2. Includes dose delays, skips, and reductions. Non-rash skin adverse events included skin exfoliation, dry skin, palmar plantar erythrodysesthesia syndrome, and pruritus. There were no grade ≥3 events. These events occurred in the same patient. Nail-related adverse events included nail discoloration, nail disorder, onycholysis, onychomadesis, onychoclasis, and nail ridging. Rash adverse events included rash, maculopapular rash, erythematous rash, and erythema. AE, adverse event; RP2D, recommended phase 2 dose; SE, standard error. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK (slide 9) TRIlogy-1 Update: Outpatient Dosing of Ram Was Feasible with No New Safety Concerns Flow diagram: 30 patients designated to receive SUD and first treatment dose outpatient -> 93.3% (n=28) completed SUD outpatient -> 6.7% (n=2) received SUD inpatient: 1 disease burden, 1 patient concern -> 4 required HCRU for AEs ≤2 days post SUD: 1 Out of home medical care, 3 Hospitalization -> 1 CRS (gr 1), febrile neutropenia (gr 3) -> 1 noncardiac chest pain (gr 3) -> 1 tumor pain (gr 3) -> 79.3% (n=23) completed first treatment dose outpatient -> 20.7% (n=6) received first treatment dose inpatient: 2 disease burden, 1 severe weather, 3 AEs -> 2 required HCRU for AEs ≤2 days post first treatment dose: 1 Out of home medical care, 1 Hospitalization 96.6% received prophylactic tocilizumab before SUD CRS mainly gr 1 - SUD: 10.0% (all gr 1) - Treatment dose 1: 6.7% (all gr 1) With prophylactic tocilizumab, CRS at SUD and first treatment dose was infrequent and all grade 1 Patients with events leading to hospitalization at SUD denoted with purple boxes did not receive the first treatment dose outpatient due to the same or related events. One patient died due to refractory disease after SUD and did not receive the first treatment dose; the percentage receiving the first treatment dose is calculated using n=29 as the denominator. Due to grade 1 CRS. Due to grade 1 CRS, grade 2 diarrhea, and grade 1 headache. Due to grade 2 bone pain. Patient received the first treatment dose inpatient due to grade 3 neutropenia. One grade 3 event occurred at dose 2 in a nonresponder with PD. AE, adverse event; CRS, cytokine release syndrome; HCRU, healthcare resource utilization; PD, progressive disease; Ram, ramantamig; SUD, step-up dose. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK (slide 14)

TRIlogy-1 Update: Study Design and Outpatient Considerations Key eligibility criteria: - Triple-class exposed RRMM - ≥18 years of age - ECOG PS of 0 or 1 - Prior BCMA- or GPRC5D-targeted therapy permitted with sponsor approval Treatment at the Ram RP2D: 5 mg SC SUD -> 100 mg SC Q4W treatment dose -> Option to switch to 100 mg SC Q8W(a) - Prophylactic tocilizumab permitted(b) - Infection prophylaxis recommended(c) Outpatient dosing at the start of treatment was evaluated in dedicated RP2D cohorts Endpoints: Primary: Safety Secondary: ORR(d) Exploratory: PFS,(d) MRD negativity Outpatient dosing outcomes: Outpatient completion of SUD and dose 1, CRS, HCRU within 2 days of outpatient Outpatient administration at start of treatment (SUD and first treatment dose): Protocol-defined patient selection and eligibility: - Patients eligible for outpatient dosing regardless of tumor burden - No rapidly progressing disease or deterioration in neurologic status - Outpatient dosing at the start of treatment was not offered to all RP2D cohorts Protocol-defined monitoring and management: - Assessments for CRS, ICANS, and other clinically significant events - Stringent criteria for hospitalization for CRS Footnotes: aPatients can switch to Q8W with sponsor approval. bTocilizumab could be used prophylactically or considered for fever after other causes are excluded; early use could be considered in patients at high risk of severe CRS. cInfection prophylaxes were administered per institutional guidelines. dResponse and disease progression were assessed by International Myeloma Working Group 2016 criteria. BCMA, B-cell maturation antigen; CRS, cytokine release syndrome; ECOG PS, Eastern Cooperative Oncology Group performance status; GPRC5D, G protein-coupled receptor class C group 5 member D; HCRU, healthcare resource utilization; ICANS, immune effector cell-associated neurotoxicity syndrome; MRD, measurable residual disease; ORR, overall response rate; PFS, progression-free survival; Q8W, every 8 weeks; Ram, ramantamig; RP2D, recommended phase 2 dose; RRMM, relapsed/refractory multiple myeloma; SC, subcutaneous; SUD, step-up dose. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK 4 TRIlogy-1 Update: Expanded RP2D Cohort Disposition 56 patients received Ram at the RP2D (Expanded RP2D) -> 30 patients at RP2D designated to receive SUD and 1st treatment dose in outpatient setting 14 (25.0%) patients discontinued treatment: - 9 (16.1%) progressive disease - 1 (1.8%) physician decision(a) - 1 (1.8%) adverse event - 3 (5.4%) death (all deemed unrelated to Ram by the investigator)(b) 42 (75.0%) patients remained on treatment Median (range) follow-up: 17.0 (0.4-34.1) months 20 additional patients (all BCMA/GPRC5D-naive) at the RP2D vs prior reports(1) Clinical cut-off date: July 15, 2026. aPatient had grade 3 myelomatous meningitis, which did not meet IMWG 2016 criteria for disease progression and discontinued treatment due to physician decision. bCauses of death leading to treatment discontinuation included AEs of pneumonia in 2 patients (these deaths occurred 89 and 55 days after the last dose of Ram, were not considered treatment-emergent, and adjudicated as unrelated to Ram by the investigator) and other (refractory disease; this patient had prior BCMA-direct therapy exposure and received teclistamab as the immediate prior line of therapy). AE, adverse event; BCMA, B-cell maturation antigen; GPRC5D, G protein-coupled receptor class C group 5 member D; IMWG, International Myeloma Working Group; Ram, ramantamig; RP2D, recommended phase 2 dose; SUD, step-up dose. 1. Krishnan A, et al. Blood 2025;146(Supplement 1):4042. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK 5 TRIlogy-1 Update: Baseline Characteristics Consistent With Triple-Class Exposed RRMM Table 1 (left): Characteristic | RP2D Expanded (n=56) | RP2D Outpatient (n=30) Age, years, median (range): 68 (43-87) | 69 (47-87) Male, n (%): 34 (60.7) | 21 (70.0) Race, n (%): White: 40 (71.4) | 21 (70.0) Black/African American: 2 (3.6) | 2 (6.7) Asian: 1 (1.8) | 0 Multiple: 2 (3.6) | 0 Unknown/not reported: 11 (19.6) | 8 (26.7) Extramedullary plasmacytomas ≥1,(a) n (%): 4 (7.1) | 2 (6.7) High-risk cytogenetics,(b) n (%): 18 (41.9) | 12 (57.1) ISS stage,(c) n (%): I: 27 (48.2) | 15 (50.0) II: 22 (39.3) | 11 (36.7) III: 7 (12.5) | 4 (13.3) Years since diagnosis, median (range)(d): 7.0 (0.1-18.7) | 7.0 (0.1-11.3) Median prior LOT, n (range): 3 (1-11) | 3 (1-11) Table 2 (right): Characteristic | RP2D Expanded (n=56) | RP2D Outpatient (n=30) Exposure status, n (%): Triple-class(e): 56 (100.0) | 30 (100.0) Penta-drug(f): 22 (39.3) | 12 (40.0) BCMA/GPRC5D exposed: 9 (16.1) | 4 (13.3) Prior BCMA: 8 (14.3) | 4 (13.3) Prior GPRC5D: 1 (1.8) | 0 Antibody-drug conjugate: 2 (3.6) | 1 (3.3) CAR-T therapy: 4 (7.1) | 3 (10.0) Bispecific antibody: 6 (10.7) | 2 (6.7) Refractory status, n (%): PI: 30 (53.6) | 16 (53.3) IMiD: 52 (92.9) | 26 (86.7) Anti-CD38 mAb: 54 (96.4) | 27 (90.0) Triple-class(e): 29 (51.8) | 15 (50.0) Penta-drug(f): 3 (5.4) | 1 (3.3) To last LOT: 51 (91.1) | 25 (83.3) Patient characteristics generally similar among all patients and those in the outpatient dosing cohort Clinical cut-off date: July 15, 2026. aGe1 nonradiated, bone-independent lesion ≥2 cm. bCytogenetic risk is based on FISH or karyotype testing. Defined as del(17p), t(4;14), or t(14;16). Percentages are based on patients with available data (n=46 in the expanded RP2D cohort and n=21 in the outpatient cohort). cISS staging is derived based on serum β2-microglobulin and albumin. dBased on n=55 patients at the RP2D and n=29 patients in the outpatient dosing cohort. ePI, IMiD, and anti-CD38 mAb. f≥2 PIs, ≥2 IMiDs, and anti-CD38 mAb. BCMA, B-cell maturation antigen; CAR, chimeric antigen receptor; FISH, fluorescence in situ hybridization; GPRC5D, G protein-coupled receptor class C group 5 member D; IMiD, immunomodulatory drug; ISS, International Staging System; LOT, line of therapy; mAb, monoclonal antibody; PI, proteasome inhibitor; RP2D, recommended phase 2 dose. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK 6

TRIlogy-1 Update: Safety Consistent with Prior Reports [ref 1] Table - Common TEAEs, >=30% (any Grade) or 10% (Grade 3/4), n (%) | Expanded RP2D (n=56): Any Grade | Grade 3/4 Hematologic TEAEs Neutropenia | 27 (48.2) | 23 (41.8) Lymphopenia | 21 (37.5) | 20 (35.7) Nonhematologic TEAEs Infections | 48 (85.7) | 14 (25.0) Non-rash skin[b] | 38 (67.9) | 0 Taste changes[c] | 37 (66.1) | -- Nail related[d] | 36 (64.3) | 0 CRS | 22 (39.3) | 1 (1.8) Fatigue | 22 (39.3) | 4 (7.1) Cough | 21 (37.5) | 0 Diarrhea | 21 (37.5) | 2 (3.6) Injection site erythema | 19 (33.9) | 0 Hypertension | 10 (17.9) | 7 (12.5) Right-side bullets: - 1 DLT (grade 4 neutropenia) - CRS mostly low grade (grade 1, 30.4%; grade 2, 7.1%) - All at SUD or treatment dose 1, except for 1 grade 3 event at dose 2 in a nonresponder with PD - 52% received prophylactic tocilizumab at SUD - No ICANS - 1 treatment discontinuation due to AEs[e] - No treatment-related deaths Black banner: CRS remains predominantly low grade and confined to SUD and first treatment dose Footnote: Clinical cut-off date: July 15, 2026. Median follow-up: 17.0 (0.4-34.1) months. aTEAEs graded by CTCAE v5.0; CRS per ASTCT criteria. bSkin exfoliation, dry skin, pruritus, and palmar-plantar erythrodysesthesia syndrome. cTaste changes included dysgeusia, ageusia, hypogeusia, and taste disorder. Per the CTCAE, the maximum severity for taste changes is grade 2. dNail-related adverse events included nail discoloration, nail disorder, onycholysis, onychomadesis, onychoclasis, and nail ridging. ePruritus and maculopapular rash. AE, adverse event; ASTCT, American Society for Transplantation and Cellular Therapy; CRS, cytokine release syndrome; CTCAE, Common Terminology Criteria for Adverse Events; DLT, dose limiting toxicity; ICANS, immune effector cell-associated neurotoxicity syndrome; PD, progressive disease; RP2D, recommended phase 2 dose; SUD, step-up dose; TEAE, treatment-emergent adverse event. 1. Krishnan A, et al. Blood 2025;146(Supplement 1):4042. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK. [Slide 7] TRIlogy-1 Update: Taste Changes, Skin AEs, Nail AEs, and Weight Loss Were Low Grade Table - AE, n (%) | Expanded RP2D (n=56) Taste changes[a] | 37 (66.1) Grade 1/2, % | 42.9 / 23.2 Leading to discontinuation of treatment | 0 Leading to dose modification[b] | 2 (3.6) Non-rash skin[c] | 38 (67.9) Grade 1/2,[d] % | 50.0 / 17.9 Leading to discontinuation of treatment | 1 (1.8)[e] Leading to dose modification[b] | 1 (1.8) Nail related[f] | 36 (64.3) Grade 1/2,[d] % | 62.5 / 1.8 Leading to discontinuation of treatment | 0 Leading to dose modification[b] | 0 Rash related[g] | 15 (26.8) Grade 1/2, % | 16.1 / 8.9 Grade >=3 | 1 (1.8) Leading to discontinuation of treatment | 1 (1.8)[e] Leading to dose modification[b] | 0 Right chart: "Mean weight loss from baseline was <5%, independent of taste changes" - line chart, y-axis "Mean change in weight from baseline, % (SE)" from -8 to 8, x-axis "Dose" (SUD, 1-21). Two series: Patients with taste changes (red circles) and Patients with no taste changes (black triangles), both hovering near 0 declining slightly to about -3 to -5% by dose 20-21, with error bars. Number of patients at each dose point (partial, from table under chart): Patients with taste changes: 0, 37, 31, 34, 34, 31, 30, 33, 30, 27, 22, 18, 15, 14, 12, 11, 11 Patients with no taste changes: 19, 16, 18, 16, 13, 16, 13, 11, 7, 9, 10, 7, 7, 7, 6, 5, 6, 6, 5 Black banner: Majority of taste changes, skin, and nail AEs were grade 1, with only 1 patient discontinuing treatment; mean weight loss from baseline was <5% Footnote: Clincal cut-off: July 15, 2026. Median follow-up: 17.0 (0.4-34.1) months. aTaste changes included dysgeusia, ageusia, hypogeusia, and taste disorder. Per the Common Terminology Criteria for Adverse Events, the maximum severity for taste changes is grade 2. bIncludes dose delays, skips, and reductions. cNon-rash skin adverse events included skin exfoliation, dry skin, palmar plantar erythrodysesthesia syndrome, and pruritus. dThere were no grade >=3 events. eThese events occurred in the same patient. fNail-related adverse events included nail discoloration, nail disorder, onycholysis, onychomadesis, onychoclasis, and nail ridging. gRash adverse events included rash, maculopapular rash, erythematous rash, and erythema. AE, adverse event; RP2D, recommended phase 2 dose; SE, standard error. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK. [Slide 9] TRIlogy-1 Update: Durable, Deepening Responses[a] Over Time Even With Switch to Q8W dosing Swimmer-plot chart: y-axis "Expanded RP2D BCMA/GPRC5D naive" (individual patient rows), x-axis "Months" 0-36. Bars colored by response category per legend: sCR (dark red), CR (red), VGPR (blue), PR (light blue), MR (dark blue), SD (green), PD (dark purple). Arrow (->) = On Treatment as of July 15, 2026. Symbols: diamond D/C-PD, diamond D/C-AE, diamond D/C-Patient withdrawal, diamond Death, diamond D/C-Physician decision, yellow circle = Switch to Q8W dosing, star = MRD negative (10^-6). Label on right: "Median DOR: NR" Black banner: 61.4% of responders switched to Q8W dosing, with responses sustained or deepening post switch Footnote: Clinical cut-off: July 15, 2026. Median follow-up: 21.8 months. aResponse was assessed by the investigator using IMWG 2016 criteria. AE, adverse event; BCMA, B-cell maturation antigen; CR, complete response; D/C, discontinued; GPRC5D, G protein-coupled receptor class C group 5 member D; MR, minimal response; MRD, minimal residual disease; NR, not reached; PD, progressive disease; PR, partial response; Q8W, every 8 weeks; RP2D, recommended phase 2 dose; sCR, stringent complete response; SD, stable disease; VGPR, very good partial response. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK. [Slide 11] TRIlogy-1 Update: High Rate of Deep Responses[a] and MRD Negativity Bar chart: "Expanded RP2D - BCMA/GPRC5D naive", y-axis "ORR, %" 0-100. Single stacked bar "BCMA/GPRC5D naive (n=47)": PR 2.1, VGPR 12.8, CR 14.9, sCR 63.8. Bracket labels: >=CR 78.7, >=VGPR 91.5. Header value above bar: 93.6 (44/47). Right bullets: - 19/20 (95%) evaluable BCMA/GPRC5D-naive patients MRD negative at 10^-5 - 14/15 (93.3%) at 10^-6 Table: | BCMA/GPRC5D naive (n=44) Time to first response, months, median (range) | 1.1 (0.3-5.2) Time to best response, months, median (range) | 6.9 (0.3-22.9) Black banner: Nearly 80% >=CR rate and >90% of evaluable patients MRD negative (10^-6) Footnote: Clinical cut-off: July 15, 2026. Median follow-up: 21.8 months. aResponse was assessed by the investigator using IMWG 2016 criteria. BCMA, B-cell maturation antigen; CR, complete response; GPRC5D, G protein-coupled receptor class C group 5 member D; MRD, minimal residual disease; ORR, overall response rate; PR, partial response; RP2D, recommended phase 2 dose; sCR, stringent complete response; VGPR, very good partial response. Presented by JV Matous at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, UK. [Slide 10]

Updated NCT04751877 Data in 1L MM: anti-CD38 mAbs - IMS 2026 Trials: Perseus (EMN) | IMROZ (sanofi) | NCT04751877 (CHU Poitiers) | CEPHEUS (Janssen, Pharmaceutical Companies of Johnson & Johnson) Treatment Regimen: Perseus: Daratumumab + Bortezomib + Lenalidomide + Dexamethasone vs Bortezomib + Lenalidomide + Dexamethasone IMROZ: Isatuximab + Bortezomib + Lenalidomide + Dexamethasone vs Bortezomib + Lenalidomide + Dexamethasone NCT04751877: Isatuximab + Lenalidomide + Dexamethasone +/- Bortezomib vs Isatuximab + Lenalidomide + Dexamethasone CEPHEUS: Isatuximab + Bortezomib + Lenalidomide + Dexamethasone vs Bortezomib + Lenalidomide + Dexamethasone Median Follow-Up: Perseus: 47.5 months | IMROZ: 59.7 months | NCT04751877: 48.9 months | CEPHEUS: 76.0 months PFS: Perseus: 48 mo: 84.3% vs 67.7% IMROZ: 60 mo: 63.2% vs 45.2% NCT04751877: 18 mo: 87.2% vs 85.5%; 24 mo: 85.2% vs 80%; 36 mo: 78% vs 72.3%; 48 mo: 74% vs 65.7% CEPHEUS: 54 mo: 68.1% vs 49.5%; 72 mo: 59.3% vs 38.3% ORR%: Perseus: 96.6% vs 93.8% | IMROZ: 91.3% vs 92.3% | NCT04751877: 85% vs 78% | CEPHEUS: 97% vs 93% MRD negativity rate: Perseus: 75.2% vs 47.5% | IMROZ: 58.1% vs 43.6% | NCT04751877: 12 mo: 51% vs 21% | CEPHEUS: 61.1% vs 40% ≥CR rate: Perseus: 87.9% vs 70.1% | IMROZ: 74.7% vs 64.1% | NCT04751877: 58% vs 31% | CEPHEUS: 80.6% vs 61.4% Adapted from Dr. Akhil Santhosh (X 2025) LARVOL

[Poster PA-288, IMS 23rd Annual Meeting - Long-Term (>=5-Year) Remission and Survival With Ciltacabtagene Autoleucel in Relapsed/Refractory Multiple Myeloma With 1-3 Prior Lines of Therapy: CARTITUDE-2 Cohort A. Adam D. Cohen et al.] CARTITUDE-2 (NCT04133636) Cohort A: patients with 1-3 prior LOT, exposed to a PI and lenalidomide-refractory. Median follow-up 60.7 months, all patients >=5 years from cilta-cel infusion. Target dose 0.75x10^6 CAR+ viable T cells/kg. Key takeaway: Long-term phase 2 data suggest earlier use of cilta-cel in patients with 1-3 prior LOT may be associated with higher likelihood of durable 5-year remissions vs later-line use (1 in 2 patients in earlier lines vs 1 in 3 in later lines). Results: With median follow-up 60.7 months (range 3.3-63.1), 10 of 20 patients (50%) remained alive and progression-free without further anti-myeloma treatment >=5 years after cilta-cel infusion. 18/20 patients (90%) received cilta-cel infusion (per correlative analysis note). Deaths due to AE <5yr n=2; withdrawal without PD <5yr n=1. Baseline characteristics by durable response group (>=5yr alive/PFS n=10 vs PD/death by yr5 n=7): trend toward lower baseline tumor burden in durable responders; 60% of durable responders had decrease in tumor burden from screening to infusion (vs 42.9% in PD/death group); 5/7(71.4%) durable responders had >=1 high-risk cytogenetic feature at baseline (4/7 had >=2) vs 3/7(42.9%)/3/7(42.9%) in PD/death group; 5/10(50%) durable responders were refractory to daratumumab; median time from start of last LOT to progression numerically higher in durable responders (12.8 vs 5.7 months). Efficacy: PFS median 60.5 months (95% CI 12.9-NE), 5-yr PFS rate 54.2% (CART-2 cohort A n=20) vs CARTITUDE-1 33.3% (5-yr PFS 33.7%). OS median NR (19-NE), 5-yr OS rate 69.2% (CART-2) vs CARTITUDE-1 51.3%. 95% (19/20) achieved CR/sCR as best response; one patient had minimal response, died due to COVID-19 pneumonia. Three patients alive/PFS at >=5yr assessed by bone marrow at 5yr, all MRD-negative (10^-6 threshold). Biomarker correlates of durable response: patients alive/PFS at >=5yr had significantly higher CD4+ naive T-cell levels at apheresis (P=0.036) and higher CAR+CD4+ central memory T cells at Tmax post-infusion (P=0.019) vs those who died/progressed <5yr. Safety update since last CARTITUDE-2 report (N=20), new since 30-month report: second primary malignancy (not related to cilta-cel) 5(25.0%) - AML 1, breast cancer 1, prostate cancer 1, basal cell carcinoma 1, melanoma in situ 1, squamous cell carcinoma 1. CAR T-related neurotoxicity: cranial nerve palsies 2, parkinsonism 2(10.0%). Overall mortality 7(35.0%). Death due to AE 1(5.0%). Conclusions: 10/20 patients (50%) in CARTITUDE-2 cohort A remained alive/PFS at >=5 years without further treatment. Median PFS 60.5 months, 5-year OS rate 69.2%. No new cases of neurotoxicity reported in long-term follow-up.

Safety summary n (%) unless stated | N=24 Median follow-up, months (range) | 47.5 (0.8-53.3) AE of any Gr | 24 (100) Gr 3-4 AE | 17 (70.8) Serious AE | 15 (66.7) Gr 5 AE excluding PD | 2 (8.3) - Sepsis* and cardiovascular event (not treatment related) AE leading to cevostamab discontinuation | 2 (8.3) - Pneumonia (treatment related) and HLH and AML (not treatment related) n (%) of patients with AEs of interest by Grade (bar chart, Gr1-5): Neutropenia: 54.2% Anemia: 33.4% Thrombocytopenia: 20.9% CRS: 75.0% Infections: 75.0% Rash: 29.1% With the exception of rash, no substantial skin conditions, hair and nail changes or mucosal toxicities were observed One Gr 1 AE of ICANS was reported The fixed-duration, modified regimen, with a QW/Q2W dosing intensity in C1-6, had a manageable overall safety profile Data cut-off: January 23, 2026. *Histopathological assessment suggested event occurred in the setting of PD. Group term neutropenia, neutrophil count decreased and febrile neutropenia. Group term thrombocytopenia and platelet count decreased. MedDRA System Organ Class Infections and Infestations. MedDRA high-level terms rashes, eruptions and exanthems NEC, exfoliative conditions, dermatitis ascribed to specific agent, acnes, dermatitis and eczema, and bullous conditions. HLH, hemophagocytic lymphohistiocytosis. Response rates Response rates (%) bar chart - PR, VGPR, CR, sCR ORR: 70.8% (95% CI: 53.8, 94.0) sCR: 20.8% CR: 12.5% VGPR: 25.0% PR: 12.5% VGPR+: 58.3% N=24 MRD-negative CR+ rates at the 10^-6 level 8 patients achieved CR+ (33.3%) 4 were MRD evaluable All were MRD negative at the 10^-6 level The fixed-duration, modified regimen was highly active (ORR: 70.8%; VGPR+ rate: 58.3%). All MRD-evaluable CR+ patients were MRD negative at the 10^-6 level. Data cut-off: January 23, 2026. Evolution of response over time Evolution of response over time among responders (swimmer plot, up to 17 responders, duration of follow-up in months) Legend: Overall response = sCR, CR, VGPR, PR, MR, SD, PD, No response reported End of treatment or study status/reason: Death, Completed, AE, PD, Physician decision, Ongoing Responses deepened over time - Median time to first PR was 1.0 month, VGPR was 3.3 months and CR+ was 9.7 months 11 of 17 responders completed 13 cycles of treatment in PR+ - 7 in CR+ - 3 in VGPR, including 1 who subsequently achieved CR+ - 1 in PR 9 of 11 responders who completed 13 cycles in PR+ had responses >40 months, including all 7 CR+ patients The fixed-duration, modified regimen induced prolonged responses that continued after treatment completion. All 7 patients who completed 13 cycles in CR+ had responses >40 months. Data cut-off: January 23, 2026. Survival outcomes PFS (KM curve, All patients N=24): Patients remaining at risk: 24 21 15 15 13 13 12 12 10 10 10 10 10 10 10 10 10 10 9 5 3 1 NE (months 0-52) mPFS in the entire cohort, months (95% CI): 12.4 (2.8, NE) 36-month PFS rate, % (95% CI): 41.7 (21.9, 61.4) OS (KM curve, All patients N=24): Patients remaining at risk: 24 22 22 20 19 19 17 17 17 16 16 16 16 16 16 16 16 16 16 16 15 8 4 1 NE mOS in the entire cohort, months (95% CI): NR (17.3, NE) 36-month OS rate, % (95% CI): 66.7 (47.8, 85.5) Survival outcomes were promising, with a plateau observed in the probability of PFS and OS events in the absence of treatment Data cut-off: January 23, 2026.

Arm B randomized dose-expansion overview Patients: RRMM with ≥1 prior line including an IMiD and a PI; Pom-refractory patients excluded Objectives: Primary: Safety, RP2D; Secondary: Activity, pharmacokinetics, pharmacodynamics Dosing regimens: Patients were randomized 1:1 to the 70 mg or 105 mg cevostamab target-dose level at enrolment Cevostamab was initiated with double step-up dosing (0.3/3.3 mg) in a 14-day pre-phase or triple step-up dosing (0.3/1.2/3.6 mg) in a 21-day pre-phase* Schema across C1, C2-4, C5-6, C7+: 70 mg IV cevostamab: D1, D15 dosing in C1, C2-4, C5-6; D1 only in C7+ 105 mg IV cevostamab: D1, D15 dosing in C1, C2-4, C5-6; D1 only in C7+ 2 mg oral Pom: D1-21 continuous across cycles 20 mg IV/oral dex: D1, D8, D15, D22 across cycles *Patients were hospitalized for ≥48 hours after each cevostamab infusion in the pre-phase. Treatment was continued until disease progression or unacceptable toxicity. IV dex was given as premedication for all cevostamab doses in the pre-phase and C1. Optional from C5. Baseline characteristics by target-dose level n (%) unless stated | 70 mg (N=29) | 105 mg (N=25) Median age, years (range): 65 (40-76) | 65 (48-80) Male: 14 (48.3) | 14 (56.0) ECOG PS 0: 14 (48.3) | 14 (56.0); PS 1: 15 (51.7) | 11 (44.0) ISS stage I: 17 (60.7)* | 15 (62.5); II: 8 (28.6)* | 8 (33.3); III: 3 (10.7)* | 1 (4.2) HR cytogenetics, n (%) of patients with a conclusive assay result: 3/20 (15.0) | 8/22 (36.4) Extramedullary disease: 4 (13.8) | 3 (12.0) Median time from diagnosis, years (range): 6.6 (0.4-16.5) | 4.0 (1.1-26.0) n (%) unless stated | 70 mg (N=29) | 105 mg (N=25) Median number of prior lines of therapy, n (range): 2 (1-6) | 2 (1-6) Triple-class exposed: 21 (72.4) | 14 (56.0) Triple-class refractory: 9 (31.0) | 6 (24.0) Penta-drug refractory: 5 (17.9) | 3 (12.0) Refractory to last prior therapy: 21 (72.4) | 17 (68.0) The median number of prior lines of therapy was 2 Most patients were triple-class exposed and refractory to their last prior therapy Characteristics were comparable between groups, except for a higher rate of HR cytogenetics in the 105 mg cohort and triple-class exposure in the 70 mg cohort Data cut-off: June 12, 2026. Includes t(4;14), t(14;16) and del(17p). ≥1 IMiD, ≥1 PI and ≥1 anti-CD38 antibody. ≥2 IMiDs, ≥2 PIs and ≥1 anti-CD38 antibody. DOR and PFS by target-dose level DOR (KM curve, treatment group 70 mg N=29 vs 105 mg N=25, censored marks): Patients remaining at risk 70mg: 25 22 20 20 18 17 17 16 12 8 5 3 1 NE; 105mg: 22 20 20 17 16 15 12 9 7 3 1 NE 70 mg (n=25) | 105 mg (n=22) Median DOR, months (95% CI): NR (NE-NE) | NR (15.0-NE) 18-month DOR, % (95% CI): 74.0 (55.9-92.0) | 68.2 (48.7-87.6) PFS (KM curve, treatment group 70 mg N=29 vs 105 mg N=25): Patients remaining at risk 70mg: 29 28 25 23 21 19 18 17 13 9 6 3 1 NE; 105mg: 25 23 21 19 18 15 11 9 7 3 1 NE 70 mg (N=29) | 105 mg (N=25) Median PFS*, months (95% CI): NR (15.6-NE) | NR (16.9-NE) 18-month PFS, % (95% CI): 67.4 (49.8-84.9) | 66.0 (46.7-85.3) After a median follow-up of ~20 months, median PFS was not reached in both cohorts. 18-month DOR and PFS rates were generally comparable between dose groups. Data cut-off: June 12, 2026. *In entire cohort. Safety summary by target-dose level n (%) unless stated | 70 mg (N=29) | 105 mg (N=25) Median follow-up, months (range): 20.2 (12.7-27.5) | 20.4 (3.3-27.3) Any Gr AE: 29 (100) | 25 (100) Gr 3-4 AE: 24 (82.8) | 24 (96.0) Serious AE: 17 (58.6) | 15 (60.0) Gr 5 AE excluding PD: 1 (3.4)* | 1 (4.0) AE leading to discontinuation - Cevostamab: 2 (6.9) | 2 (8.0); Pom: 8 (27.6) | 8 (32.0); Dex: 0 | 2 (8.0) n (%) with AEs of interest by Grade (mirrored bar chart 70mg vs 105mg, Gr1-5): Neutropenia: 79.3% (70mg) | 84.0% (105mg) Anemia: 34.5% | 40.0% Thrombocytopenia: 24.1% | 32.0% CRS: 72.4% | 68.0% Infections: 75.9% | 80.0% Rash: 34.5% | 56.0% With the exception of rash, no substantial skin conditions, hair and nail changes or mucosal toxicities were observed Low ICANS incidence: Single cases of Gr 1 and Gr 2 only (n=1 each in 105 mg cohort) Overall rates of AEs were generally comparable between dose groups, with a lower incidence of Gr 3-4 events and rash in the 70 mg cohort Data cut-off: June 12, 2026. *Septic shock (treatment related). Intracranial haemorrhage (not treatment related) and weight increased (treatment related) (n=1 each). Clear cell renal cell carcinoma (not treatment related) and diffuse large B-cell lymphoma (treatment related) (n=1 each). Infective exacerbation of chronic obstructive airways disease (not treatment related) (n=1).

International Myeloma Society Arm B randomized dose-expansion overview Patients - RRMM with >=1 prior line including an IMiD and a PI - Pom-refractory patients excluded Objectives - Primary: Safety, RP2D - Secondary: Activity, pharmacokinetics, pharmacodynamics Dosing regimens - Patients were randomized 1:1 to the 70 mg or 105 mg cevostamab target-dose level at enrolment - Cevostamab was initiated with double step-up dosing (0.3/3.3 mg) in a 14-day pre-phase or triple step-up dosing (0.3/1.2/3.6 mg) in a 21-day pre-phase* [Schema] R 1:1 -> 70 mg IV cevostamab / 105 mg IV cevostamab (D1, D15 across cycles C1, C2-4, C5-6, C7+); 2 mg oral Pom D1-21 each cycle; 20 mg IV/oral dex [dosing days small print, not fully legible] *Patients were hospitalized for >=48 hours after each cevostamab infusion in the pre-phase. [footnote small print partially illegible] Treatment was continued until disease progression or unacceptable toxicity. IV dex was given as premedication for all cevostamab doses in the pre-phase and C1. Optional from C5. International Myeloma Society Response rates by target-dose level Response rates (%) by target-dose level — PR | VGPR | CR | sCR 70 mg (N=29): ORR: 86.2% (95% CI: 71.9-100); CR+: 62.1%; VGPR+: 75.9%; stacked bar: 44.8% (CR/sCR block) + 17.2% + 13.8% (VGPR) + 10.3% (PR) 105 mg (N=25): ORR: 88.0% (95% CI: 73.3-100); CR+: 64.0%; VGPR+: 76.0%; stacked bar: 44.0% + 20.0% + 12.0% (VGPR) + 12.0% (PR) Median time on study, months (range): 70 mg (N=29): 20.2 (12.7-27.5); 105 mg (N=25): 20.4 (3.3-27.3) Cevostamab plus pom-dex induced high ORR, VGPR+ and CR+ rates at both target doses Response rates were comparable between dose groups Data cut-off: June 12, 2026 [date partially legible] International Myeloma Society Safety summary by target-dose level [Left table] n (%) unless stated | 70 mg (N=29) | 105 mg (N=25) Median follow-up, months (range) | 20.2 (12.7-27.5) | 20.4 (3.3-27.3) Any Gr AE | 29 (100) | 25 (100) Gr 3-4 AE | 24 (82.8) | 24 (96.0) Serious AE | 17 (58.6) | 15 (60.0) Gr 5 AE excluding PD | 1 (3.4)* | 1 (4.0)† AE leading to discontinuation Cevostamab | 2 (6.9)‡ | 2 (8.0)§ Pom | 8 (27.6) | 8 (32.0) Dex | 0 | 2 (8.0) [Right chart] n (%) with AEs of interest by Grade — 70 mg vs 105 mg (Gr 1 / Gr 2 / Gr 3 / Gr 4 / Gr 5) Neutropenia¶ 79.3% | 84.0% Anaemia 34.5% | 40.0% Thrombocytopenia** 24.1% | 32.0% CRS 72.4% | 68.0% Infections†† 75.9% | 80.0% Rash‡‡ 34.5% | 56.0% Patients (%) • With the exception of rash, no substantial skin conditions, hair and nail changes or mucosal toxicities were observed • Low ICANS incidence: Single cases of Gr 1 and Gr 2 only (n=1 each in 105 mg cohort) Overall rates of AEs were generally comparable between dose groups, with a lower incidence of Gr 3-4 events and rash in the 70 mg cohort Data cut-off: June 12, 2025 [remainder of footnote largely illegible]: *Septic shock (treatment related). †Intracranial haemorrhage (not treatment related). ‡Clear cell renal cell carcinoma (not treatment related) and weight increased (treatment related) (n=1 each). §Diffuse large B-cell lymphoma (treatment related) and infective exacerbation of chronic obstructive airways disease (not treatment related) (n=1 each). ¶Group term neutropenia, neutrophil count decreased and febrile neutropenia. **Group term thrombocytopenia and platelet count decreased. ††MedDRA System Organ Class Infections and Infestations. ‡‡MedDRA high-level terms rashes, eruptions and exanthems NEC, exfoliative conditions, dermatitis ascribed to specific agent, acnes, dermatitis and eczema, and bullous conditions.

PHASE 1B DATA Cevostamab Plus Pom-Dex Yields Durable Responses in BCMA-Naive Myeloma Extended follow-up from the phase 1b CAMMA 1 trial (NCT04910568) supported selection of doses moving into phase 3 testing. OBJECTIVE RESPONSE RATE: 86.2%-88.0% (70 mg vs 105 mg cevostamab + pom-dex, n=29 vs n=25) COMPLETE RESPONSE OR BETTER: 62.1%-64.0% (Comparable across both dose levels) MRD NEGATIVITY IN CR+ PTS: 92.3%-93.8% (Among MRD-evaluable responders, 10^-5 threshold, NGS) 18-MONTH PFS RATE: 66.0%-67.4% (Median PFS not yet reached at ~20 months of follow-up) Source: Spencer A, et al. Presented at the 23rd International Myeloma Society (IMS) Annual Meeting; September 23-26, 2026; Glasgow, Scotland. OncLive / OncLive.com

23rd Annual MEETING & EXPOSITION - International Myeloma Society THURSDAY, SEPTEMBER 24, 2026 | 11:00-12:00 Abstract Session 3 Co-chairs: Meral Beksac and Martin Kortum - Double hit ultra-high risk myeloma treated with Isa-VRD plus cyclophosphamide induction and Isa-VRD consolidation: QoL analysis of the UKMRA RADAR trial in newly diagnosed transplant eligible patients - Karthik Ramasamy - Updated safety and efficacy of ramantamig (Ram; JNJ-5322) at the recommended phase 2 dose (RP2D), demonstrating feasibility of outpatient dosing (OPD) in relapsed/refractory multiple myeloma (RRMM) - Jeffrey Matous - DREAMM-9 Patient-Reported Vision-Related Function with Extended Belantamab Mafodotin Dosing Intervals with Bortezomib, Lenalidomide, Dexamethasone (BVRd) in Transplant Ineligible Multiple Myeloma - Hang Quach - A fixed-duration, modified cevostamab dosing regimen induces durable remissions that continue after completion in BCMA-naive pts with relapsed/refractory multiple myeloma (RRMM): CAMMA 1 Arm A results - Amrita Krishnan - Overall Survival, Progression and Non-Relapse Mortality With Teclistamab Plus Daratumumab (Tec-Dara) vs Dara-Based Triplets in Relapsed/Refractory Multiple Myeloma (RRMM): MajesTEC-3 Post Hoc Analysis - Luciano Costa Panelist photos shown: Dr. Beksac, Dr. Kortum, Dr. Matous, Dr. Ramasamy, Dr. Costa, Dr. Krishnan, Dr. Quach 23rd Annual MEETING & EXPOSITION - International Myeloma Society THURSDAY, SEPTEMBER 24, 2026 | 11:00-12:00 Abstract Session 4 Co-chairs: Lugui Qiu and Niels van de Donk - Augmenting TRAIL-dependent apoptosis enables BCMA CAR-T cells to overcome resistance in multiple myeloma - Mansour Poorebrahim - First-in-human phase Ia/Ib study of QLS32015 (GPRC5DxCD3) in patients (pts) with relapsed/refractory multiple myeloma (RRMM): updated safety and efficacy data - Jian Cui - An AI-Designed BCMA/FcRL5 Targeting TriTE Overcomes Antigen Escape and Outperforms BCMA-Directed Therapy in Multiple Myeloma - Jessica Encinas Mayoral - A CD4-Dominant, T-Helper-Skewed T Cell State Underlies Movement and Neurocognitive Toxicity After BCMA-Directed CAR-T Therapy in Multiple Myeloma - Theresia Akhlaghi - Single-cell antigen:TCR mapping reveals tumor-specific effector T-cell responses in multiple myeloma - David Avigan Panelist photos shown: Dr. Akhlaghi, Dr. Avigan, Dr. Cui, Dr. Encinas Mayoral, Dr. Poorebrahim, Dr. Qiu, Dr. van de Donk

Response rates by target-dose level Bar chart, y-axis "Patients (%)" 0-100: 70 mg (N=29): ORR 86.2% (95% CI: 71.9-100). Stacked: PR 10.3%, VGPR 13.8%, CR 17.2%, sCR 44.8%. Brackets: CR+: 62.1%, VGPR+: 75.9% 105 mg (N=25): ORR 88.0% (95% CI: 73.3-100). Stacked: PR 12.0%, VGPR 12.0%, CR 20.0%, sCR 44.0%. Brackets: CR+: 64.0%, VGPR+: 76.0% Right table: | 70 mg (N=29) | 105 mg (N=25) Median time on study, months (range) | 20.2 (12.7-27.5) | 20.4 (3.3-27.3) Right box: Cevostamab plus pom-dex induced high ORR, VGPR+ and CR+ rates at both target doses. Response rates were comparable between dose groups. Data cut-off: June 12, 2026. [no slide number visible] DOR and PFS by target-dose level Left panel "DOR": Kaplan-Meier curve, y-axis "Survival probability (%)" 0-100, x-axis "Time from start of treatment (months)" 0-28. Two lines: 70 mg (N=29, blue), 105 mg (N=25, orange), with censored marks (+). Legend: Treatment group 70 mg (N=29), 105 mg (N=25), Censored. "Patients remaining at risk" table: 70 mg: 25, 24, 23, 23, 22, 22, 21, 18, 15, 8, 6, 7, 5, 3, 1, NE (at months 0,2,4,6,8,10,12,14,16,18,20,22,24,26,28... approx) 105 mg: 22, 22, 20, 20, 18, 17, 17, 16, 12, 6, 3, 5, 3, 1, NE Right panel "PFS": Kaplan-Meier curve, same axes, two lines 70 mg (N=29, blue), 105 mg (N=25, orange), censored marks. "Patients remaining at risk" table: 70 mg: 29, 28, 25, 24, 23, 22, 21, 21, 18, 16, 10, 6, 4, 3, 1, NE 105 mg: 25, 25, 23, 21, 21, 19, 18, 16, 9, 6, 6, 4, 3, 1, NE Summary table: | 70 mg (n=25) | 105 mg (n=22) [DOR columns] Median DOR, months (95% CI) | NR (NE-NE) | NR (15.0-NE) 18-month DOR, % (95% CI) | 74.0 (55.9-92.0) | 68.2 (48.7-87.6) | 70 mg (N=29) | 105 mg (N=25) [PFS columns] Median PFS*, months (95% CI) | NR (15.6-NE) | NR (16.9-NE) 18-month PFS, % (95% CI) | 67.4 (49.8-84.9) | 66.0 (46.7-85.3) Bottom banner: After a median follow-up of ~20 months, median PFS was not reached in both cohorts. 18-month DOR and PFS rates were generally comparable between dose groups. Data cut-off: June 12, 2026. *In entire cohort. Safety summary by target-dose level Left table: n (%) unless stated | 70 mg (N=29) | 105 mg (N=25) Median follow-up, months (range) | 20.2 (12.7-27.5) | 20.4 (3.3-27.3) Any Gr AE | 29 (100) | 25 (100) Gr 3-4 AE [highlighted] | 24 (82.8) | 24 (96.0) Serious AE | 17 (58.6) | 15 (60.0) Gr 5 AE excluding PD | 1 (3.4)* | 1 (4.0)dagger AE leading to discontinuation: Cevostamab | 2 (6.9)Dagger | 2 (8.0)section Pom | 8 (27.6) | 8 (32.0) Dex | 0 | 2 (8.0) Right chart "n (%) with AEs of interest by Grade" - horizontal tornado/butterfly bar chart, 70 mg (left, 100-0%) vs 105 mg (right, 0-100%), graded Gr1-Gr5 (color scale): Neutropenia[para]: 79.3% (70mg) vs 84.0% (105mg) Anaemia: 34.5% vs 40.0% Thrombocytopenia**: 24.1% vs 32.0% CRS: 72.4% vs 68.0% Infections[dagger dagger]: 75.9% vs 80.0% Rash[Dagger Dagger]: 34.5% vs 56.0% Bullets: With the exception of rash, no substantial skin conditions, hair and nail changes or mucosal toxicities were observed. Low ICANS incidence: Single cases of Gr 1 and Gr 2 only (n=1 each in 105 mg cohort). Bottom banner: Overall rates of AEs were generally comparable between dose groups, with a lower incidence of Gr 3-4 events and rash in the 70 mg cohort. Footnote: Data cut-off: June 12, 2026. *Septic shock (treatment related). daggerIntracranial haemorrhage (not treatment related). DaggerClear cell renal cell carcinoma (not treatment related) and weight increased (treatment related) (n=1 each). sectionDiffuse large B-cell lymphoma (treatment related) and infective exacerbation of chronic obstructive airways disease (not treatment related) (n=1 each). paragraphGroup term neutropenia, neutrophil count decreased and febrile neutropenia. **Group term thrombocytopenia and platelet count decreased. dagger daggerMedDRA System Organ Class Infections and Infestations. Dagger DaggerMedDRA high-level terms rashes, eruptions and exanthems NEC, exfoliative conditions, dermatitis ascribed to specific agent, acnes, dermatitis and eczema, and bullous conditions. CRS by step-up dosing regimen Left chart "Double step-up dosing (N=28)": Any Gr CRS: 22 (78.6); Gr 2+ CRS: 6 (21.4); Gr 3+ CRS: 1 (3.6) Bar chart, y-axis "Patients (%)" 0-100, x-axis categories: D1 0.3mg (Pre-phase), D2-4 3.3mg (Pre-phase), D8 70/105mg (Pre-phase), D1 70/105mg (C1+). Grade 1/2/3 stacked bars (approx values: D1 0.3mg ~25% mostly Gr1; D2-4 3.3mg ~53% mixed Gr1/2; D8 70/105mg ~50% mixed Gr1/2; D1 70/105mg (C1+) ~28% mixed Gr1/2/3). Right chart "Triple step-up dosing (N=26)": Any Gr CRS: 16 (61.5); Gr 2+ CRS: 2 (7.7); Gr 3+ CRS: 0 Bar chart, same axes, categories: D1 0.3mg, D2-4 1.2mg, D8 3.6mg, D9-11 70/105mg (Pre-phase), D1 70/105mg (C1+). (approx values: D1 0.3mg ~20%; D2-4 1.2mg ~42%; D8 3.6mg ~11%; D9-11 70/105mg ~33%; D1 70/105mg (C1+) ~18%). Legend: Gr 1, Gr 2, Gr 3 (shaded purple scale) Right table "n (%) with CRS who received tocilizumab or corticosteroids for CRS treatment": | Double step-up dosing (N=28) | Triple step-up dosing (N=26) Tocilizumab | 10 (35.7) | 2 (7.7) Corticosteroids | 12 (42.9) | 1 (3.8) Tocilizumab and corticosteroids | 6 (21.4) | 1 (3.8) Bottom banner: CRS was less common, with <10% Gr 2+ events and no Gr 3 events, in the triple step-up cohort compared with the double step-up cohort. Tocilizumab use was <10% in the triple step-up cohort. Data cut-off: June 12, 2026.

[Slide 1 - Parkinsonism with Cilta-cel, mechanism: potential on-target, off-tumor effect on basal ganglia] Cross-trial table (CARTITUDE-1 / CARTITUDE-4 / RWE USMMIC / RWE USMMIC-2): Incidence: 6% / 0.5% / 2% / 3% (23/766). Clinical features: tremor/bradykinesia/rigidity/gait difficulty (CARTITUDE-1, USMMIC, USMMIC-2), tremor/bradykinesia/rigidity (CARTITUDE-4), USMMIC-2 adds personality change. Median time to onset: 27 days (14-108) / 85 days / 30 days (13-69) / 32 days (6-277). Sources: Martin et al. J Clin Oncol 41:1265-1274, 2023. San-Miguel et al. NEJM 389:335-347, 2023. Sidana et al. Blood. 2025;145(1):85-97. Sidana S et al, ASH 2025, manuscript in preparation 2026. Cohen et al Blood Cancer J. 2022;12:32. Hosoya et al Blood 2025. [Slide 2 - Parkinsonism: Treatment and Outcomes] N=23 patients with Parkinsonism. Resolved or improved: 55%. Resolved: 30% (7). Improved: 22% (5). Median time to resolution: 29 days (range 7-427). Patient-level treatment/outcome grid (23 patients) showing treatments received (corticosteroids, IVIG, anakinra, ruxolitinib, tocilizumab, cyclophosphamide, dasatinib, intrathecal chemotherapy, high-dose methotrexate, anti-parkinsonian medication, other supportive/symptom care, no treatment documented) mapped against Favorable/Unfavorable/Unknown outcome. Source: Sidana S et al. Manuscript in preparation, 2026. [Slide 3 - IEC-Colitis: My Management Approach] Diagnosis: Clinical and Pathological - GI infectious work up (infections can co-occur); Endoscopy + biopsies (duodenum: 80%; including normal mucosa); Assess lymphocytic infiltrate for clonality and CAR-T presence. First line: Early treatment initiation is key; High-dose Cytoxan (1-2 g/m2): preferred for fit patients. Other options: Cyclosporine (target trough 300-400 ng/mL); Ruxolitinib; Biologics for IBD: infliximab, vedolizumab, ustekinumab. Supportive Care: Often need parenteral nutrition; Aggressive antimicrobial prophylaxis; If on steroids for >5-7 days, antifungal prophylaxis. [Slide 4 - Cerebellar Toxicity with GPRC5D CAR-T and BsAb: Dizziness-Ataxia Syndrome] Potential low-level expression in inferior olivary nucleus of medulla oblongata. Cross-trial table (Arlo-cel N=84 / MonumenTAL-3 Tal-DP/Tal-D / MonumenTAL-1 Monotherapy): Any-grade 12% / 15%/12% / 5%. Grade 3: 7% / 3%/2% / 0.5%. Median time to onset: 30.5 days / 292/326 days / ~114 days. Recovery/resolution: "4 resolved, but recurred; all active" / 10.8%/16.9% / 54.2%. Incidence: 12-14%; Early for CAR-T, risk increases over time for BsAb. Diagnosis and Treatment: Diagnosis is clinical; Talquetamab: Hold drug, if rechallenged after resolution use a lower dose; Arlo-cel: Consider cytotoxic therapy and JAK inhibitors, ~Parkinsonism. Sources: Bal S et al. Blood. 2026;148:1240-1250. Mailankody et al. N Engl J Med 387:1196-1206. Mina R et al. N Engl J Med. 2026;395:671-683. Janakiram M et al. Blood Cancer J. 2025;15:135.

[Slide 1 - Incidence and severity of infections across different approved BsAbs and CAR T trials, cross-trial compilation table] BsAbs late-line relapse (triple-class exposed) - trial/drug/All grade/Grade 3-4/Grade 5: MajesTEC-1/Teclistamab: 78.8% / 55.2% / 12.7%. MagnetisMM-3/Elranatamab: 69.9% / 39.8% / 6.5%. LinkerMM-1/Linvoseltamab: 75% / 48% / NR. MonumenTAL-1/Talquetamab: 66.2% / 14.5% / 1.4%. RedirecTT-1/Tec-Tal: 89% / 64% / 12%. BsAbs early-line relapse (1-3 prior lines): MajesTEC-9/Teclistamab: 82.8% / 41.6% / 5.5%. MajesTEC-3/Tec-Dara: 96.5% / 41.6% / 5.5%. MonumenTAL-3/Tal-Dara: 83.0% / 29.2% / 1.5%. MonumenTAL-3/Tal-Dara-Pom: 84.3% / 37.7% / 0.7%. CAR T late-line relapse: CARTITUDE-1/Cilta-Cel: All grade 58%, Gr3-4 20%, Gr5 NR. KarMMa/Ide-Cel: 69%, 22%, NR (#, deaths from AEs not otherwise specified 7% total). CAR T early-line relapse: CARTITUDE-4/Cilta-Cel: 59%, 24%, 4%. KarMMa-3/Ide-Cel: 56%, 22%, NR (*, death from AEs 6% total). "Infections remain a major safety concern with T-cell-redirecting therapies, particularly with continuous bispecific antibody treatment." Table provided for ease of viewing, no comparisons intended. Sources: Moreau P et al NEJM 2022; Lesokhin AM et al Nat Med 2023;29(9):2259-2267; Lee HC et al Clin Lymphoma Myeloma Leuk 2026 Feb, 26(2):e201-e212,e8; Chari A et al NEJM 2022; Cohen Y et al Engl J Med 2025;392:138-49; Touzeau C et al N Engl J Med 2026;395:440-53; Costa LJ et al N Eng J Med 2026;394:739-52; Mina R et al N Engl J Med 2026;395:671-83; Einsele H et al, Lancet Oncol 2026;27:254-68; Ailawadhi S et al, Blood 2024 (144)23; Berdeja J et al, Lancet 2021;398:314-24; Munshi NC et al, N Engl J Med 2021 Feb 25;384(8):705-716. [Slide 2 - Pillars of infection prevention in patients receiving T-cell redirecting therapies] Table: Immune defect/risk -> Type of infection -> Preventive strategy. Lymphopenia/Neutropenia -> Bacterial infections/PJP, approx first 3-6 months -> Antiviral + PJP prophylaxis, G-CSF support. Immunosuppressive therapy (steroids, IL-6 blockage, cyclophosphamide) -> Bacterial infections/Sepsis/fungal infections/CMV -> Consider antibacterial/antifungal. B-cell aplasia/Hypogammaglobulinemia -> Bacterial/Viral (Respiratory virus, VZV/HSV)/CMV -> Immunoglobulin replacement, Vaccination. T-cell dysfunction -> Other opportunistic infections, Viral infections, fungal (in selected high risk patients) -> Fixed treatment duration/Schedule optimization. Sources: Rodriguez-Otero P et al, Lancet Oncology May 2024; O'Neill C, et al. eJHaem. 2023;4:811-822; Friedrich MJ, et al. Cancer Cell. 2023;41:711-725.e6; Frerichs KA et al, Blood Advances 2024; Mohan M et al, Blood Cancer Journal (2025) 15:74; Tix T et al, Molecular Therapy (2025)33:7; Richardson T et al, Blood Advances (2025) 9:6; Nath K et al, Blood Cancer Journal (2024) 14:88; Reynolds G et al, Blood advances (2023) 7:19; Banerjee R et al, Blood Advances (2025) 9:18; Dr Rodriguez Otero personal opinion.

Phase 3 evidence has moved inmune therapy into early relapse [Table: STUDY | COMPARISON | forest plot (FAVOURS EXPERIMENTAL / FAVOURS CONTROL) | PFS HR | OS HR] CARTITUDE-4 | Cilta-cel vs DPd/PVd | PFS HR 0.29 | OS HR 0.55 DREAMM-7 | BVd vs DVd | PFS HR 0.41 | OS HR 0.58 DREAMM-8 | BPd vs PVd | PFS HR 0.52 | OS HR 0.77 MajesTEC-3 | Tec-Dara vs DPd/DVd | PFS HR 0.17 | OS HR 0.46 MonumenTAL-3 | Tal-DP vs DPd | PFS HR 0.28 | OS HR 0.47 (MonumenTAL-3) | Tal-D vs DPd | PFS HR 0.33 | OS HR 0.51 MajesTEC-9 | Teclistamab vs PVd/Kd | PFS HR 0.29 | OS HR 0.60 MonumenTAL-6 | Tal-Tec vs EPd/PVd | PFS HR 0.11 | OS HR NR (MonumenTAL-6) | Tal-Pom vs EPd/PVd | PFS HR 0.27 | OS HR NR CERVINO | Etentamig vs investigator's choice | PFS HR 0.40 | OS HR NR RANDOMIZED PHASE 3 TRIALS WITHOUT PUBLIC NUMERICAL HRs MagnetisMM-5 | Elranatamab +/- Dara vs DPd | Positive top-line result; no numerical data reported [Forest plot x-axis tick labels largely illegible]

FIRST RELAPSE Why CAR T belongs earlier Maximize early kill, then return time to the patient 48% MRD-negative by day 56 62% vs 18% overall MRD negativity at 10⁻⁵ Deepest early cytoreduction ENGINEERED IMMUNE RESET Lymphodepletion reduces suppressive immune cells; one infusion delivers a large, activated population that expands in vivo. A fundamentally different biologic attack than continuous antibody exposure +11.6 point QoL advantage at month 12 Global health status: +10.1 with cilta-cel vs −1.5 with SOC Treatment-free runway, less ongoing supportive care, and a PFS2/OS advantage SURVIVORSHIP IS PART OF THE STRATEGY Track immune reconstitution, IVIG and infection needs, revaccination, late effects and return to life. CARTITUDE-4; Mina et al, Lancet Haematol 2025; Popat et al, Blood 2024; updated OS analysis, Lancet 2026 Emerging CAR T platforms show deep early responses Latest publicly reported datasets; populations and follow-up differ Dataset | Anito-cel iMMagine-1 | Arlo-cel early-line phase 1 | AZD0120 DURGA-1 | KLN-1010 inMMyCAR Platform / target | Ex vivo BCMA | Ex vivo GPRC5D | Ex vivo BCMA + CD19 | In vivo BCMA Patients | 117 evaluable | 24 evaluable, 31 treated | 23 evaluable | 18 dosed Prior lines | Median 5 | Median 2, 1-3 allowed | Median 4 | Not reported Follow-up | Median 12.6 mo | Median 18.3 mo | Median 3.9 mo | 6 pts ≥4 mo, first pt >10 mo ORR | 97% | 96% | 96% | 100% all evaluable CR / sCR | 68% | 67% | 78.3% | 4 sCR / 6 pts with ≥4 mo MRD negative at 10⁻⁵ | 93% (70/75 evaluable) | 56.3% MRD-negative CR | 94% (16/17 evaluable) | 100% all evaluable 12-month PFS | 79% | 74% | Not reported | Not reported Deep responses now extend from ex vivo platforms to first-in-human in vivo CAR T Patel K, ASH 2025 #256. Bal S, Haematologica 2026. Richard S, Blood 2025 #269. Ho J, ASCO 2026. Cross-trial comparisons are descriptive. MRD denominators and follow-up maturity differ. Take-home messages • CAR T belongs at first relapse for eligible patients, before disease biology and treatment burden narrow the opportunity. • Earlier-line cilta-cel delivers deep real-world responses, with 79% progression-free survival at 12 months. • Earlier referral preserves more active holding-therapy options and helps protect T-cell fitness before collection. • Sequence intentionally: use holding therapy while access is secured, collect early, then bridge to control disease before infusion. • The platform is expanding beyond conventional BCMA CAR T to GPRC5D, dual-target, and in vivo approaches. Refer early, collect early, and preserve the path to CAR T

International Myeloma Society CARTITUDE-4 subgroup analysis: High risk vs Standard risk [PFS Kaplan-Meier plot] Patients progression free and alive, % vs Progression-free survival, months (0-45). Curves: High risk, Cilta-cel; Standard risk, Cilta-cel; High risk, SOC; Standard risk, SOC. Patients at risk rows shown [small print, not fully legible]. [OS Kaplan-Meier plot] Patients alive, % vs Overall survival, months (0-45). Curves: High risk, Cilta-cel; Standard risk, Cilta-cel; High risk, SOC; Standard risk, SOC. Patients at risk rows shown [small print, not fully legible]. PFS table: Cilta-cel SOC High risk, months (95% CI) 37.1 (26.7-NE) 10.3 (7.6-12.6) Standard risk, months (95% CI) NR (NE-NE) 20.6 (11.2-33.6) OS table: Cilta-cel SOC High risk, months (95% CI) NR (NE-NE) 38.0 (34.0-NE) Standard risk, months (95% CI) NR (NE-NE) NR (34.7-NE) The median follow-up was 33.6 months (range, 0.1-45.0). NR, not reached; SOC, standard of care. Sidana S, et al. ASCO 2025 International Myeloma Society RWE: Cilta-cel outcomes in earlier versus later lines German nationwide registry of standard-of-care cilta-cel, 2022-2025 OUTCOME EARLIER LINE (1-3 prior lines, n = 177) LATER LINE (>3 prior lines, n = 429) 12-month PFS 79% (95% CI, 71-87%) 70% (95% CI, 65-76%) Overall response rate 91% 88% Complete response 63% 54% Non-ICANS neurotoxicity 3% 8% Maintained CR was associated with 100% PFS at 12 months in both cohorts Gagelmann N, et al. J Hematol Oncol. 2026;19:46. doi:10.1186/s13045-026-01806-6 Nonrandomized registry analysis. The investigators did not perform formal comparisons between cohorts. International Myeloma Society CARTITUDE-4 TME: Patients With 1 or 2 pLOT vs 3 pLOT had a More Immunocompetent TME at Baseline - Patients with 1 or 2 vs 3 pLOT had a more immunocompetent TME at baseline, as suggested by higher levels of costimulatory molecules, CD4+ T-cell antigen-presenting machinery (MHCII), and TCRA pathway - Higher levels of these gene signatures in TME at baseline trended with longer PFS (data not shown) Earlier use of cilta-cel may support longer PFS by leveraging a more immunocompetent TME at baseline Baseline TME gene set enrichment scores grouped by pLOT [Box plots: Coactivation molecules (P<0.05, P<0.05, ns), Costimulatory receptors (P.adj<0.05, P<0.05, ns), MHCII pathway (P<0.05, ns, ns), TCRA pathway (P<0.05, P<0.05, ns); Enrichment score -1.0 to 0.8; groups 1 pLOT (n=49), 2 pLOT (n=57), 3 pLOT (n=36)] Cilta-cel, ciltacabtagene autoleucel; MHC, major histocompatibility complex; pLOT, prior line of therapy; TCRA, T-cell receptor activation; TME, tumor microenvironment. Presented by S Parekh at the 67th American Society of Hematology (ASH) Annual Meeting; December 6-9, 2025; Orlando, FL, USA

[Slide 1 - IBEX study schema, NCT06107738, Karmanos Cancer Institute + Indiana University + Cedars Sinai + University of Michigan] n=42 MRD-evaluable patients (enrolled 46). Dara-containing induction, 90-150 days after ASCT, PR+ any MRD status. Iberdomide 0.75mg/d x21d (28-day cycles, no escalation) + SC Dara 1800mg QW(C1-2), Q2W(C3-6), Q4W(C7+), for 24 months (max 26 cycles) or progressive disease. Primary endpoint: sustained MRD(-) using 10^-5 cutoff, ClonoSEQ or EuroFlow @ 6,12,18,24 months. Key secondary: conversion MRD(+) to MRD(-), IMWG response category improvement rate, PFS/OS at 12/24 months, QoL/PRO (EORTC-QLQ-C30, EQ-5D-5L), safety. Exploratory: serial MRD-FIX, immune profiling. [Slide 2 - Safety: Kinetics of Neutropenia] Per-patient grid tracking neutropenia grade (1/2/3) by cycle (baseline-C6) at both 0.75mg and 1.0mg dose levels, with IBER dose reduction markers (n=3, 14%), dose increase (n=1, 5%), and G-CSF added (n=2, 10%). 2 additional patients had IBER dose reductions after cycle 6. [Slide 3 - Efficacy] Baseline to on-study response shift (Unknown/CR/VGPR/PR/PD categories) showing improvement, with 5/21 improved IMWG response category. Serial MRD measurements (n=13) tracked by patient (KCI-02 through KCI-17) at Baseline/6mo/12mo/18mo/24mo, color-coded by MRD depth (<10^-4 red, >=10^-4-<10^-3 orange, >=10^-3-<10^-6 light green, >=10^-6 dark green). 8/11 baseline MRD(+) patients became MRD(-) (>=10^-6 threshold implied); 3/9 patients achieved sustained MRD(-) status (marked with asterisk: KCI-02, KCI-06, KCI-09). [Slide 4 - Conclusions] IBER/DARA maintenance post-ASCT is generally well-tolerated: dose reductions for hematologic toxicity uncommon (15%); neutropenia most pronounced during C1-2 (weekly DARA dosing); infections common but most Gr1-2 (only 1 hospitalization); diarrhea less common, less severe than typical for LEN maintenance. Improvements in IMWG response and MRD status observed: sustained MRD(-) in 3/9 pts (33%) with >=1yr follow-up; 6/9 pts (67%) with >=1yr F/U currently MRD(-) (all MRD(+) at start). Enrollment is ongoing (27 patients). NCT06107738.

DON'T GET LOST AT IMS 2026 GLASGOW - SEPTEMBER 23-26 THE MUST-SEE ROADMAP PEOPLE SCIENCE PROGRESS TOGETHER | BIGGER SCIENCE A BRIGHTER TOMORROW | GLASGOW WELCOMES A BRIGHTER TOMORROW THURSDAY - SEPT 24 5:00-6:00 PM - HALL 5 MRD & MAINTENANCE (MRD: DEEPER ANSWERS BRIGHTER TOMORROWS) - MIDAS: Isa-iberdomide - CASSIOPEIA: 5-year sustained MRD - GEM2014MAIN: MRD-guided stopping - IBEX: Iber-Dara post-ASCT FRIDAY - SEPT 25 3:00-5:00 PM - HALL 5 DO NOT MISS THE PLENARY - DO NOT MISS (BOLD SCIENCE CHANGES LIVES) - CERVINO: Etentamig vs standard therapy - PERSEUS: Does long-term PFS become OS? - EXCALIBER-RRMM: IberDd vs DVd - First-in-human CD19/BCMA in-vivo CAR-T SAME QUESTIONS. A BRIGHTER TOMORROW. SATURDAY - SEPT 26 8:30-9:30 AM - HALL 5 THE NEXT CELLULAR-THERAPY ERA (CELLULAR THERAPIES: REAL POTENTIAL FOR REAL PEOPLE) - Updated CEPHEUS MRD - BCMAxGPRC5D dual-target CAR-T - IASO206: IV in-vivo BCMA CAR-T 11:00 AM-12:00 PM - HALL 5 - Post-CAR-T elranatamab - Delayed cilta-cel neurotoxicity - MRD-guided T-cell redirection 11:00 AM-12:00 PM - LOMOND - CAR-T enterocolitis biology 1:00-2:00 PM - LOMOND - LV169: lentiviral in-vivo BCMA CAR delivery Same goal. Brighter lives. WHAT TO WATCH (DIFFERENT PERSPECTIVES A CLEARER FUTURE) - Durability - OS - sustained MRD - infection burden - Access - treatment-free time - late toxicity - In-vivo CAR-T: insertion sites, off-target transduction, vector shedding & persistence Go for the randomized data and the biology--not every new response rate. Single malt only AFTER the final session. SAME CURIOSITY FURTHER TOGETHER | GREAT SCIENCE DESERVES A DRAM

Study schema (NCT06645678) Phase I/II RRMM patients with >=2 prior lines, including 1 PI and lenalidomide, ECOG 0-2 Part 1 (safety, boxed): Elranatamab 76mg weekly* + Mezigdomide TBD, q4 weeks x24 cycles (N=12) Part 2 (expansion): Elranatamab + Mezigdomide q4 weeks x24 cycles (N=63) Maintenance: Mezigdomide until progression *Weekly up to C6 -> if PR or better response after C6 then Q2W -> if CR or better response after C12 then Q4W Dosing timeline: Elranatamab 2 step-up dosing D1(12mg), D4(32mg), D8(76mg) SC 76mg C1+; D1,D4,D8 then D1,D8,D15,D22 (C1+ QW schedule) then D1,D15 (C7+ Q2W schedule if PR or better) through EOT Mezigdomide: QD D1-21, Reset D22-D28, per RP2D, all cycles starting with C1D1, through EOT Side notes: 3+3 dose escalation Phase 1b, starting mezigdomide dose at 0.3mg -> 0.6mg -> 1mg; Prior anti-BCMA exposure not allowed; Primary prophylaxis with peg-GCSF; Bactrim, acyclovir prophylaxis mandatory; IVIG use according to IgG level 23rd International Myeloma Society Annual Meeting, slide 4 Response rates Data cutoff (2026-Jul-31); Median FU duration from C0D1 13.7M (range 8.6-19.3M); Median time to response 17 days (11-45d), to best response 5.6M (3.3-5.9M) Per mezigdomide dose (VGPR/sCR stacked bars, % of patients): Total (13/13): 100% - VGPR 7.7%, sCR 92.3% (>=CR: 92.3%) MEZI 0.3mg (7/7): 100% - VGPR 14.3%, sCR 85.7% MEZI 0.6mg (6/6): 100% - VGPR 0%, sCR 100.0% In subgroups (VGPR/sCR stacked bars): Prior TCE - Yes (4/4): 100%, VGPR 0, sCR 100.0; No (9/9): 100%, VGPR 11.1, sCR 88.9 Triple class refractory - Yes (7/7): 100%, VGPR 0, sCR 100.0; No (6/6): 100%, VGPR 16.7, sCR 83.3 Penta-refractory - Yes (6/6): 100%, VGPR 0, sCR 100.0; No (7/7): 100%, VGPR 14.3, sCR 85.7 Soft tissue EMD - Yes (3/3): 100%, VGPR 0, sCR 100; No (10/10): 100%, VGPR 10, sCR 90 *NGS [LymphoTrack, Invivoscribe] MRD data available in 11/13 patients: 10^-5 MRD negative CR in 10/11 (90.9%) 23rd International Myeloma Society Annual Meeting, slide 7 Baseline characteristics Characteristics | Total (N=13) | Elra + Mezi 0.3mg (N=7) | Elra + Mezi 0.6mg (N=6) Median age, years (range) | 69 (51-76) | 69 (54-74) | 70 (51-76) Male, n (%) | 9 (69.2) | 6 (85.7) | 3 (50.0) Soft tissue EMD, n (%) | 3 (23.1) | 2 (28.6) | 1 (16.7) High cytogenetic risk, n (%) | 5 (38.5) | 2 (28.6) | 3 (50.0) R-ISS, I/II/III, n (%) | 7(53.8)/6(46.2)/0 | 4(57.1)/3(42.9)/0 | 3(50.0)/3(50.0)/0 ECOG PS, 0/1/2, n (%) | 4(30.8)/8(61.5)/1(7.7) | 3(42.9)/4(57.1)/0 | 1(16.7)/4(66.7)/1(16.7) IMWG frailty score, frail, n(%) | 2 (15.4) | 0 | 2 (33.3) Prior lines of tx, median (range) | 4 (2-7) | 4 (2-6) | 4 (2-7) Prior ASCT | 8 (61.5) | 5 (71.4) | 3 (50.0) Prior anti-CD38 mAb exposure | 7 (53.8) | 3 (42.9) | 4 (66.7) Prior TCE exposure | 4 (30.8) | 2 (28.6) | 2 (33.3) GPRC5D/CD3 | 3 | 1 | 2 FcRH5/CD3 | 1 | 1 | 0 Abbreviations: EMD, extramedullary disease; R-ISS, revised International Staging System; ECOG PS, Eastern Cooperative Oncology Group Performance Status; IMWG, International Myeloma Working Group; tx, treatment; ASCT, autologous stem cell transplantation; mAb, monoclonal antibody; TCE, T-cell engager 23rd International Myeloma Society Annual Meeting, slide 6 Treatment related adverse events Data cutoff (2026-Jul-31) Event, n (%) | Total (N=13) Any grade/Gr3-4 | Elra+Mezi 0.3mg (N=7) Any grade/Gr3-4 | Elra+Mezi 0.6mg (N=6) Any grade/Gr3-4 Any AE | 13(100)/13(100) | 7(100)/7(100) | 6(100)/6(100) Hematologic: Thrombocytopenia | 7(53.8)/7(53.8) | 3(42.9)/3(42.9) | 4(66.7)/4(66.7) Neutropenia | 11(84.6)/11(84.6) | 6(85.7)/6(85.7) | 5(83.3)/5(83.3) Febrile neutropenia | 0/0 | 0/0 | 0/0 Non-hematologic: CRS | 10(76.9)/0 | 5(71.4)/0 | 5(83.3)/0 ICANS | 0/0 | 0/0 | 0/0 Fatigue/weakness | 11(84.6)/1(7.7) | 5(71.4)/1(14.3) | 6(100)/0 Decreased appetite | 7(53.8)/0 | 4(57.1)/0 | 3(50.0)/0 Infection | 6(46.2)/4(30.8) | 4(57.1)/3(42.9) | 2(33.3)/1(16.7) Nausea | 5(38.5)/0 | 4(57.1)/0 | 1(16.7)/0 Liver enzyme elevation | 5(38.5)/3(23.1) | 1(14.3)/0 | 4(66.7)/3(50.0) Pyrexia | 5(38.5)/0 | 2(28.6)/0 | 3(50.0)/0 Skin rash | 4(30.8)/0 | 1(14.3)/0 | 3(50)/0 Arthralgia/bone pain | 4(30.8)/0 | 2(28.6)/0 | 2(33.3)/0 23rd International Myeloma Society Annual Meeting, slide 9

International Myeloma Society New 2nd line treatments according to patient characteristics Columns: Tec +Ig (MajesTec 9) | Tec Dara +Ig (MajesTec 3) | Tal Dara (MonumenTal 3) | Tal Dara Pom (MonumenTal 3) | Mezi Kd (Successor 2) | [Press release only:] Tal Pom (MonumenTal 6) | Tec Tal +Ig (MonumenTal 6) | Elra +Ig (MagnetisMM 5) CD38 refractory: Yes | Possibly* | Possibly* | Possibly* | Yes | Yes | Yes | ? High risk: Possibly | Yes | Possibly | Possibly | Yes | Yes | Yes | ? EMD: Possibly | Possibly | Possibly | Possibly | Possibly | Possibly | Yes | ? Frail: Yes | Caution | Possibly | Caution | Yes | Possibly | Yes | ? High Infection risk: Possibly | Caution | Yes | Caution | Possibly | Possibly | Caution | ? Tolerability: Yes | Yes | Possibly | Possibly | Yes | Possibly | Caution | ? [Note: row-to-cell alignment approximate due to photo angle] Yes = Preferred / data supports; Possibly = Conditional / Limited data; Caution = possible but caution required Tec = Teclistamab | Tal = Talquetamab | Elra = Elranatamab | Dara = Daratumumab | Pom = Pomalidomide | Ig = Immunoglobulin | Mezi = mezigdomide | K = carilzomib | *based on TRIMM-2 data Based on trial eligibility + published efficacy/safety signals for BCMA & GPRC5D bispecifics. Table represents personal opinion and based on limited evidence! Costa LJ, et al. N Engl J Med. 2026;394(8):739-752, Touzeau C, et al. N Engl J Med. 2026, Mina R, et al. N Engl J Med. 2026, Dimopoulos et al., The Lancet 2026, Johnson & Johnson. press release, 23 July 2026, Pfizer. press release, 29 April 2026.

International Myeloma Society Response rates Data cutoff (2026-Jul-31) Median FU duration from C0D1 13.7 M (range 8.6-19.3M) Median time to response 17 days (11-45d), to best response 5.6M (3.3-5.9M) Per mezigdomide dose (Patients (%), VGPR / >=CR stacked bars): Total: 100% (13/13) — VGPR 7.7, >=CR 92.3 (bracket: >=CR: 92.3%) MEZI 0.3mg: 100% (7/7) — VGPR 14.3, >=CR 85.7 MEZI 0.6mg: 100% (6/6) — >=CR 100.0 In subgroups (all bars 100%): Prior TCE — Yes 100% (4/4) [bar split illegible]; No 100% (9/9) — 88.9 / 11.1 Triple class refractory — Yes 100% (7/7) [bar split illegible]; No 100% (6/6) — 83.3 / 16.7 Penta-refractory — Yes 100% (6/6) [bar split illegible]; No 100% (7/7) — 85.7 / 14.3 Soft tissue EMD — Yes 100% (3/3) — 100; No 100% (10/10) — 90 / 10 Legend: VGPR / >=CR [wording partially illegible] *NGS (LymphoTrack, Invivoscribe) MRD data available in 11/13 patients : 10^-5 MRD negative CR in 10/11 (90.9%) 23rd International Myeloma Society Annual Meeting — 7 International Myeloma Society Treatment related adverse events Data cutoff (2026-Jul-31) Event, n (%) | Total (N=13) Any grade / Gr 3/4 | Elra + Mezi 0.3mg (N=7) Any grade / Gr 3/4 | Elra + Mezi 0.6mg (N=6) Any grade / Gr 3/4 Any AE | 13 (100) / 13 (100) | 7 (100) / 7 (100) | 6 (100) / 6 (100) Hematologic Thrombocytopenia | 7 (53.8) / 7 (53.8) | 3 (42.9) / 3 (42.9) | 4 (66.7) / 4 (66.7) Neutropenia | 11 (84.6) / 11 (84.6) | 6 (85.7) / 6 (85.7) | 5 (83.3) / 5 (83.3) Febrile neutropenia | 0 / 0 | 0 / 0 | 0 / 0 Non-hematologic CRS | 10 (76.9) / 0 | 5 (71.4) / 0 | 5 (83.3) / 0 ICANS | 0 / 0 | 0 / 0 | 0 / 0 Fatigue/ weakness | 11 (84.6) / 1 (7.7) | 5 (71.4) / 1 (14.3) | 6 (100) / 0 Decreased appetite | 7 (53.8) / 0 | 4 (57.1) / 0 | 3 (50.0) / 0 Infection | 6 (46.2) / 4 (30.8) | 4 (57.1) / 3 (42.9) | 2 (33.3) / 1 (16.7) Nausea | 5 (38.5) / 0 | 4 (57.1) / 0 | 1 (16.7) / 0 Liver enzyme elevation | 5 (38.5) / 3 (23.1) | 1 (14.3) / 0 | 4 (66.7) / 3 (50.0) Pyrexia | 5 (38.5) / 0 | 2 (28.6) / 0 | 3 (50.0) / 0 Skin rash | 4 (30.8) / 0 | 1 (14.3) / 0 | 3 (50) / 0 Arrthralgia /bone pain | 4 (30.8) / 0 | 2 (28.6) / 0 | 2 (33.3) / 0 23rd International Myeloma Society Annual Meeting — 9 International Myeloma Society Complementary activity of ELRA + MEZI #1: Mezigdomide augments and sustains elranatamab-driven T-cell activation and effector function (patient sample analyses, N=13) Panel a: CD4+ : CD8+ ratio (% of T cells / % of T cells) across study visits C0D1, C0D8, C1D1, C1D8, C1D15, C3D1, CR (thin lines, individual patients) Panel b: CD8+ memory subset (% of CD8+ T cells) — EM, TEMRA, CM curves across study visits Panel c: CD4+ memory subset (% of CD4+ T cells) — EM, CM, Naive, TEMRA curves across study visits Panel d: Ki67+ cells (% of parent lineage) — up 3.68-fold, q=0.007 (red arrows at C0D8 and C1D8) Panel e: CD57+ KLRG1+ cells (% of parent lineage) — down 0.45-fold, q=0.007 (CD8, CD4) Panel f: Marker+ cells (% of CD8+ T cells) — HLA-DR, CD38, ICOS (ICOS up 1.70-fold, q=0.008), EOMES Panel g: IFN-gamma (pg/mL), C1D1 to C1D8 — [13.48-fold, 95% CI [illegible], q=0.013] [partially illegible] Panel h: Fold change C1D1->C1D8 by soluble analyte — all pairs (n=13), plasma->plasma (n=6) [partially illegible]; [4.3-fold, p=0.007] [partially illegible] 23rd International Myeloma Society Annual Meeting — 12

Response rates Data cutoff (2026-Jul-31) Median FU duration from C0D1 13.7 M (range 8.6-19.3M) Median time to response 17 days (11-45d), to best response 5.6M (3.3-5.9M) Left bar chart "Per mezigdomide dose" (VGPR/sCR stacked, % Patients): Total: 100% (13/13) - VGPR 7.7, sCR 92.3 (bracket: >=CR: 92.3%) MEZI 0.3mg: 100% (7/7) - VGPR 14.3, sCR 85.7 MEZI 0.6mg: 100% (6/6) - VGPR 0, sCR 100.0 Right bar chart "In subgroups" (VGPR/sCR stacked, % Patients), by prior TCE / triple class refractory / penta-refractory / soft tissue EMD, Yes vs No: Prior TCE: Yes 100% (4/4) - VGPR 0, sCR 100.0 | No 100% (9/9) - VGPR 11.1, sCR 88.9 Triple class refractory: Yes 100% (7/7) - VGPR 0, sCR 100.0 | No 100% (6/6) - VGPR 16.7, sCR 83.3 Penta-refractory: Yes 100% (6/6) - VGPR 0, sCR 100.0 | No 100% (7/7) - VGPR 14.3, sCR 85.7 Soft tissue EMD: Yes 100% (3/3) - VGPR 0, sCR 100 | No 100% (10/10) - VGPR 10, sCR 90 Footnote: *NGS [LymphoTrack, Invivoscribe] MRD data available in 11/13 patients: 10^-5 MRD negative CR in 10/11 (90.9%) International Myeloma Society - 23rd International Myeloma Society Annual Meeting [Slide 7] Duration of response Data cutoff (2026-Jul-31) 10/13 patients remain on study: 1 discontinued d/t patient's decision; 2 d/t physician's decision Swimmer plot, y-axis: Mezi dose | Patient (SNU-S01 through SNU-S15, 13 rows), x-axis: Months from first infusion (COD1) 0-21. Rows (dose, patient, status): 0.3, SNU-S01, blue bar ending ~9mo with Death (X) marker, best response sCR (star) 0.3, SNU-S03, blue bar extending to ~19mo, ongoing (arrow), best response sCR 0.3, SNU-S02, blue bar extending to ~19mo, ongoing (arrow), best response sCR 0.3, SNU-S04, blue bar to ~10mo (End of treatment tick) then gray "survival follow-up after EOT" to ~17.5mo, ongoing (arrow) 0.6, SNU-S05, orange bar extending to ~14mo, ongoing (arrow), sCR 0.6, SNU-S06, orange bar extending to ~15.5mo, ongoing (arrow), sCR 0.6, SNU-S07, orange bar extending to ~15.5mo, ongoing (arrow), sCR 0.3, SNU-S09, blue bar to ~11mo then gray survival follow-up to ~12.5mo, Death (X) marker, sCR 0.3, SNU-S10, blue bar extending to ~14mo, ongoing (arrow), sCR 0.3, SNU-S11, blue bar extending to ~12.5mo, ongoing (arrow), sCR 0.6, SNU-S12, orange bar to ~9.5mo, ongoing (arrow), sCR 0.6, SNU-S14, orange bar to ~7.5mo, ongoing (arrow), sCR 0.6, SNU-S15, orange bar to ~9mo, ongoing (arrow), sCR Each row also marks First response (PR, triangle) and VGPR (diamond) and Best response (star). Legend: On study treatment - 0.3 mg (blue), On study treatment - 0.6 mg (orange), Survival follow-up after EOT (gray), VGPR (diamond), First response (PR, triangle), Best response (star), End of treatment (vertical tick), Follow-up ongoing (arrow), Death (X) International Myeloma Society Annual Meeting [Slide 8] Treatment related adverse events Data cutoff (2026-Jul-31) Table - Event, n (%) | Total (N=13): Any grade / Gr 3/4 | Elra + Mezi 0.3mg (N=7): Any grade / Gr 3/4 | Elra + Mezi 0.6mg (N=6): Any grade / Gr 3/4 Any AE | 13 (100) / 13 (100) | 7 (100) / 7 (100) | 6 (100) / 6 (100) Hematologic: Thrombocytopenia | 7 (53.8) / 7 (53.8) | 3 (42.9) / 3 (42.9) | 4 (66.7) / 4 (66.7) Neutropenia | 11 (84.6) / 11 (84.6) | 6 (85.7) / 6 (85.7) | 5 (83.3) / 5 (83.3) Febrile neutropenia | 0 / 0 | 0 / 0 | 0 / 0 Non-hematologic: CRS | 10 (76.9) / 0 | 5 (71.4) / 0 | 5 (83.3) / 0 ICANS | 0 / 0 | 0 / 0 | 0 / 0 Fatigue/weakness | 11 (84.6) / 1 (7.7) | 5 (71.4) / 1 (14.3) | 6 (100) / 0 Decreased appetite | 7 (53.8) / 0 | 4 (57.1) / 0 | 3 (50.0) / 0 Infection | 6 (46.2) / 4 (30.8) | 4 (57.1) / 3 (42.9) | 2 (33.3) / 1 (16.7) Nausea | 5 (38.5) / 0 | 4 (57.1) / 0 | 1 (16.7) / 0 Liver enzyme elevation | 5 (38.5) / 3 (23.1) | 1 (14.3) / 0 | 4 (66.7) / 3 (50.0) Pyrexia | 5 (38.5) / 0 | 2 (28.6) / 0 | 3 (50.0) / 0 Skin rash | 4 (30.8) / 0 | 1 (14.3) / 0 | 3 (50) / 0 Arrthralgia /bone pain | 4 (30.8) / 0 | 2 (28.6) / 0 | 2 (33.3) / 0 International Myeloma Society Annual Meeting [Slide 9] Infection Data cutoff (2026-Jul-31) Table - Parameter, n (%) | Total (N=13) | Elra + Mezi 0.3mg (N=7) | Elra + Mezi 0.6mg (N=6) Patients with infection | 6 (46.2) | 4 (57.1) | 2 (33.3) Upper respiratory tract | 3 (23.1) | 1 (14.3) | 2 (33.3) Lower respiratory tract | 4 (30.8) | 3 (42.9) | 1 (16.7) Gastrointestinal tract | 1 (7.7) | 1 (14.3) | 0 Ocular | 3 (23.1) | 2 (28.6) | 1 (16.7) Patient with viral infections CMV disease (retinitis) | 2 (15.4) | 2 (28.6) | 0 Respiratory viruses (influenza A) | 1 (7.7) | 1 (14.3) | 0 Patients with fungal infections | 1 (7.7) | 1 (14.3) | 0 International Myeloma Society Annual Meeting [Slide 10]

DON'T GET LOST AT IMS 2026 GLASGOW - SEPTEMBER 23-26 THE MUST-SEE ROADMAP PEOPLE SCIENCE PROGRESS TOGETHER | BIGGER SCIENCE A BRIGHTER TOMORROW | GLASGOW WELCOMES A BRIGHTER TOMORROW THURSDAY - SEPT 24 5:00-6:00 PM - HALL 5 MRD & MAINTENANCE (MRD: DEEPER ANSWERS BRIGHTER TOMORROWS) - MIDAS: Isa-iberdomide - CASSIOPEIA: 5-year sustained MRD - GEM2014MAIN: MRD-guided stopping - IBEX: Iber-Dara post-ASCT FRIDAY - SEPT 25 3:00-5:00 PM - HALL 5 DO NOT MISS THE PLENARY - DO NOT MISS (BOLD SCIENCE CHANGES LIVES) - CERVINO: Etentamig vs standard therapy - PERSEUS: Does long-term PFS become OS? - EXCALIBER-RRMM: IberDd vs DVd - First-in-human CD19/BCMA in-vivo CAR-T SAME QUESTIONS. A BRIGHTER TOMORROW. SATURDAY - SEPT 26 8:30-9:30 AM - HALL 5 THE NEXT CELLULAR-THERAPY ERA (CELLULAR THERAPIES: REAL POTENTIAL FOR REAL PEOPLE) - Updated CEPHEUS MRD - BCMAxGPRC5D dual-target CAR-T - IASO206: IV in-vivo BCMA CAR-T 11:00 AM-12:00 PM - HALL 5 - Post-CAR-T elranatamab - Delayed cilta-cel neurotoxicity - MRD-guided T-cell redirection 11:00 AM-12:00 PM - LOMOND - CAR-T enterocolitis biology 1:00-2:00 PM - LOMOND - LV169: lentiviral in-vivo BCMA CAR delivery Same goal. Brighter lives. WHAT TO WATCH (DIFFERENT PERSPECTIVES A CLEARER FUTURE) - Durability - OS - sustained MRD - infection burden - Access - treatment-free time - late toxicity - In-vivo CAR-T: insertion sites, off-target transduction, vector shedding & persistence Go for the randomized data and the biology--not every new response rate. Single malt only AFTER the final session. SAME CURIOSITY FURTHER TOGETHER | GREAT SCIENCE DESERVES A DRAM

[Slide 1 - Parkinsonism with Cilta-cel, mechanism: potential on-target, off-tumor effect on basal ganglia] Cross-trial table (CARTITUDE-1 / CARTITUDE-4 / RWE USMMIC / RWE USMMIC-2): Incidence: 6% / 0.5% / 2% / 3% (23/766). Clinical features: tremor/bradykinesia/rigidity/gait difficulty (CARTITUDE-1, USMMIC, USMMIC-2), tremor/bradykinesia/rigidity (CARTITUDE-4), USMMIC-2 adds personality change. Median time to onset: 27 days (14-108) / 85 days / 30 days (13-69) / 32 days (6-277). Sources: Martin et al. J Clin Oncol 41:1265-1274, 2023. San-Miguel et al. NEJM 389:335-347, 2023. Sidana et al. Blood. 2025;145(1):85-97. Sidana S et al, ASH 2025, manuscript in preparation 2026. Cohen et al Blood Cancer J. 2022;12:32. Hosoya et al Blood 2025. [Slide 2 - Parkinsonism: Treatment and Outcomes] N=23 patients with Parkinsonism. Resolved or improved: 55%. Resolved: 30% (7). Improved: 22% (5). Median time to resolution: 29 days (range 7-427). Patient-level treatment/outcome grid (23 patients) showing treatments received (corticosteroids, IVIG, anakinra, ruxolitinib, tocilizumab, cyclophosphamide, dasatinib, intrathecal chemotherapy, high-dose methotrexate, anti-parkinsonian medication, other supportive/symptom care, no treatment documented) mapped against Favorable/Unfavorable/Unknown outcome. Source: Sidana S et al. Manuscript in preparation, 2026. [Slide 3 - IEC-Colitis: My Management Approach] Diagnosis: Clinical and Pathological - GI infectious work up (infections can co-occur); Endoscopy + biopsies (duodenum: 80%; including normal mucosa); Assess lymphocytic infiltrate for clonality and CAR-T presence. First line: Early treatment initiation is key; High-dose Cytoxan (1-2 g/m2): preferred for fit patients. Other options: Cyclosporine (target trough 300-400 ng/mL); Ruxolitinib; Biologics for IBD: infliximab, vedolizumab, ustekinumab. Supportive Care: Often need parenteral nutrition; Aggressive antimicrobial prophylaxis; If on steroids for >5-7 days, antifungal prophylaxis. [Slide 4 - Cerebellar Toxicity with GPRC5D CAR-T and BsAb: Dizziness-Ataxia Syndrome] Potential low-level expression in inferior olivary nucleus of medulla oblongata. Cross-trial table (Arlo-cel N=84 / MonumenTAL-3 Tal-DP/Tal-D / MonumenTAL-1 Monotherapy): Any-grade 12% / 15%/12% / 5%. Grade 3: 7% / 3%/2% / 0.5%. Median time to onset: 30.5 days / 292/326 days / ~114 days. Recovery/resolution: "4 resolved, but recurred; all active" / 10.8%/16.9% / 54.2%. Incidence: 12-14%; Early for CAR-T, risk increases over time for BsAb. Diagnosis and Treatment: Diagnosis is clinical; Talquetamab: Hold drug, if rechallenged after resolution use a lower dose; Arlo-cel: Consider cytotoxic therapy and JAK inhibitors, ~Parkinsonism. Sources: Bal S et al. Blood. 2026;148:1240-1250. Mailankody et al. N Engl J Med 387:1196-1206. Mina R et al. N Engl J Med. 2026;395:671-683. Janakiram M et al. Blood Cancer J. 2025;15:135.

[Slide 1 - Incidence and severity of infections across different approved BsAbs and CAR T trials, cross-trial compilation table] BsAbs late-line relapse (triple-class exposed) - trial/drug/All grade/Grade 3-4/Grade 5: MajesTEC-1/Teclistamab: 78.8% / 55.2% / 12.7%. MagnetisMM-3/Elranatamab: 69.9% / 39.8% / 6.5%. LinkerMM-1/Linvoseltamab: 75% / 48% / NR. MonumenTAL-1/Talquetamab: 66.2% / 14.5% / 1.4%. RedirecTT-1/Tec-Tal: 89% / 64% / 12%. BsAbs early-line relapse (1-3 prior lines): MajesTEC-9/Teclistamab: 82.8% / 41.6% / 5.5%. MajesTEC-3/Tec-Dara: 96.5% / 41.6% / 5.5%. MonumenTAL-3/Tal-Dara: 83.0% / 29.2% / 1.5%. MonumenTAL-3/Tal-Dara-Pom: 84.3% / 37.7% / 0.7%. CAR T late-line relapse: CARTITUDE-1/Cilta-Cel: All grade 58%, Gr3-4 20%, Gr5 NR. KarMMa/Ide-Cel: 69%, 22%, NR (#, deaths from AEs not otherwise specified 7% total). CAR T early-line relapse: CARTITUDE-4/Cilta-Cel: 59%, 24%, 4%. KarMMa-3/Ide-Cel: 56%, 22%, NR (*, death from AEs 6% total). "Infections remain a major safety concern with T-cell-redirecting therapies, particularly with continuous bispecific antibody treatment." Table provided for ease of viewing, no comparisons intended. Sources: Moreau P et al NEJM 2022; Lesokhin AM et al Nat Med 2023;29(9):2259-2267; Lee HC et al Clin Lymphoma Myeloma Leuk 2026 Feb, 26(2):e201-e212,e8; Chari A et al NEJM 2022; Cohen Y et al Engl J Med 2025;392:138-49; Touzeau C et al N Engl J Med 2026;395:440-53; Costa LJ et al N Eng J Med 2026;394:739-52; Mina R et al N Engl J Med 2026;395:671-83; Einsele H et al, Lancet Oncol 2026;27:254-68; Ailawadhi S et al, Blood 2024 (144)23; Berdeja J et al, Lancet 2021;398:314-24; Munshi NC et al, N Engl J Med 2021 Feb 25;384(8):705-716. [Slide 2 - Pillars of infection prevention in patients receiving T-cell redirecting therapies] Table: Immune defect/risk -> Type of infection -> Preventive strategy. Lymphopenia/Neutropenia -> Bacterial infections/PJP, approx first 3-6 months -> Antiviral + PJP prophylaxis, G-CSF support. Immunosuppressive therapy (steroids, IL-6 blockage, cyclophosphamide) -> Bacterial infections/Sepsis/fungal infections/CMV -> Consider antibacterial/antifungal. B-cell aplasia/Hypogammaglobulinemia -> Bacterial/Viral (Respiratory virus, VZV/HSV)/CMV -> Immunoglobulin replacement, Vaccination. T-cell dysfunction -> Other opportunistic infections, Viral infections, fungal (in selected high risk patients) -> Fixed treatment duration/Schedule optimization. Sources: Rodriguez-Otero P et al, Lancet Oncology May 2024; O'Neill C, et al. eJHaem. 2023;4:811-822; Friedrich MJ, et al. Cancer Cell. 2023;41:711-725.e6; Frerichs KA et al, Blood Advances 2024; Mohan M et al, Blood Cancer Journal (2025) 15:74; Tix T et al, Molecular Therapy (2025)33:7; Richardson T et al, Blood Advances (2025) 9:6; Nath K et al, Blood Cancer Journal (2024) 14:88; Reynolds G et al, Blood advances (2023) 7:19; Banerjee R et al, Blood Advances (2025) 9:18; Dr Rodriguez Otero personal opinion.

Phase 3 evidence has moved inmune therapy into early relapse [Table: STUDY | COMPARISON | forest plot (FAVOURS EXPERIMENTAL / FAVOURS CONTROL) | PFS HR | OS HR] CARTITUDE-4 | Cilta-cel vs DPd/PVd | PFS HR 0.29 | OS HR 0.55 DREAMM-7 | BVd vs DVd | PFS HR 0.41 | OS HR 0.58 DREAMM-8 | BPd vs PVd | PFS HR 0.52 | OS HR 0.77 MajesTEC-3 | Tec-Dara vs DPd/DVd | PFS HR 0.17 | OS HR 0.46 MonumenTAL-3 | Tal-DP vs DPd | PFS HR 0.28 | OS HR 0.47 (MonumenTAL-3) | Tal-D vs DPd | PFS HR 0.33 | OS HR 0.51 MajesTEC-9 | Teclistamab vs PVd/Kd | PFS HR 0.29 | OS HR 0.60 MonumenTAL-6 | Tal-Tec vs EPd/PVd | PFS HR 0.11 | OS HR NR (MonumenTAL-6) | Tal-Pom vs EPd/PVd | PFS HR 0.27 | OS HR NR CERVINO | Etentamig vs investigator's choice | PFS HR 0.40 | OS HR NR RANDOMIZED PHASE 3 TRIALS WITHOUT PUBLIC NUMERICAL HRs MagnetisMM-5 | Elranatamab +/- Dara vs DPd | Positive top-line result; no numerical data reported [Forest plot x-axis tick labels largely illegible]

International Myeloma Society New 2nd line treatments according to patient characteristics Columns: Tec +Ig (MajesTec 9) | Tec Dara +Ig (MajesTec 3) | Tal Dara (MonumenTal 3) | Tal Dara Pom (MonumenTal 3) | Mezi Kd (Successor 2) | [Press release only:] Tal Pom (MonumenTal 6) | Tec Tal +Ig (MonumenTal 6) | Elra +Ig (MagnetisMM 5) CD38 refractory: Yes | Possibly* | Possibly* | Possibly* | Yes | Yes | Yes | ? High risk: Possibly | Yes | Possibly | Possibly | Yes | Yes | Yes | ? EMD: Possibly | Possibly | Possibly | Possibly | Possibly | Possibly | Yes | ? Frail: Yes | Caution | Possibly | Caution | Yes | Possibly | Yes | ? High Infection risk: Possibly | Caution | Yes | Caution | Possibly | Possibly | Caution | ? Tolerability: Yes | Yes | Possibly | Possibly | Yes | Possibly | Caution | ? [Note: row-to-cell alignment approximate due to photo angle] Yes = Preferred / data supports; Possibly = Conditional / Limited data; Caution = possible but caution required Tec = Teclistamab | Tal = Talquetamab | Elra = Elranatamab | Dara = Daratumumab | Pom = Pomalidomide | Ig = Immunoglobulin | Mezi = mezigdomide | K = carilzomib | *based on TRIMM-2 data Based on trial eligibility + published efficacy/safety signals for BCMA & GPRC5D bispecifics. Table represents personal opinion and based on limited evidence! Costa LJ, et al. N Engl J Med. 2026;394(8):739-752, Touzeau C, et al. N Engl J Med. 2026, Mina R, et al. N Engl J Med. 2026, Dimopoulos et al., The Lancet 2026, Johnson & Johnson. press release, 23 July 2026, Pfizer. press release, 29 April 2026.

[Slide 1 - Incidence and severity of infections across different approved BsAbs and CAR T trials, cross-trial compilation table] BsAbs late-line relapse (triple-class exposed) - trial/drug/All grade/Grade 3-4/Grade 5: MajesTEC-1/Teclistamab: 78.8% / 55.2% / 12.7%. MagnetisMM-3/Elranatamab: 69.9% / 39.8% / 6.5%. LinkerMM-1/Linvoseltamab: 75% / 48% / NR. MonumenTAL-1/Talquetamab: 66.2% / 14.5% / 1.4%. RedirecTT-1/Tec-Tal: 89% / 64% / 12%. BsAbs early-line relapse (1-3 prior lines): MajesTEC-9/Teclistamab: 82.8% / 41.6% / 5.5%. MajesTEC-3/Tec-Dara: 96.5% / 41.6% / 5.5%. MonumenTAL-3/Tal-Dara: 83.0% / 29.2% / 1.5%. MonumenTAL-3/Tal-Dara-Pom: 84.3% / 37.7% / 0.7%. CAR T late-line relapse: CARTITUDE-1/Cilta-Cel: All grade 58%, Gr3-4 20%, Gr5 NR. KarMMa/Ide-Cel: 69%, 22%, NR (#, deaths from AEs not otherwise specified 7% total). CAR T early-line relapse: CARTITUDE-4/Cilta-Cel: 59%, 24%, 4%. KarMMa-3/Ide-Cel: 56%, 22%, NR (*, death from AEs 6% total). "Infections remain a major safety concern with T-cell-redirecting therapies, particularly with continuous bispecific antibody treatment." Table provided for ease of viewing, no comparisons intended. Sources: Moreau P et al NEJM 2022; Lesokhin AM et al Nat Med 2023;29(9):2259-2267; Lee HC et al Clin Lymphoma Myeloma Leuk 2026 Feb, 26(2):e201-e212,e8; Chari A et al NEJM 2022; Cohen Y et al Engl J Med 2025;392:138-49; Touzeau C et al N Engl J Med 2026;395:440-53; Costa LJ et al N Eng J Med 2026;394:739-52; Mina R et al N Engl J Med 2026;395:671-83; Einsele H et al, Lancet Oncol 2026;27:254-68; Ailawadhi S et al, Blood 2024 (144)23; Berdeja J et al, Lancet 2021;398:314-24; Munshi NC et al, N Engl J Med 2021 Feb 25;384(8):705-716. [Slide 2 - Pillars of infection prevention in patients receiving T-cell redirecting therapies] Table: Immune defect/risk -> Type of infection -> Preventive strategy. Lymphopenia/Neutropenia -> Bacterial infections/PJP, approx first 3-6 months -> Antiviral + PJP prophylaxis, G-CSF support. Immunosuppressive therapy (steroids, IL-6 blockage, cyclophosphamide) -> Bacterial infections/Sepsis/fungal infections/CMV -> Consider antibacterial/antifungal. B-cell aplasia/Hypogammaglobulinemia -> Bacterial/Viral (Respiratory virus, VZV/HSV)/CMV -> Immunoglobulin replacement, Vaccination. T-cell dysfunction -> Other opportunistic infections, Viral infections, fungal (in selected high risk patients) -> Fixed treatment duration/Schedule optimization. Sources: Rodriguez-Otero P et al, Lancet Oncology May 2024; O'Neill C, et al. eJHaem. 2023;4:811-822; Friedrich MJ, et al. Cancer Cell. 2023;41:711-725.e6; Frerichs KA et al, Blood Advances 2024; Mohan M et al, Blood Cancer Journal (2025) 15:74; Tix T et al, Molecular Therapy (2025)33:7; Richardson T et al, Blood Advances (2025) 9:6; Nath K et al, Blood Cancer Journal (2024) 14:88; Reynolds G et al, Blood advances (2023) 7:19; Banerjee R et al, Blood Advances (2025) 9:18; Dr Rodriguez Otero personal opinion.

Phase 3 evidence has moved inmune therapy into early relapse [Table: STUDY | COMPARISON | forest plot (FAVOURS EXPERIMENTAL / FAVOURS CONTROL) | PFS HR | OS HR] CARTITUDE-4 | Cilta-cel vs DPd/PVd | PFS HR 0.29 | OS HR 0.55 DREAMM-7 | BVd vs DVd | PFS HR 0.41 | OS HR 0.58 DREAMM-8 | BPd vs PVd | PFS HR 0.52 | OS HR 0.77 MajesTEC-3 | Tec-Dara vs DPd/DVd | PFS HR 0.17 | OS HR 0.46 MonumenTAL-3 | Tal-DP vs DPd | PFS HR 0.28 | OS HR 0.47 (MonumenTAL-3) | Tal-D vs DPd | PFS HR 0.33 | OS HR 0.51 MajesTEC-9 | Teclistamab vs PVd/Kd | PFS HR 0.29 | OS HR 0.60 MonumenTAL-6 | Tal-Tec vs EPd/PVd | PFS HR 0.11 | OS HR NR (MonumenTAL-6) | Tal-Pom vs EPd/PVd | PFS HR 0.27 | OS HR NR CERVINO | Etentamig vs investigator's choice | PFS HR 0.40 | OS HR NR RANDOMIZED PHASE 3 TRIALS WITHOUT PUBLIC NUMERICAL HRs MagnetisMM-5 | Elranatamab +/- Dara vs DPd | Positive top-line result; no numerical data reported [Forest plot x-axis tick labels largely illegible]

International Myeloma Society New 2nd line treatments according to patient characteristics Columns: Tec +Ig (MajesTec 9) | Tec Dara +Ig (MajesTec 3) | Tal Dara (MonumenTal 3) | Tal Dara Pom (MonumenTal 3) | Mezi Kd (Successor 2) | [Press release only:] Tal Pom (MonumenTal 6) | Tec Tal +Ig (MonumenTal 6) | Elra +Ig (MagnetisMM 5) CD38 refractory: Yes | Possibly* | Possibly* | Possibly* | Yes | Yes | Yes | ? High risk: Possibly | Yes | Possibly | Possibly | Yes | Yes | Yes | ? EMD: Possibly | Possibly | Possibly | Possibly | Possibly | Possibly | Yes | ? Frail: Yes | Caution | Possibly | Caution | Yes | Possibly | Yes | ? High Infection risk: Possibly | Caution | Yes | Caution | Possibly | Possibly | Caution | ? Tolerability: Yes | Yes | Possibly | Possibly | Yes | Possibly | Caution | ? [Note: row-to-cell alignment approximate due to photo angle] Yes = Preferred / data supports; Possibly = Conditional / Limited data; Caution = possible but caution required Tec = Teclistamab | Tal = Talquetamab | Elra = Elranatamab | Dara = Daratumumab | Pom = Pomalidomide | Ig = Immunoglobulin | Mezi = mezigdomide | K = carilzomib | *based on TRIMM-2 data Based on trial eligibility + published efficacy/safety signals for BCMA & GPRC5D bispecifics. Table represents personal opinion and based on limited evidence! Costa LJ, et al. N Engl J Med. 2026;394(8):739-752, Touzeau C, et al. N Engl J Med. 2026, Mina R, et al. N Engl J Med. 2026, Dimopoulos et al., The Lancet 2026, Johnson & Johnson. press release, 23 July 2026, Pfizer. press release, 29 April 2026.

[Slide 1 - GEM2014MAIN: Discontinuation of Rd or IRd - patients who were MRD- at second year of maintenance (N=271), median follow-up 7.5 years] KM curves for PFS by MRD status at 2yr (T5), comparing Continue Lenalidomide vs Stop Lenalidomide within MRD- and MRD+ groups. Median PFS: Stop Lenalidomide (Not reached) vs Continue Lenalidomide (3.5 years), p<0.0001. At-risk table shown through year 9. [Slide 2 - MRD negativity achievement is an independent prognostic factor for prolonged PFS] Subgroup KM curves by ISS stage and cytogenetic risk, continue vs stop lenalidomide maintenance. Median PFS by ISS: Continue Maintenance ISS1 5.7yr, ISS2 5.9yr, ISS3 2.9yr (Stop Maintenance not reached). By cytogenetic risk: Continue Maintenance-HR CG 2.8yr, Continue Maintenance-SR CG 4.2yr (Stop Maintenance not reached). [Slide 3 - Half of patients who were MRD- at 2 years and discontinued maintenance remained alive and progression-free after 7 years, N=84] After 7 years of maintenance discontinuation: 3 patients had disease progression, 1 died without progression. PFS KM curve (Sustained MRDneg n=64 vs Variable MRD status n=20) shows Sustained MRDneg group staying near 95-100% through 1.5 years vs Variable MRD status group dropping to ~88%.

DON'T GET LOST AT IMS 2026 GLASGOW - SEPTEMBER 23-26 THE MUST-SEE ROADMAP PEOPLE SCIENCE PROGRESS TOGETHER | BIGGER SCIENCE A BRIGHTER TOMORROW | GLASGOW WELCOMES A BRIGHTER TOMORROW THURSDAY - SEPT 24 5:00-6:00 PM - HALL 5 MRD & MAINTENANCE (MRD: DEEPER ANSWERS BRIGHTER TOMORROWS) - MIDAS: Isa-iberdomide - CASSIOPEIA: 5-year sustained MRD - GEM2014MAIN: MRD-guided stopping - IBEX: Iber-Dara post-ASCT FRIDAY - SEPT 25 3:00-5:00 PM - HALL 5 DO NOT MISS THE PLENARY - DO NOT MISS (BOLD SCIENCE CHANGES LIVES) - CERVINO: Etentamig vs standard therapy - PERSEUS: Does long-term PFS become OS? - EXCALIBER-RRMM: IberDd vs DVd - First-in-human CD19/BCMA in-vivo CAR-T SAME QUESTIONS. A BRIGHTER TOMORROW. SATURDAY - SEPT 26 8:30-9:30 AM - HALL 5 THE NEXT CELLULAR-THERAPY ERA (CELLULAR THERAPIES: REAL POTENTIAL FOR REAL PEOPLE) - Updated CEPHEUS MRD - BCMAxGPRC5D dual-target CAR-T - IASO206: IV in-vivo BCMA CAR-T 11:00 AM-12:00 PM - HALL 5 - Post-CAR-T elranatamab - Delayed cilta-cel neurotoxicity - MRD-guided T-cell redirection 11:00 AM-12:00 PM - LOMOND - CAR-T enterocolitis biology 1:00-2:00 PM - LOMOND - LV169: lentiviral in-vivo BCMA CAR delivery Same goal. Brighter lives. WHAT TO WATCH (DIFFERENT PERSPECTIVES A CLEARER FUTURE) - Durability - OS - sustained MRD - infection burden - Access - treatment-free time - late toxicity - In-vivo CAR-T: insertion sites, off-target transduction, vector shedding & persistence Go for the randomized data and the biology--not every new response rate. Single malt only AFTER the final session. SAME CURIOSITY FURTHER TOGETHER | GREAT SCIENCE DESERVES A DRAM

Table 1. Characteristics of the Patients at Baseline.* Characteristic | Dose Level 1 (N=32) | Dose Level 2 (N=6) | Total (N=38) Age Median (range) - yr | 66 (44-76) | 60 (52-65) | 66 (44-76) >=65 yr - no. (%) | 19 (59) | 1 (17) | 20 (53) Sex - no. (%) Male | 18 (56) | 5 (83) | 23 (61) Female | 14 (44) | 1 (17) | 15 (39) Race - no. (%) [dagger] White | 28 (88) | 4 (67) | 32 (84) Black | 3 (9) | 1 (17) | 4 (11) Asian | 1 (3) | 0 | 1 (3) Other | 0 | 1 (17) | 1 (3) ECOG performance-status score - no. (%) [double dagger] 0 | 9 (28) | 3 (50) | 12 (32) 1 | 23 (72) | 3 (50) | 26 (68) High-risk prognostic features - no. (%) Extramedullary disease [section] | 9 (28) | 3 (50) | 12 (32) Bone marrow plasma cells >=60% [paragraph] | 8 (25) | 4 (67) | 12 (32) Beta2-microglobulin level >=5.5 mg/liter [paragraph] | 5 (16) | 2 (33) | 7 (18) High-risk cytogenetic abnormality [double vertical bar] | 9 (28) | 2 (33) | 11 (29) High-risk cytogenetic abnormality inclusive of 1q gain | 21 (66) | 5 (83) | 26 (68) Median no. of previous lines of therapy (range) | 5 (3-7) | 4 (3-14) | 4 (3-14) Triple-class-refractory disease - no. (%)** | 32 (100) | 6 (100) | 38 (100) Penta-drug-refractory disease - no. (%) [dagger dagger] | 21 (66) | 5 (83) | 26 (68) Previous autologous stem-cell transplantation - no. (%) | 24 (75) | 4 (67) | 28 (74) Previous BCMA-directed therapy - no. (%) [double dagger double dagger] | 1 (3) | 0 | 1 (3) Median time since diagnosis (range) - yr | 6.5 (1.5-14.9) | 6.9 (1.7-11.0) | 6.5 (1.5-14.9) Bridging therapy - no. (%) | 22 (69) | 6 (100) | 28 (74) Footnotes: * Patients in the dose level 1 cohort received 100x10^6 chimeric antigen receptor (CAR) T cells, and those in the dose level 2 cohort received 300x10^6 CAR T cells. Percentages may not total 100 because of rounding. dagger Race was reported by the patient. double dagger Eastern Cooperative Oncology Group (ECOG) performance-status scores are assessed on a scale from 0 to 5, with higher scores indicating greater disability. section Extramedullary disease was defined as the presence of a non-bone-based plasmacytoma at the screening or baseline visit. paragraph Data are according to the worse value observed at the screening or baseline visit. double vertical bar High-risk cytogenetic abnormality was defined as the presence of del 17p, t(14;16), or t(4;14). ** Triple-class-refractory disease was defined as disease refractory to an immunomodulatory agent, a proteasome inhibitor, and an anti-CD38 monoclonal antibody. dagger dagger Penta-drug-refractory disease was defined as disease refractory to at least two unique immunomodulatory drugs, two unique proteasome inhibitors, and one anti-CD38 therapy. double dagger double dagger B-cell maturation antigen (BCMA)-directed therapy was defined as a BCMA-directed antibody-drug conjugate. Figure S1. Patient Disposition. Flow diagram: Enrolled & Leukapheresed n=40 -> Successful Manufacture of Anito-cel n=40 -> [Discontinued: Infection, n=1] -> Lymphodepletion n=39 -> [Discontinued: Hypoxia/Heart Failure, n=1] -> Total Dosed n=38 -> Dose level 1: 100 x 10^6 CAR+ T cells, n=32; Dose Escalation, n=6; Expansion Cohort, n=26 -> Dose level 2: 300 x 10^6 CAR+ T cells, n=6; Dose Escalation, n=6 4-panel NEJM Kaplan-Meier figure. Panel A - Progression-free Survival among All Patients: KM curve, y-axis "Percentage of Patients" 0-100, x-axis "Months" 0-60. Table: No. of Patients | Median Progression-free Survival (95% CI), mo: All Patients 38 | 30.2 (16.6-36.2). No. at Risk: 38, 34, 28, 24, 19, 13, 7, 3, 2, 0, 0 (at 0,6,12,18,24,30,36,42,48,54,60 mo approx). Panel B - Overall Survival among All Patients: KM curve, same x-axis. Table: No. of Patients | Median Overall Survival (95% CI), mo: All Patients 38 | NR (33.3-NE). No. at Risk: 38, 37, 36, 31, 30, 23, 17, 8, 5, 2, 0. Panel C - Progression-free Survival According to Presence or Absence of Extramedullary Disease: KM curves for All patients (black), EMD (red), No EMD (blue), x-axis Months 0-51. Table: No. of Patients | Median PFS (95% CI), mo: All Patients 38 | 30.2 (16.6-36.2); EMD 12 | 36.2 (10.2-NE); No EMD 26 | 27.9 (16.0-34.3). No. at Risk: All patients 38,38,34,28,26,24,23,19,18,13,12,7,5,3,2,2,0; EMD 12,12,10,10,7,7,7,5,5,5,5,3,2,2,0; No EMD 26,26,24,21,19,17,16,14,13,8,7,4,3,0,0,0,0 (approx per time points shown). Panel D - Progression-free Survival According to Beta2-Microglobulin Level: KM curves for All patients (black), Beta2M <5.5 mg/liter (purple), Beta2M >=5.5 mg/liter (green), x-axis Months 0-51. Table: No. of Patients | Median PFS (95% CI), mo: All Patients 38 | 30.2 (16.6-36.2); Beta2M >=5.5 mg/liter 7 | 30.2 (11.0-NE); Beta2M <5.5 mg/liter 31 | 27.9 (16.0-39.9). No. at Risk: All patients 38,38,34,28,26,24,23,19,18,13,12,7,5,3,2,2,0; Beta2M >=5.5 mg/liter 7,7,7,7,6,5,4,4,4,2,1,0,0,0,0,0,0; Beta2M <5.5 mg/liter 31,31,27,21,20,19,19,15,14,11,11,7,5,3,2,2,0.

OncLive PHASE 2 LINKER-SMM1 TRIAL Linvoseltamab Yields Deep, MRD-Negative Responses in High-Risk Smoldering Myeloma Single-agent, fixed-duration linvoseltamab-gcpt (Lynozyfic) produced deep, durable responses with manageable safety in patients with high-risk smoldering multiple myeloma (N = 40) in the phase 2 LINKER-SMM1 trial. 100% Objective response rate among response-evaluable patients (n = 39) 85% Complete response or better - 97% achieved ≥VGPR; 13.3-mo median follow-up 100% MRD negativity (10-5) among all MRD-evaluable patients 45% Cytokine release syndrome - predominantly grade 1; no ICANS observed Grade 3 infections occurred in 20% of patients (no grade 4 infections); grade ≥3 TEAEs occurred in 70% of patients, with neutropenia the most common. No new safety signal. Source: Rodriguez-Otero P, Mateos M-V, et al. Presented at the 23rd International Myeloma Society (IMS) Annual Meeting & Exposition; September 23-26, 2026; Glasgow, Scotland. OncLive.com

International Myeloma Society 23rd Annual MEETING & EXPOSITION September 23-26, 2026 . Glasgow, Scotland WEDNESDAY, SEPTEMBER 23, 2026 | 12:00-12:50 Poster Discussion Co-chairs: Masahiro Abe, Eileen Boyle, Hartmut Goldschmidt, Peter Voorhees Updated safety and efficacy results for linvoseltamab (LINVO) in patients (pts) with high-risk smoldering multiple myeloma (HR-SMM): Phase 2 LINKER-SMM1 trial Paula Rodriguez-Otero Functional implications of FCRL5 mutations for cevostamab immune escape in multiple myeloma Umair Munawar Matching-adjusted indirect comparisons (MAICs) of efficacy and safety outcomes for anitocabtagene autoleucel versus ciltacabtagene autoleucel in 4L+ relapsed and/or refractory multiple myeloma (RRMM) Surbhi Sidana Prognostic Impact of [18F] FDG PET/CT Before and After BCMA-Directed CAR T-Cell Therapy in Relapsed/Refractory Multiple Myeloma: A Multicenter Analysis Luis-Esteban Tamariz-Amador Second Primary Malignancies After Autologous Transplantation for Multiple Myeloma in the Modern Treatment Era: Incidence, Risk Factors, and Prognostic Impact Using a Competing-Risk Approach Damian Mikulski Efficacy and Safety of Daratumumab Plus VCd in Newly Diagnosed Amyloid Light-chain Amyloidosis with Cardiac Involvement: A Pooled Analysis of Asian Patients from AMY2009 and AMY3001 Jin Lu POLB 001, an oral p38 MAPK inhibitor for the prevention of cytokine release syndrome Liam Tremble Proteogenomic Integration Reveals Predictive Signatures of Ibrutinib Response in Waldenstrom Macroglobulinemia Christos Vlachos Latent Effector T Cells Mediate Immunotherapy Responses in the Bone Marrow Microenvironment Niklas Kehl ALTITUDE - ALTernating Induction Therapies to achieve Undetectable Disease Endpoints: Ph 1/2 Dara-RVd/Teclistamab-RVd in Transplant Eligible Standard-Risk Newly Diagnosed Multiple Myeloma Neha Korde [Photo captions] Dr. Abe / Dr. Boyle / Dr. Goldschmidt / Dr. Kehl / Dr. Korde / Dr. Lu / Dr. Mikulski / Dr. Munawar / Dr. Rodriguez-Otero / Dr. Sidana / Dr. Tamariz-Amador / Dr. Tremble / Dr. Vlachos / Dr. Voorhees

DON'T GET LOST AT IMS 2026 GLASGOW - SEPTEMBER 23-26 THE MUST-SEE ROADMAP PEOPLE SCIENCE PROGRESS TOGETHER | BIGGER SCIENCE A BRIGHTER TOMORROW | GLASGOW WELCOMES A BRIGHTER TOMORROW THURSDAY - SEPT 24 5:00-6:00 PM - HALL 5 MRD & MAINTENANCE (MRD: DEEPER ANSWERS BRIGHTER TOMORROWS) - MIDAS: Isa-iberdomide - CASSIOPEIA: 5-year sustained MRD - GEM2014MAIN: MRD-guided stopping - IBEX: Iber-Dara post-ASCT FRIDAY - SEPT 25 3:00-5:00 PM - HALL 5 DO NOT MISS THE PLENARY - DO NOT MISS (BOLD SCIENCE CHANGES LIVES) - CERVINO: Etentamig vs standard therapy - PERSEUS: Does long-term PFS become OS? - EXCALIBER-RRMM: IberDd vs DVd - First-in-human CD19/BCMA in-vivo CAR-T SAME QUESTIONS. A BRIGHTER TOMORROW. SATURDAY - SEPT 26 8:30-9:30 AM - HALL 5 THE NEXT CELLULAR-THERAPY ERA (CELLULAR THERAPIES: REAL POTENTIAL FOR REAL PEOPLE) - Updated CEPHEUS MRD - BCMAxGPRC5D dual-target CAR-T - IASO206: IV in-vivo BCMA CAR-T 11:00 AM-12:00 PM - HALL 5 - Post-CAR-T elranatamab - Delayed cilta-cel neurotoxicity - MRD-guided T-cell redirection 11:00 AM-12:00 PM - LOMOND - CAR-T enterocolitis biology 1:00-2:00 PM - LOMOND - LV169: lentiviral in-vivo BCMA CAR delivery Same goal. Brighter lives. WHAT TO WATCH (DIFFERENT PERSPECTIVES A CLEARER FUTURE) - Durability - OS - sustained MRD - infection burden - Access - treatment-free time - late toxicity - In-vivo CAR-T: insertion sites, off-target transduction, vector shedding & persistence Go for the randomized data and the biology--not every new response rate. Single malt only AFTER the final session. SAME CURIOSITY FURTHER TOGETHER | GREAT SCIENCE DESERVES A DRAM

Updated NCT04751877 Data in 1L MM: anti-CD38 mAbs - IMS 2026 Trials: Perseus (EMN) | IMROZ (sanofi) | NCT04751877 (CHU Poitiers) | CEPHEUS (Janssen, Pharmaceutical Companies of Johnson & Johnson) Treatment Regimen: Perseus: Daratumumab + Bortezomib + Lenalidomide + Dexamethasone vs Bortezomib + Lenalidomide + Dexamethasone IMROZ: Isatuximab + Bortezomib + Lenalidomide + Dexamethasone vs Bortezomib + Lenalidomide + Dexamethasone NCT04751877: Isatuximab + Lenalidomide + Dexamethasone +/- Bortezomib vs Isatuximab + Lenalidomide + Dexamethasone CEPHEUS: Isatuximab + Bortezomib + Lenalidomide + Dexamethasone vs Bortezomib + Lenalidomide + Dexamethasone Median Follow-Up: Perseus: 47.5 months | IMROZ: 59.7 months | NCT04751877: 48.9 months | CEPHEUS: 76.0 months PFS: Perseus: 48 mo: 84.3% vs 67.7% IMROZ: 60 mo: 63.2% vs 45.2% NCT04751877: 18 mo: 87.2% vs 85.5%; 24 mo: 85.2% vs 80%; 36 mo: 78% vs 72.3%; 48 mo: 74% vs 65.7% CEPHEUS: 54 mo: 68.1% vs 49.5%; 72 mo: 59.3% vs 38.3% ORR%: Perseus: 96.6% vs 93.8% | IMROZ: 91.3% vs 92.3% | NCT04751877: 85% vs 78% | CEPHEUS: 97% vs 93% MRD negativity rate: Perseus: 75.2% vs 47.5% | IMROZ: 58.1% vs 43.6% | NCT04751877: 12 mo: 51% vs 21% | CEPHEUS: 61.1% vs 40% ≥CR rate: Perseus: 87.9% vs 70.1% | IMROZ: 74.7% vs 64.1% | NCT04751877: 58% vs 31% | CEPHEUS: 80.6% vs 61.4% Adapted from Dr. Akhil Santhosh (X 2025) LARVOL

Photo of auditorium screen, slide "Conclusions" (UKMRA - UK Myeloma Research Alliance): - The VRAT demonstrated a high negative predictive value (98.2%) for detecting significant ocular adverse events with BVd - This may be an effective method of detecting significant keratopathy and visual acuity changes without the need for ocular examination - A lower positive predictive value suggests a conservative approach - There may be dose holds in the absence of significant keratopathy - The VRAT may be subjective to patient/HCP interpretation - Training is required of the questionnaire to ensure accuracy and reproducibility - Confirmatory data will be provided by the ongoing Stage 2 - Assessment of VRAT safety in the absence of ocular examinations - Reduced frequency BVd schedule was well tolerated - Belamaf Q8W followed by Q12W maintenance with weekly bortezomib associated with few dose modifications and interruptions HCP: Health Care Professional

International Myeloma Society ProMMise: Study Schema UKMRA UK Myeloma Research Alliance Relapsed myeloma 1-3 prior lines of therapy -> Stage 1: BelaMaf* + bortezomib + dexamethasone, N=40 - Eye symptom questionnaire (VRAT) - Ocular Examination Stop/go — Endpoint: Toxicity Stage 2: BelaMaf* + bortezomib + dexamethasone, N=60 - Eye symptom questionnaire (VRAT) X Ocular Examination Endpoints: Primary: Safety Secondary: Response rate (>=VGPR), PFS, OS Dosing Schedule: Cycles 1-6 | Cycle 7+; Belamaf Q8W [then] Belamaf Q12W; Weekly Bortezomib (x4) + Dexamethasone (x8); Observation and response assessment only Reduced intensity schedule: - Belamaf Q8 for 6 months then Q12 - Bortezomib weekly for 6 months BelaMaf: Belantamab Mafodotin; VRAT: vision-related anamnestic tool International Myeloma Society Predictive Value of VRAT UKMRA UK Myeloma Research Alliance 83.3% Sensitivity — 5 / 6 true ocular events flagged 75.7% Specificity — 56 / 74 true negatives cleared 98.2% Negative Predictive Value — 56 / 57 VRAT-negatives confirmed 21.7% Positive Predictive Value — 5 / 23 VRAT-positives confirmed Total of 80 paired VRAT and ocular assessments, only 1 false negative VRAT: vision-related anamnestic tool International Myeloma Society Conclusions UKMRA UK Myeloma Research Alliance - The VRAT demonstrated a high negative predictive value (98.2%) for detecting significant ocular adverse events with BVd - This may be an effective method of detecting significant keratopathy and visual acuity changes without the need for ocular examination - A lower positive predictive value suggests a conservative approach - There may be dose holds in the absence of significant keratopathy - The VRAT may be subjective to patient/ HCP interpretation - Training is required of the questionnaire to ensure accuracy and reproducibility - Confirmatory data will be provided by the ongoing Stage 2 - Assessment of VRAT safety in the absence of ocular examinations - Reduced frequency BVd schedule was well tolerated - Belamaf Q8W followed by Q12W maintenance with weekly bortezomib was associated with few dose modifications and interruptions HCP: Health Care Professional

ProMMise: Study Schema (UKMRA - UK Myeloma Research Alliance) Relapsed myeloma, 1-3 prior lines of therapy -> Stage 1: BelaMaf* + bortezomib + dexamethasone, N=40 - Eye symptom questionnaire (VRAT): check - Ocular Examination: check Stop/go, Endpoint: Toxicity Stage 2: BelaMaf* + bortezomib + dexamethasone, N=60 - Eye symptom questionnaire (VRAT): check - Ocular Examination: X (not performed) Endpoints: Primary: Safety Secondary: Response rate (≥VGPR), PFS, OS Dosing Schedule diagram: Cycles 1-6 | Cycle 7+ Belamaf: Q8W (cycles 1-6) then Q12W (cycle 7+, 4 years post C1D1) Weekly Bortezomib (x4) + Dexamethasone (x8) during cycles 1-6; Observation and response assessment only in cycle 7+ Reduced intensity schedule: - Belamaf Q8 for 6 months then Q12 - Bortezomib weekly for 6 months BelaMaf: Belantamab Mafodotin; VRAT: vision-related anamnestic tool Adverse events Table (left): Adverse event | Any grade n (%) | Grade 3-4 n (%) Eye disorders*: 30 (75.0%) | 13 (32.5%) Fatigue: 16 (40.0%) | 0 (0.0%) Constipation: 13 (32.5%) | 0 (0.0%) Diarrhoea: 13 (32.5%) | 2 (5.0%) Insomnia: 11 (27.5%) | 1 (2.5%) Pain: 11 (27.5%) | 0 (0.0%) Platelet count decreased: 10 (25.0%) | 8 (20.0%) Cough: 9 (22.5%) | 0 (0.0%) Nausea: 8 (20.0%) | 0 (0.0%) Table (right): Adverse event | Any grade n (%) | Grade 3-4 n (%) Alanine aminotransferase increased: 7 (17.5%) | 2 (5.0%) Infections and infestations: 7 (17.5%) | 0 (0.0%) Peripheral neuropathy: 6 (15.0%) | 1 (2.5%) Oedema limbs: 6 (15.0%) | 0 (0.0%) Urinary tract infection: 5 (12.5%) | 1 (2.5%) Dyspnoea: 5 (12.5%) | 0 (0.0%) Lung infection: 2 (5.0%) | 1 (2.5%) Neutrophil count decreased: 2 (5.0%) | 1 (2.5%) Other: 33 (82.5%) | 12 (30.0%) *Includes cataract, blurred vision, dry eye, and related terms Predictive Value of VRAT 83.3% - Sensitivity - 5/6 true ocular events flagged 75.7% - Specificity - 56/74 true negatives cleared 98.2% - Negative Predictive Value - 56/57 VRAT-negatives confirmed 21.7% - Positive Predictive Value - 5/23 VRAT-positives confirmed Total of 80 paired VRAT and ocular assessments, only 1 false negative VRAT: vision-related anamnestic tool Treatment Compliance: Belantamab Mafodotin Dose reductions: Cycles per participant with modification Mean (SD): 0.1 (0.27) Median (IQR): 0.0 (0.0, 0.0) Range: 0 - 1 N: 40 Reason for dose modification: Toxicity 1 (33.3%), Other 2 (66.7%), Total 3 (100.0%) 3 (8%) patients required dose reduction Dose interruptions: Cycles per participant with modification Mean (SD): 0.6 (1.24) Median (IQR): 0.0 (0.0, 1.0) Range: 0 - 7 N: 40 Reason for dose modification: Toxicity 19 (79.2%), Other 5 (20.8%), Total 24 (100.0%) 14 (35%) patients had a dose interruption Low rates of dose reductions and interruptions with Q8W followed by Q12W Belamaf dosing

QoL Completion - 68/70 participants consented to QoL data collection - Data is immature for 12 and 24 months as follow-up is ongoing (Complete case set analysis; Descriptive analysis) Table, columns: Baseline | End of Induction | 100 days Post-ASCT | 6 Months | 12 Months | 24 Months | Total Questionnaire completed? Yes: 64 (94.1%) | 54 (79.4%) | 44 (64.7%) | 36 (52.9%) | 25 (36.8%) | 1 (1.5%) | 224 (54.9%) No: 4 (5.9%) | 14 (20.6%) | 24 (35.3%) | 32 (47.1%) | 43 (63.2%) | 67 (98.5%) | 184 (45.1%) Total: 68 (100%) | 68 (100%) | 68 (100%) | 68 (100%) | 68 (100%) | 68 (100%) | 408 (100%) Reason not completed: Reason Unknown: 2 (50.0%) | 7 (50.0%) | 14 (58.3%) | 24 (75.0%) | 26 (60.5%) | 0 (0.0%) | 73 (39.7%) Timepoint not yet reached: 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 8 (18.6%) | 56 (83.6%) | 64 (34.8%) Participant Death: 0 (0.0%) | 0 (0.0%) | 3 (12.5%) | 4 (12.5%) | 6 (14.0%) | 7 (10.4%) | 20 (10.9%) Participant Withdrawal: 0 (0.0%) | 1 (7.1%) | 1 (4.2%) | 3 (9.4%) | 3 (7.0%) | 4 (6.0%) | 12 (6.5%) Either not completed or misplaced by site: 1 (25.0%) | 1 (7.1%) | 3 (12.5%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 5 (2.7%) Admin error: 0 (0.0%) | 3 (21.4%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 3 (1.6%) Lost in post: 0 (0.0%) | 2 (14.3%) | 1 (4.2%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 3 (1.6%) Other: 1 (25.0%) | 0 (0.0%) | 0 (0.0%) | 1 (3.1%) | 0 (0.0%) | 0 (0.0%) | 2 (1.1%) Participant refused: 0 (0.0%) | 0 (0.0%) | 1 (4.2%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 1 (0.5%) Site missed in error: 0 (0.0%) | 0 (0.0%) | 1 (4.2%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 1 (0.5%) Total: 4 (100%) | 14 (100%) | 24 (100%) | 32 (100%) | 43 (100%) | 67 (100%) | 184 (100%) (Note: percentages for "Timepoint not yet reached" 24-month value shown on slide as 56 (63.6%) though row total math suggests possible slide rounding; transcribed exactly as displayed: 56 (63.6%).)

23rd Annual MEETING & EXPOSITION - International Myeloma Society THURSDAY, SEPTEMBER 24, 2026 | 11:00-12:00 Abstract Session 3 Co-chairs: Meral Beksac and Martin Kortum - Double hit ultra-high risk myeloma treated with Isa-VRD plus cyclophosphamide induction and Isa-VRD consolidation: QoL analysis of the UKMRA RADAR trial in newly diagnosed transplant eligible patients - Karthik Ramasamy - Updated safety and efficacy of ramantamig (Ram; JNJ-5322) at the recommended phase 2 dose (RP2D), demonstrating feasibility of outpatient dosing (OPD) in relapsed/refractory multiple myeloma (RRMM) - Jeffrey Matous - DREAMM-9 Patient-Reported Vision-Related Function with Extended Belantamab Mafodotin Dosing Intervals with Bortezomib, Lenalidomide, Dexamethasone (BVRd) in Transplant Ineligible Multiple Myeloma - Hang Quach - A fixed-duration, modified cevostamab dosing regimen induces durable remissions that continue after completion in BCMA-naive pts with relapsed/refractory multiple myeloma (RRMM): CAMMA 1 Arm A results - Amrita Krishnan - Overall Survival, Progression and Non-Relapse Mortality With Teclistamab Plus Daratumumab (Tec-Dara) vs Dara-Based Triplets in Relapsed/Refractory Multiple Myeloma (RRMM): MajesTEC-3 Post Hoc Analysis - Luciano Costa Panelist photos shown: Dr. Beksac, Dr. Kortum, Dr. Matous, Dr. Ramasamy, Dr. Costa, Dr. Krishnan, Dr. Quach 23rd Annual MEETING & EXPOSITION - International Myeloma Society THURSDAY, SEPTEMBER 24, 2026 | 11:00-12:00 Abstract Session 4 Co-chairs: Lugui Qiu and Niels van de Donk - Augmenting TRAIL-dependent apoptosis enables BCMA CAR-T cells to overcome resistance in multiple myeloma - Mansour Poorebrahim - First-in-human phase Ia/Ib study of QLS32015 (GPRC5DxCD3) in patients (pts) with relapsed/refractory multiple myeloma (RRMM): updated safety and efficacy data - Jian Cui - An AI-Designed BCMA/FcRL5 Targeting TriTE Overcomes Antigen Escape and Outperforms BCMA-Directed Therapy in Multiple Myeloma - Jessica Encinas Mayoral - A CD4-Dominant, T-Helper-Skewed T Cell State Underlies Movement and Neurocognitive Toxicity After BCMA-Directed CAR-T Therapy in Multiple Myeloma - Theresia Akhlaghi - Single-cell antigen:TCR mapping reveals tumor-specific effector T-cell responses in multiple myeloma - David Avigan Panelist photos shown: Dr. Akhlaghi, Dr. Avigan, Dr. Cui, Dr. Encinas Mayoral, Dr. Poorebrahim, Dr. Qiu, Dr. van de Donk

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 (EMN28 · NCT05257083) | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN (IFM 2025-01 · NCT06918002) | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 (EMN41 · MMY3003) | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 (NCT07735637) | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4-6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 (KT-US-715-0795) | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 5 Who might still need an alkylator? Profile | Why an alkylator might still matter | Evidence today t(11;14) | Overexpresses cyclin D1. BCL2 dependence -> cells more "primed" for apoptosis -> hypothesis: more sensitive to pro-apoptotic stress including melphalan | Hypothesis? MIDAS small numbers t(11;14): n=32 Arm A, n=46 Arm B Very high risk (biallelic TP53, ≥ 2 CGS criteria, PCL) | Proliferative disease: cytotoxic pressure may complement T-cell engagement | Tandem ASCT case by case, no robust data Extramedullary disease, high tumor burden | Rapid debulking alongside cell therapy (bridging, CRS risk) | Hypothesis? Antigen escape (BCMA loss) | Argues first for dual targeting (BCMA × GPRC5D, BCMA/CD19) rather than an alkylator | Being tested in the multi-target trials T-cell fitness, marrow microenvironment | Exhausted or immune-cold marrow: T-cell engagement may not suffice; Mel may modify the niche | Hypothesis: biomarker to test? The new trials need prespecified biology-based subgroups and translational cohorts: MRD kinetics, T-cell fitness, target expression International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 7

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 EMN28 · NCT05257083 | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN IFM 2025-01 · NCT06918002 | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 EMN41 · MMY3003 | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 NCT07735637 | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4–6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 KT-US-715-0795 | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society | 23rd International Myeloma Society Annual Meeting

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 (EMN28 · NCT05257083) | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN (IFM 2025-01 · NCT06918002) | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 (EMN41 · MMY3003) | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 (NCT07735637) | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4-6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 (KT-US-715-0795) | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 5 Who might still need an alkylator? Profile | Why an alkylator might still matter | Evidence today t(11;14) | Overexpresses cyclin D1. BCL2 dependence -> cells more "primed" for apoptosis -> hypothesis: more sensitive to pro-apoptotic stress including melphalan | Hypothesis? MIDAS small numbers t(11;14): n=32 Arm A, n=46 Arm B Very high risk (biallelic TP53, ≥ 2 CGS criteria, PCL) | Proliferative disease: cytotoxic pressure may complement T-cell engagement | Tandem ASCT case by case, no robust data Extramedullary disease, high tumor burden | Rapid debulking alongside cell therapy (bridging, CRS risk) | Hypothesis? Antigen escape (BCMA loss) | Argues first for dual targeting (BCMA × GPRC5D, BCMA/CD19) rather than an alkylator | Being tested in the multi-target trials T-cell fitness, marrow microenvironment | Exhausted or immune-cold marrow: T-cell engagement may not suffice; Mel may modify the niche | Hypothesis: biomarker to test? The new trials need prespecified biology-based subgroups and translational cohorts: MRD kinetics, T-cell fitness, target expression International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 7

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 EMN28 · NCT05257083 | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN IFM 2025-01 · NCT06918002 | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 EMN41 · MMY3003 | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 NCT07735637 | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4–6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 KT-US-715-0795 | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society | 23rd International Myeloma Society Annual Meeting

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 (EMN28 · NCT05257083) | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN (IFM 2025-01 · NCT06918002) | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 (EMN41 · MMY3003) | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 (NCT07735637) | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4-6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 (KT-US-715-0795) | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 5 Who might still need an alkylator? Profile | Why an alkylator might still matter | Evidence today t(11;14) | Overexpresses cyclin D1. BCL2 dependence -> cells more "primed" for apoptosis -> hypothesis: more sensitive to pro-apoptotic stress including melphalan | Hypothesis? MIDAS small numbers t(11;14): n=32 Arm A, n=46 Arm B Very high risk (biallelic TP53, ≥ 2 CGS criteria, PCL) | Proliferative disease: cytotoxic pressure may complement T-cell engagement | Tandem ASCT case by case, no robust data Extramedullary disease, high tumor burden | Rapid debulking alongside cell therapy (bridging, CRS risk) | Hypothesis? Antigen escape (BCMA loss) | Argues first for dual targeting (BCMA × GPRC5D, BCMA/CD19) rather than an alkylator | Being tested in the multi-target trials T-cell fitness, marrow microenvironment | Exhausted or immune-cold marrow: T-cell engagement may not suffice; Mel may modify the niche | Hypothesis: biomarker to test? The new trials need prespecified biology-based subgroups and translational cohorts: MRD kinetics, T-cell fitness, target expression International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 7

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 EMN28 · NCT05257083 | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN IFM 2025-01 · NCT06918002 | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 EMN41 · MMY3003 | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 NCT07735637 | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4–6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 KT-US-715-0795 | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society | 23rd International Myeloma Society Annual Meeting

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 (EMN28 · NCT05257083) | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN (IFM 2025-01 · NCT06918002) | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 (EMN41 · MMY3003) | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 (NCT07735637) | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4-6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 (KT-US-715-0795) | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 5 Who might still need an alkylator? Profile | Why an alkylator might still matter | Evidence today t(11;14) | Overexpresses cyclin D1. BCL2 dependence -> cells more "primed" for apoptosis -> hypothesis: more sensitive to pro-apoptotic stress including melphalan | Hypothesis? MIDAS small numbers t(11;14): n=32 Arm A, n=46 Arm B Very high risk (biallelic TP53, ≥ 2 CGS criteria, PCL) | Proliferative disease: cytotoxic pressure may complement T-cell engagement | Tandem ASCT case by case, no robust data Extramedullary disease, high tumor burden | Rapid debulking alongside cell therapy (bridging, CRS risk) | Hypothesis? Antigen escape (BCMA loss) | Argues first for dual targeting (BCMA × GPRC5D, BCMA/CD19) rather than an alkylator | Being tested in the multi-target trials T-cell fitness, marrow microenvironment | Exhausted or immune-cold marrow: T-cell engagement may not suffice; Mel may modify the niche | Hypothesis: biomarker to test? The new trials need prespecified biology-based subgroups and translational cohorts: MRD kinetics, T-cell fitness, target expression International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 7

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 EMN28 · NCT05257083 | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN IFM 2025-01 · NCT06918002 | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 EMN41 · MMY3003 | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 NCT07735637 | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4–6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 KT-US-715-0795 | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society | 23rd International Myeloma Society Annual Meeting

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 (EMN28 · NCT05257083) | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN (IFM 2025-01 · NCT06918002) | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 (EMN41 · MMY3003) | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 (NCT07735637) | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4-6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 (KT-US-715-0795) | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 5 Who might still need an alkylator? Profile | Why an alkylator might still matter | Evidence today t(11;14) | Overexpresses cyclin D1. BCL2 dependence -> cells more "primed" for apoptosis -> hypothesis: more sensitive to pro-apoptotic stress including melphalan | Hypothesis? MIDAS small numbers t(11;14): n=32 Arm A, n=46 Arm B Very high risk (biallelic TP53, ≥ 2 CGS criteria, PCL) | Proliferative disease: cytotoxic pressure may complement T-cell engagement | Tandem ASCT case by case, no robust data Extramedullary disease, high tumor burden | Rapid debulking alongside cell therapy (bridging, CRS risk) | Hypothesis? Antigen escape (BCMA loss) | Argues first for dual targeting (BCMA × GPRC5D, BCMA/CD19) rather than an alkylator | Being tested in the multi-target trials T-cell fitness, marrow microenvironment | Exhausted or immune-cold marrow: T-cell engagement may not suffice; Mel may modify the niche | Hypothesis: biomarker to test? The new trials need prespecified biology-based subgroups and translational cohorts: MRD kinetics, T-cell fitness, target expression International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 7

A new comparator: T-cell redirection Then: Mel 200 + ASCT vs the same drug classes (PI ± IMiD) → Now: Mel 200 + ASCT vs a new mechanism (T-cell engagement) Trial | Replaces ASCT with | Control arm | N | Status EMagine / CARTITUDE-6 EMN28 · NCT05257083 | Dara-VRd → cilta-cel (BCMA CAR-T) | Dara-VRd → ASCT | ~750 | Enrolment complete ELLEN IFM 2025-01 · NCT06918002 | D-VRd × 4 → elranatamab + len × 6 (BCMA × CD3; no ASCT) | D-VRd × 4 → ASCT → D-VRd × 2 | 824 | Recruiting TRIlogy-6 EMN41 · MMY3003 | Ramantamig + Dara, ≤ 48 months (BCMA × GPRC5D × CD3) | DVRd → ASCT → DVRd × 2 → DR | ~1,000 | Not yet open DURGA-6 NCT07735637 | AZD0120 (BCMA/CD19 dual CAR-T) | ASCT after 4–6 cycles of induction; both then lenalidomide | 750 | Opening 2026 iMMagine-6 KT-US-715-0795 | anti-CD38VRd → anito-cel (BCMA CAR-T) | Induction → ASCT → maintenance | — | Not yet open Primary read-outs pending: no first-line randomized data yet for any of these strategies. International Myeloma Society | 23rd International Myeloma Society Annual Meeting

Phase 3 evidence has moved inmune therapy into early relapse [Table: STUDY | COMPARISON | forest plot (FAVOURS EXPERIMENTAL / FAVOURS CONTROL) | PFS HR | OS HR] CARTITUDE-4 | Cilta-cel vs DPd/PVd | PFS HR 0.29 | OS HR 0.55 DREAMM-7 | BVd vs DVd | PFS HR 0.41 | OS HR 0.58 DREAMM-8 | BPd vs PVd | PFS HR 0.52 | OS HR 0.77 MajesTEC-3 | Tec-Dara vs DPd/DVd | PFS HR 0.17 | OS HR 0.46 MonumenTAL-3 | Tal-DP vs DPd | PFS HR 0.28 | OS HR 0.47 (MonumenTAL-3) | Tal-D vs DPd | PFS HR 0.33 | OS HR 0.51 MajesTEC-9 | Teclistamab vs PVd/Kd | PFS HR 0.29 | OS HR 0.60 MonumenTAL-6 | Tal-Tec vs EPd/PVd | PFS HR 0.11 | OS HR NR (MonumenTAL-6) | Tal-Pom vs EPd/PVd | PFS HR 0.27 | OS HR NR CERVINO | Etentamig vs investigator's choice | PFS HR 0.40 | OS HR NR RANDOMIZED PHASE 3 TRIALS WITHOUT PUBLIC NUMERICAL HRs MagnetisMM-5 | Elranatamab +/- Dara vs DPd | Positive top-line result; no numerical data reported [Forest plot x-axis tick labels largely illegible]

International Myeloma Society New 2nd line treatments according to patient characteristics Columns: Tec +Ig (MajesTec 9) | Tec Dara +Ig (MajesTec 3) | Tal Dara (MonumenTal 3) | Tal Dara Pom (MonumenTal 3) | Mezi Kd (Successor 2) | [Press release only:] Tal Pom (MonumenTal 6) | Tec Tal +Ig (MonumenTal 6) | Elra +Ig (MagnetisMM 5) CD38 refractory: Yes | Possibly* | Possibly* | Possibly* | Yes | Yes | Yes | ? High risk: Possibly | Yes | Possibly | Possibly | Yes | Yes | Yes | ? EMD: Possibly | Possibly | Possibly | Possibly | Possibly | Possibly | Yes | ? Frail: Yes | Caution | Possibly | Caution | Yes | Possibly | Yes | ? High Infection risk: Possibly | Caution | Yes | Caution | Possibly | Possibly | Caution | ? Tolerability: Yes | Yes | Possibly | Possibly | Yes | Possibly | Caution | ? [Note: row-to-cell alignment approximate due to photo angle] Yes = Preferred / data supports; Possibly = Conditional / Limited data; Caution = possible but caution required Tec = Teclistamab | Tal = Talquetamab | Elra = Elranatamab | Dara = Daratumumab | Pom = Pomalidomide | Ig = Immunoglobulin | Mezi = mezigdomide | K = carilzomib | *based on TRIMM-2 data Based on trial eligibility + published efficacy/safety signals for BCMA & GPRC5D bispecifics. Table represents personal opinion and based on limited evidence! Costa LJ, et al. N Engl J Med. 2026;394(8):739-752, Touzeau C, et al. N Engl J Med. 2026, Mina R, et al. N Engl J Med. 2026, Dimopoulos et al., The Lancet 2026, Johnson & Johnson. press release, 23 July 2026, Pfizer. press release, 29 April 2026.

FIRST RELAPSE Why CAR T belongs earlier Maximize early kill, then return time to the patient 48% MRD-negative by day 56 62% vs 18% overall MRD negativity at 10⁻⁵ Deepest early cytoreduction ENGINEERED IMMUNE RESET Lymphodepletion reduces suppressive immune cells; one infusion delivers a large, activated population that expands in vivo. A fundamentally different biologic attack than continuous antibody exposure +11.6 point QoL advantage at month 12 Global health status: +10.1 with cilta-cel vs −1.5 with SOC Treatment-free runway, less ongoing supportive care, and a PFS2/OS advantage SURVIVORSHIP IS PART OF THE STRATEGY Track immune reconstitution, IVIG and infection needs, revaccination, late effects and return to life. CARTITUDE-4; Mina et al, Lancet Haematol 2025; Popat et al, Blood 2024; updated OS analysis, Lancet 2026 Emerging CAR T platforms show deep early responses Latest publicly reported datasets; populations and follow-up differ Dataset | Anito-cel iMMagine-1 | Arlo-cel early-line phase 1 | AZD0120 DURGA-1 | KLN-1010 inMMyCAR Platform / target | Ex vivo BCMA | Ex vivo GPRC5D | Ex vivo BCMA + CD19 | In vivo BCMA Patients | 117 evaluable | 24 evaluable, 31 treated | 23 evaluable | 18 dosed Prior lines | Median 5 | Median 2, 1-3 allowed | Median 4 | Not reported Follow-up | Median 12.6 mo | Median 18.3 mo | Median 3.9 mo | 6 pts ≥4 mo, first pt >10 mo ORR | 97% | 96% | 96% | 100% all evaluable CR / sCR | 68% | 67% | 78.3% | 4 sCR / 6 pts with ≥4 mo MRD negative at 10⁻⁵ | 93% (70/75 evaluable) | 56.3% MRD-negative CR | 94% (16/17 evaluable) | 100% all evaluable 12-month PFS | 79% | 74% | Not reported | Not reported Deep responses now extend from ex vivo platforms to first-in-human in vivo CAR T Patel K, ASH 2025 #256. Bal S, Haematologica 2026. Richard S, Blood 2025 #269. Ho J, ASCO 2026. Cross-trial comparisons are descriptive. MRD denominators and follow-up maturity differ. Take-home messages • CAR T belongs at first relapse for eligible patients, before disease biology and treatment burden narrow the opportunity. • Earlier-line cilta-cel delivers deep real-world responses, with 79% progression-free survival at 12 months. • Earlier referral preserves more active holding-therapy options and helps protect T-cell fitness before collection. • Sequence intentionally: use holding therapy while access is secured, collect early, then bridge to control disease before infusion. • The platform is expanding beyond conventional BCMA CAR T to GPRC5D, dual-target, and in vivo approaches. Refer early, collect early, and preserve the path to CAR T

FIGURE 1. Study design Isatuximab SC + Kd: 28-day cycles (Parts 1 & 2) - Isatuximab SC: 1400 mg weekly in C1, then biweekly - Carfilzomib: 20 mg/m2 D1-2, then 56 mg/m2 D8-9, D15-16 in C1, then 56 mg/m2 D1-2, D8-9, D15-16 of further cycles - Dexamethasone: 20 mg D1-2, D8-9, D15-16, D22-23 Part 1: N=8 - Manual administration ~8 mins at C1D1 and afterwards ~6 mins - Manual administration ~6 mins from C1D1 on Part 2: N=66, Randomization 1:3 - n=24: OBI C1-C3 -> Manual C4-C6 - n=42: Manual C1-C3 -> OBI C4-C6 OBI or Manual per patient choice at C7+ C, cycle; D, day; Kd, carfilzomib + dexamethasone; OBI, on-body injector; SC, subcutaneous. FIGURE 2. Progression-free survivala Kaplan-Meier curve, Y axis "Progression-free survival probability" (0-1.0), X axis "Time (Months)" (0-26) Legend: Isa SC + Kd, + Censored Median PFS: NR (95% CI: 16.2-NR) PFS at 18 months: 61.5% (95% CI: 48.7-72.0) Patients at risk (Isa SC + Kd): 74, 68, 62, 56, 54, 51, 47, 44, 40, 31, 20, 9, 2, 0 aThe median follow up duration was 21.9 months. CI, confidence interval; Kd, carfilzomib + dexamethasone; NR, not reached; PFS, progression free survival; SC, subcutaneous. FIGURE 3. Patient experience and satisfaction Line chart, Y axis "Patients satisfied with injection method (%)" (0-100), X axis "Cycle" (C1D8 through C25D15) Legend: Isa SC Manual, Isa SC OBI - both lines generally 80-100% with some dips (e.g. one point drops to ~50% labeled "Very small sample size") Isa SC Manual - Total number of patients who completed the questionnaire (by cycle): 48 41 42 46 40 43 39 29 26 27 24 9 5 7 6 6 8 8 8 7 7 6 6 6 6 5 4 4 2 2 2 Isa SC OBI - Total number of patients who completed the questionnaire (by cycle): 23 23 24 24 20 23 22 35 32 35 31 33 27 48 40 48 44 40 38 37 37 40 31 35 34 30 33 29 22 18 17 15 14 9 C, cycle; D, day; OBI, on-body injector; SC, subcutaneous. Key Conclusions - Updated results of the IZALCO study continue to show the efficacy and safety of Isa SC delivered either by manual injection or OBI, plus Kd. - Patient satisfaction with Isa SC OBI remained consistently high after patient choice of injection method at C7D1. - These safety and efficacy findings are consistent with those reported in the Phase 3 IKEMA study (Isa IV plus Kd).5 - Overall, these results support the efficacy, safety and feasibility of Isa delivered subcutaneously via OBI in patients with RRMM.

FIGURE 1. Study design Isatuximab SC + Kd: 28-day cycles (Parts 1 & 2) - Isatuximab SC: 1400 mg weekly in C1, then biweekly - Carfilzomib: 20 mg/m2 D1-2, then 56 mg/m2 D8-9, D15-16 in C1, then 56 mg/m2 D1-2, D8-9, D15-16 of further cycles - Dexamethasone: 20 mg D1-2, D8-9, D15-16, D22-23 N=8 (Part 1): Manual administration ~8 mins at C1D1 and afterwards ~6 mins Manual administration ~6 mins from C1D1 on N=66 (Part 2), Randomization 1:3: n=24 -> OBI C1-C3 -> Manual C4-C6 n=42 -> Manual C1-C3 -> OBI C4-C6 OBI or Manual per patient choice at C7+ C, cycle; D, day; Kd, carfilzomib + dexamethasone; OBI, on-body injector; SC, subcutaneous. FIGURE 2. Progression-free survival(a) Kaplan-Meier curve, Isa SC + Kd (purple line), Censored (+ marks) Y-axis: Progression-free survival probability 0-1.0 X-axis: Time (Months) 0-26 Median PFS: NR (95% CI: 16.2-NR) PFS at 18 months: 61.5% (95% CI: 48.7-72.0) Number at risk row (Isa SC + Kd): 74 68 62 56 54 51 47 44 40 31 20 9 2 0 (at months 0,2,4,6,8,10,12,14,16,18,20,22,24,26) (a)The median follow up duration was 21.9 months. CI, confidence interval; carfilzomib + dexamethasone; NR, not reached; PFS, progression-free survival; SC, subcutaneous. FIGURE 3. Patient experience and satisfaction Line chart: Patients satisfied with injection method (%), Y-axis 0-100%, X-axis Cycle 1-25(D15) Two lines: Isa SC Manual and Isa SC OBI, both fluctuating mostly in 80-100% range across cycles Annotation arrow pointing to a dip around cycle 20 area: "Very small sample size" Below chart, two rows showing "Total number of patients who completed the questionnaire" per cycle for Isa SC Manual and Isa SC OBI, values decreasing over cycles (starting around 48-49 at cycle 1, declining to single digits by cycle 25) C, cycle; D, day; OBI, on-body injector; SC, subcutaneous. Key Conclusions - Updated results of the IZALCO study continue to show the efficacy and safety of Isa SC delivered either by manual injection or OBI, plus Kd. - Patient satisfaction with Isa SC OBI remained consistently high after patient choice of injection method at C7D1. - These safety and efficacy findings are consistent with those reported in the Phase 3 IKEMA study (Isa IV plus Kd).5 - Overall, these results support the efficacy, safety and feasibility of Isa delivered subcutaneously via OBI in patients with RRMM.

[Slide 1 - Incidence and severity of infections across different approved BsAbs and CAR T trials, cross-trial compilation table] BsAbs late-line relapse (triple-class exposed) - trial/drug/All grade/Grade 3-4/Grade 5: MajesTEC-1/Teclistamab: 78.8% / 55.2% / 12.7%. MagnetisMM-3/Elranatamab: 69.9% / 39.8% / 6.5%. LinkerMM-1/Linvoseltamab: 75% / 48% / NR. MonumenTAL-1/Talquetamab: 66.2% / 14.5% / 1.4%. RedirecTT-1/Tec-Tal: 89% / 64% / 12%. BsAbs early-line relapse (1-3 prior lines): MajesTEC-9/Teclistamab: 82.8% / 41.6% / 5.5%. MajesTEC-3/Tec-Dara: 96.5% / 41.6% / 5.5%. MonumenTAL-3/Tal-Dara: 83.0% / 29.2% / 1.5%. MonumenTAL-3/Tal-Dara-Pom: 84.3% / 37.7% / 0.7%. CAR T late-line relapse: CARTITUDE-1/Cilta-Cel: All grade 58%, Gr3-4 20%, Gr5 NR. KarMMa/Ide-Cel: 69%, 22%, NR (#, deaths from AEs not otherwise specified 7% total). CAR T early-line relapse: CARTITUDE-4/Cilta-Cel: 59%, 24%, 4%. KarMMa-3/Ide-Cel: 56%, 22%, NR (*, death from AEs 6% total). "Infections remain a major safety concern with T-cell-redirecting therapies, particularly with continuous bispecific antibody treatment." Table provided for ease of viewing, no comparisons intended. Sources: Moreau P et al NEJM 2022; Lesokhin AM et al Nat Med 2023;29(9):2259-2267; Lee HC et al Clin Lymphoma Myeloma Leuk 2026 Feb, 26(2):e201-e212,e8; Chari A et al NEJM 2022; Cohen Y et al Engl J Med 2025;392:138-49; Touzeau C et al N Engl J Med 2026;395:440-53; Costa LJ et al N Eng J Med 2026;394:739-52; Mina R et al N Engl J Med 2026;395:671-83; Einsele H et al, Lancet Oncol 2026;27:254-68; Ailawadhi S et al, Blood 2024 (144)23; Berdeja J et al, Lancet 2021;398:314-24; Munshi NC et al, N Engl J Med 2021 Feb 25;384(8):705-716. [Slide 2 - Pillars of infection prevention in patients receiving T-cell redirecting therapies] Table: Immune defect/risk -> Type of infection -> Preventive strategy. Lymphopenia/Neutropenia -> Bacterial infections/PJP, approx first 3-6 months -> Antiviral + PJP prophylaxis, G-CSF support. Immunosuppressive therapy (steroids, IL-6 blockage, cyclophosphamide) -> Bacterial infections/Sepsis/fungal infections/CMV -> Consider antibacterial/antifungal. B-cell aplasia/Hypogammaglobulinemia -> Bacterial/Viral (Respiratory virus, VZV/HSV)/CMV -> Immunoglobulin replacement, Vaccination. T-cell dysfunction -> Other opportunistic infections, Viral infections, fungal (in selected high risk patients) -> Fixed treatment duration/Schedule optimization. Sources: Rodriguez-Otero P et al, Lancet Oncology May 2024; O'Neill C, et al. eJHaem. 2023;4:811-822; Friedrich MJ, et al. Cancer Cell. 2023;41:711-725.e6; Frerichs KA et al, Blood Advances 2024; Mohan M et al, Blood Cancer Journal (2025) 15:74; Tix T et al, Molecular Therapy (2025)33:7; Richardson T et al, Blood Advances (2025) 9:6; Nath K et al, Blood Cancer Journal (2024) 14:88; Reynolds G et al, Blood advances (2023) 7:19; Banerjee R et al, Blood Advances (2025) 9:18; Dr Rodriguez Otero personal opinion.

Updated CR109234 Data in 4L+ R/R MM: BsAbs - IMS 2026 Trials: CR109234, P1 (N=180), Janssen (Pharmaceutical Companies of Johnson & Johnson) | RedirecTT-1, P1/2 (N=228), Janssen (Pharmaceutical Companies of Johnson & Johnson) Status: CR109234: Recruiting | RedirecTT-1: Active, not recruiting Patient Population: CR109234: Triple-class, 51.8% penta-drug | RedirecTT-1: Triple-class, 65% penta-drug PFS: CR109234: 96.3% @12 mo (RP2D); 84.6% @18 mo (TCR Naïve); 81.9% @24 mo (TCR-naive pts in part 1) RedirecTT-1: 70% @18 mo ORR (Naive Patients): CR109234: 93.60% | RedirecTT-1: 80% (All patients) CR/sCR (Naive): CR109234: 76.60% | RedirecTT-1: 52% Grade ≥3 Infections: CR109234: 25% | RedirecTT-1: 64% CRS (Any Grade): CR109234: 39.30% | RedirecTT-1: 79% (2% Grade 3) ICANS: CR109234: Not reported | RedirecTT-1: 3% Hypogammaglobulinemia: CR109234: 52.80% | RedirecTT-1: 57% (non-IgG) Fatal Infections: CR109234: 2 patients | RedirecTT-1: 12% of patients Follow-up: CR109234: 19.4 mo (naive group); 23.5 mo (Part 1) | RedirecTT-1: 20.3 mo Adapted from a post by Dr. Al-Ola A. Abdallah, MD, originally published on X on August 9, 2025. LARVOL

Phase 3 evidence has moved inmune therapy into early relapse [Table: STUDY | COMPARISON | forest plot (FAVOURS EXPERIMENTAL / FAVOURS CONTROL) | PFS HR | OS HR] CARTITUDE-4 | Cilta-cel vs DPd/PVd | PFS HR 0.29 | OS HR 0.55 DREAMM-7 | BVd vs DVd | PFS HR 0.41 | OS HR 0.58 DREAMM-8 | BPd vs PVd | PFS HR 0.52 | OS HR 0.77 MajesTEC-3 | Tec-Dara vs DPd/DVd | PFS HR 0.17 | OS HR 0.46 MonumenTAL-3 | Tal-DP vs DPd | PFS HR 0.28 | OS HR 0.47 (MonumenTAL-3) | Tal-D vs DPd | PFS HR 0.33 | OS HR 0.51 MajesTEC-9 | Teclistamab vs PVd/Kd | PFS HR 0.29 | OS HR 0.60 MonumenTAL-6 | Tal-Tec vs EPd/PVd | PFS HR 0.11 | OS HR NR (MonumenTAL-6) | Tal-Pom vs EPd/PVd | PFS HR 0.27 | OS HR NR CERVINO | Etentamig vs investigator's choice | PFS HR 0.40 | OS HR NR RANDOMIZED PHASE 3 TRIALS WITHOUT PUBLIC NUMERICAL HRs MagnetisMM-5 | Elranatamab +/- Dara vs DPd | Positive top-line result; no numerical data reported [Forest plot x-axis tick labels largely illegible]

23rd Annual MEETING & EXPOSITION - International Myeloma Society THURSDAY, SEPTEMBER 24, 2026 | 11:00-12:00 Abstract Session 3 Co-chairs: Meral Beksac and Martin Kortum - Double hit ultra-high risk myeloma treated with Isa-VRD plus cyclophosphamide induction and Isa-VRD consolidation: QoL analysis of the UKMRA RADAR trial in newly diagnosed transplant eligible patients - Karthik Ramasamy - Updated safety and efficacy of ramantamig (Ram; JNJ-5322) at the recommended phase 2 dose (RP2D), demonstrating feasibility of outpatient dosing (OPD) in relapsed/refractory multiple myeloma (RRMM) - Jeffrey Matous - DREAMM-9 Patient-Reported Vision-Related Function with Extended Belantamab Mafodotin Dosing Intervals with Bortezomib, Lenalidomide, Dexamethasone (BVRd) in Transplant Ineligible Multiple Myeloma - Hang Quach - A fixed-duration, modified cevostamab dosing regimen induces durable remissions that continue after completion in BCMA-naive pts with relapsed/refractory multiple myeloma (RRMM): CAMMA 1 Arm A results - Amrita Krishnan - Overall Survival, Progression and Non-Relapse Mortality With Teclistamab Plus Daratumumab (Tec-Dara) vs Dara-Based Triplets in Relapsed/Refractory Multiple Myeloma (RRMM): MajesTEC-3 Post Hoc Analysis - Luciano Costa Panelist photos shown: Dr. Beksac, Dr. Kortum, Dr. Matous, Dr. Ramasamy, Dr. Costa, Dr. Krishnan, Dr. Quach 23rd Annual MEETING & EXPOSITION - International Myeloma Society THURSDAY, SEPTEMBER 24, 2026 | 11:00-12:00 Abstract Session 4 Co-chairs: Lugui Qiu and Niels van de Donk - Augmenting TRAIL-dependent apoptosis enables BCMA CAR-T cells to overcome resistance in multiple myeloma - Mansour Poorebrahim - First-in-human phase Ia/Ib study of QLS32015 (GPRC5DxCD3) in patients (pts) with relapsed/refractory multiple myeloma (RRMM): updated safety and efficacy data - Jian Cui - An AI-Designed BCMA/FcRL5 Targeting TriTE Overcomes Antigen Escape and Outperforms BCMA-Directed Therapy in Multiple Myeloma - Jessica Encinas Mayoral - A CD4-Dominant, T-Helper-Skewed T Cell State Underlies Movement and Neurocognitive Toxicity After BCMA-Directed CAR-T Therapy in Multiple Myeloma - Theresia Akhlaghi - Single-cell antigen:TCR mapping reveals tumor-specific effector T-cell responses in multiple myeloma - David Avigan Panelist photos shown: Dr. Akhlaghi, Dr. Avigan, Dr. Cui, Dr. Encinas Mayoral, Dr. Poorebrahim, Dr. Qiu, Dr. van de Donk

Photo of a person standing in front of poster PA-104 (partially occluding it), with poster PA-105 partially visible to the right. PA-104 title: "Comparative Analysis of Isa SC VRd (ISASOCUT, Once-Weekly Bortezomib) and Isa IV VRd (IMROZ, Twice-Weekly Bortezomib): 20-Month Follow-Up in Transplant-Ineligible Newly Diagnosed Multiple Myeloma Patients" Authors (partial): Arthur Bobin, Mohamad Mohty, Philippe Moreau, Hartmut Goldschmidt, Meletios Athanasios Dimopoulos, Robert Z. Orlowski, Cyrille Touzeau, Aurore Perrot, Thomas Chalopin, Mamoun Dib, Lydia Montes, Lionel Karlin, Umer Khan, Julien Bonnet, Rick Zhang, Ercem Kodas, Florence Suzan, Thierry Facon, Xavier Leleu Legible bullet points (partially visible under poster header): isatuximab (Isa) plus bortezomib (V), lenalidomide/dexamethasone (d) is approved for transplant-ineligible newly diagnosed multiple myeloma based on the phase 3 IMROZ trial; subcutaneous (SC) Isa delivered via the novel on-body injector (OBI) has been previously tested in other therapies; a study (ISASOCUT) demonstrated Isa SC delivered via the OBI; major differences between the IMROZ and ISASOCUT protocols; Isa delivered once weekly (QW) for 12 months from cycle 2 in ISASOCUT vs twice weekly (BIW) for 6 months in IMROZ; both switch to monthly administration of Isa (12 months in ISASOCUT, 18 months in IMROZ); peripheral neuropathy (PN), a key dose-limiting toxicity, dose reduction/discontinuation; an updated comparative analysis on the safety and efficacy of Isa SC (ISASOCUT; QW V) vs Isa IV (IMROZ; BIW V) at 20 months of follow-up. FIGURE 3 "Efficacy in the Isa SC OBI (QW V) and Isa IV (BIW V) cohorts" - bar chart comparing Isa SC OBI + VRd vs Isa IV + VRd across categories including >=VGPR, MRD-, and sustained MRD- (>=12 months); approximate percentage labels visible (roughly 80-90% range for >=VGPR, 30-50% range for MRD- categories) but exact values not confidently legible due to image resolution/blur. FIGURE 4 "Progression-free survival in Isa SC OBI (QW V) and Isa IV (BIW V) cohorts" - Kaplan-Meier curve with a small inset table showing "PFS per investigator at 24 months" comparing the two cohorts (percentages not legibly readable). Safety summary bullets (partially legible): Grade >=3 non-hematologic TEAEs occurred at a lower rate with Isa SC OBI (62.2%, QW V) than with Isa IV (73.8%, BIW V) (Table 2); the incidence of serious TEAEs was similar with Isa SC OBI (51.4%, QW V) and Isa IV (52.5%, BIW V); infusion-related reactions (IRRs) were less common in Isa SC OBI (5.4%, QW V) patients vs Isa IV (22.4%, BIW V); a lower incidence of grade >=2 peripheral neuropathy was observed with Isa SC OBI (29.7%, QW V) compared with Isa IV (40.7%, BIW V), but the incidence of grade >=3 peripheral neuropathy was similar; Isa SC OBI (21.6%, QW V) patients exhibited a lower incidence of peripheral neuropathy leading to V dose reduction vs Isa IV (33.5%, BIW V) patients, and a lower incidence of peripheral neuropathy leading to V discontinuation (Isa SC OBI: 5.4%; Isa IV: 9.5%). TABLE 2 "Safety summary" - comparing Isa SC OBI x VRd (N=74) vs Isa IV x VRd (N=265), row labels visible including "Treatment emergent adverse events, TEAEs", "Any TEAE", "Grade >=3 TEAEs", "Serious TEAEs", "Fatal (Grade 5) TEAEs during the treatment period", "Any TEAE leading to definitive discontinuation of all treatments", "Any TEAE leading to peripheral neuropathy", "Any TEAE leading to bortezomib dose reduction" - the numeric values in the table cells are too small/blurry to read confidently. Key Conclusions (partially legible): This comparative analysis at 20 months of follow-up found greater tolerability with Isa SC OBI plus VRd (QW V) compared with Isa IV plus VRd (BIW V) in patients with TI-NDMM; there was a lower incidence of peripheral neuropathy (including events leading to V dose reduction and discontinuation) and infusion-related reactions. Neighboring poster PA-105 (partially visible, different study): Authors "Gulibadanmu Aizezi, Jianli Xu, Gang Chen, Chun xia Han..." - Department of Hematology, the First Affiliated Hospital of Xinjiang Medical [University, cut off]. Legible text fragments: INTRODUCTION - "Multiple myeloma (MM) is a malignant hematologic tumor characterized by clonal abnormal proliferation of bone marrow plasma cells, accounting for 10%-15% of all hematologic malignancies. Standard first-line therapies-including proteasome inhibitors, immunomodulators, and novel CAR-T cell therapies-have significantly improved treatment efficacy; however, MM remains incurable. Enhancing remission rates and improving patient PFS and OS remain critical clinical challenges. Compared to traditional chemotherapy and newer drug regimens, patients receiving autologous stem cell transplantation (ASCT) exhibit superior PFS and OS outcomes." AIM (partially cut off): "To evaluate the imp[act of] cell transplantation [on] outcomes by compar[ing] characteristics and [...] eligible multiple my[eloma] who underwent ASC[T] did not during the [...] thereby informing [...] decision-making." RESULTS: "All 73 ASCT recipients achieved successful engraftment without transplant-related mortality. After a median follow-up of 37.3 months, ASCT was associated with significantly prolonged PFS compared with non-ASCT treatment (72.6 vs 43.9 months, P<0.041), while OS showed a numerical but non-significant improvement (89.7% vs 54.8%, P=0.637). In high-risk patients, ASCT improved PFS (76.7 vs 53.1 months) and yielded higher OS rates (93.3% vs 48.9%), although differences were not statistically significant. A separate subgroup analysis demonstrated significantly longer PFS with ASCT [...]" IMS 2026 conference banner also visible: "23rd Annual Meeting & Exposition, September 23-26, 2026, Glasgow, Scotland."

Updated NCT04751877 Data in 1L MM: anti-CD38 mAbs - IMS 2026 Trials: Perseus (EMN) | IMROZ (sanofi) | NCT04751877 (CHU Poitiers) | CEPHEUS (Janssen, Pharmaceutical Companies of Johnson & Johnson) Treatment Regimen: Perseus: Daratumumab + Bortezomib + Lenalidomide + Dexamethasone vs Bortezomib + Lenalidomide + Dexamethasone IMROZ: Isatuximab + Bortezomib + Lenalidomide + Dexamethasone vs Bortezomib + Lenalidomide + Dexamethasone NCT04751877: Isatuximab + Lenalidomide + Dexamethasone +/- Bortezomib vs Isatuximab + Lenalidomide + Dexamethasone CEPHEUS: Isatuximab + Bortezomib + Lenalidomide + Dexamethasone vs Bortezomib + Lenalidomide + Dexamethasone Median Follow-Up: Perseus: 47.5 months | IMROZ: 59.7 months | NCT04751877: 48.9 months | CEPHEUS: 76.0 months PFS: Perseus: 48 mo: 84.3% vs 67.7% IMROZ: 60 mo: 63.2% vs 45.2% NCT04751877: 18 mo: 87.2% vs 85.5%; 24 mo: 85.2% vs 80%; 36 mo: 78% vs 72.3%; 48 mo: 74% vs 65.7% CEPHEUS: 54 mo: 68.1% vs 49.5%; 72 mo: 59.3% vs 38.3% ORR%: Perseus: 96.6% vs 93.8% | IMROZ: 91.3% vs 92.3% | NCT04751877: 85% vs 78% | CEPHEUS: 97% vs 93% MRD negativity rate: Perseus: 75.2% vs 47.5% | IMROZ: 58.1% vs 43.6% | NCT04751877: 12 mo: 51% vs 21% | CEPHEUS: 61.1% vs 40% ≥CR rate: Perseus: 87.9% vs 70.1% | IMROZ: 74.7% vs 64.1% | NCT04751877: 58% vs 31% | CEPHEUS: 80.6% vs 61.4% Adapted from Dr. Akhil Santhosh (X 2025) LARVOL

[Main poster, PA-507] International Myeloma Society - 23rd Annual MEETING & EXPOSITION - September 23-26, 2026 . Glasgow, Scotland Treatment Patterns and Outcomes in Primary Plasma Cell Leukemia: A Single-Center Case Series Rahul Pottabathini 1, Noffar Bar 2, Elan Gorshein 2, Tara Anderson 2, Erica Stevens 2, Sabrina L. Browning 2, Natalia Neparidze 2, and Terri L. Parker 2 1 UConn - Waterbury Internal Medicine Residency Program, Waterbury, Connecticut, USA 2 Division of Medical Oncology and Hematology, Department of Internal Medicine, Yale University School of Medicine, New Haven, CT PA-507 INTRODUCTION - Primary plasma cell leukemia (pPCL) is an aggressive plasma cell neoplasm associated with poor outcomes. pPCL patients are excluded from randomized trials, and optimal treatment sequencing remains unclear[1]. - Proteasome inhibitor - immunomodulatory drug (PI-IMiD) combinations with anti-CD38 antibodies and autologous stem cell transplantation (ASCT) may improve outcomes [2]. - We report a retrospective review of pPCL patients treated with daratumumab-based quadruplet (DQ), ASCT, during induction and maintenance of pPCL. METHODS - We conducted a retrospective study of 13 patients diagnosed with pPCL per IMWG criteria at our institution from 2014 to 2026. - Baseline demographics, laboratory data, cytogenetics, Revised International Staging System (R-ISS), treatment regimens, best response per IMWG response criteria, ASCT status, and survival outcomes were studied. - High risk cytogenetics defined as presence of any one of the abnormality: del(17p)/TP53 mutation; t(4;14), t(14;16), t(14,20)-with 1q+/del(1p32). - The data cutoff for analysis was July 2026. RESULTS - The median age at diagnosis was 64 years (IQR: 57-82), 8 (62%) were female, and 11 (85%) were non-Hispanic. High-risk cytogenetics were seen in 11 (85%) patients. The median LDH was 335.0 U/L (IQR: 103-2033) and median beta-2 microglobulin was 11.0 mg/L (IQR: 2-41). R-ISS staging was III in 72%. ECOG PS status of 0-1 in 62% patients, and 2 in 31%. Eight patients (62%) underwent ASCT. - A cytotoxic regimen (CyBorD/VTD-PACE) was used initially if patients exhibited acute organ failure at presentation, followed by daratumumab-based combinations (Dara-KRd or Dara-VRd). - Among 13 patients, 11 (85%) received DQ therapy during their treatment course. 4 (31%) patients received one cycle of cytotoxic chemotherapy followed by transition to DQ as frontline therapy, 3 (23%) patients received upfront DQ, and 4 (31%) received as maintenance following ASCT. - The overall response rate was 82% of patients (n=9). Of responders, 2 (18.2%) had a complete response, 6 (54.5%) had VGPR, and 1 (9.1%) had PR. Four patients (31%) had relapsed disease, of whom one received talquetamab as a bridge to CAR-T cell therapy. - At a median follow-up of 15 months (IQR: 4-42), the median observed PFS was 8 months (IQR: 3-42), and the median OS was not reached. 4 (31%) patients had early progression and succumbed to the disease. Figure 1: Response Rates following DQ Induction - Overall Response (N=11): CR 18.2%, VGPR 54.5%, PR 9.1% Table 1. Baseline demographic and clinical characteristics (Data Cutoff: 7/1/26, N=13) Age, years median (range): 64.0 (57-82) Sex, n (%): Male 5 (38), Female 8 (62) Race, n (%): White 10 (77), Black 2 (15) Ethnicity, n (%): Non-Hispanic 11 (85), Hispanic 2 (15) Median circulating plasma cells, n (range): 30.0 (7-85) Extramedullary disease present, n (%): Yes 6 (46), No 7 (54) Median LDH (U/L), (range): 335.0 (103-2033) Median Beta-2 microglobulin (mg/L), (range): 11.0 (2-41) R-ISS staging, n (%): Stage II 1 (8), Stage III 10 (77), Unknown 2 (15) Cytogenetics, n (%): High risk 11 (85), Standard risk 1 (7), Unknown 1 (7) ECOG performance status, n (%): 0-1 8 (62), 2 4 (31), 3 1 (8) ECOG: Eastern Cooperative Oncology Group CONCLUSIONS - In this retrospective review of pPCL, DQ regimens were universally incorporated into therapy. ASCT was performed in the majority of patients. - Our study adds to the growing body of real-world evidence supporting anti-CD38-based strategies with ASCT consolidation in pPCL. REFERENCES 1. Katodritou E, Terpos E, Delimpasi S, et al. Real-world data on prognosis and outcome of primary plasma cell leukemia in the era of novel agents: a multicenter national study by the Greek Myeloma Study Group. Blood Cancer J. 2018;8(3):31. 2. Katodritou E, Kastritis E, Dalampira D, et al. Improved survival of patients with primary plasma cell leukemia with VRd or daratumumab-based quadruplets: a multicenter study by the Greek Myeloma Study Group. Am J Hematol. 2023;98(5):730-738. ACKNOWLEDGMENTS Noffar Bar has research funding from Johnson & Johnson; Sabrina L. Browning has research funding from Johnson & Johnson; Natalia Neparidze has research funding from Johnson & Johnson, GSK, and consults for Cel[illegible]. The remainder of the authors have no conflicts to disclose. CONTACT INFORMATION Terri Parker, M.D., Yale School of Medicine, New Haven, CT, USA, terri.parker@yale.edu Footer: PA-507 - Plasma Cell Precursor and Other Disorders - Rahul Pottabathini - International Myeloma Society - PosterSessionOnline [Adjacent poster, left, partially visible] [Ou]tcomes in Smoldering Multiple Myeloma - [Up]dated Results From the SPARK Study Body text mostly too small/angled to read: SPARK is a real-world, retrospective chart review study involving hematology centers across Europe ... 408 patients were included from 27 sites across 5 countries (Italy, Spain, and France mentioned) ... median follow-up was 59.5 months ... risk classification per AQUILA study criteria, Mayo 20-2-20, and IMWG 2020 ... [illegible]

Methods - SPARK is a real-world, retrospective chart review study involving hematology centers across Europe (Figure 1) - Eligible patients were adults (>=18 years) diagnosed with SMM - Data were collected from all patients who were diagnosed with SMM between January 1, 2016, and December 31, 2021, with follow-up from the date of diagnosis until December 31, 2023, death, or loss to follow-up, whichever came first - For patients who withdrew consent, all data entered into the electronic case report form up until withdrawal were considered for analysis - High-risk SMM definitions per AQUILA study criteria, Mayo 20-2-20, and IMWG scoring system have been presented previously for SPARK5 - Here, we present updated results with extended follow-up, including all available data collected up to the cut-off date of March 2026 Figure 6: Patients with high-risk SMM who progressed to MM Bar chart, Y axis "Patients, %" (0-100), grouped by risk classification, bars = CRAB (red) vs SLiM only (black): AQUILA HR (n=80): CRAB 57.5%, SLiM only 16.3% Mayo 20-2-20 HR (n=50): CRAB 62.0%, SLiM only 14.0% IMWG 2020 HR (n=25): CRAB 60.0%, SLiM only 24.0% Any-model HR (n=72): CRAB 62.5%, SLiM only 12.5% aIncluded patients with CRAB only and patients with both CRAB and SLiM. Table 3: Incidence of organ damage in patients who progressed to MM Columns: AQUILA (HR n=194, Non-HR n=25, UC n=189), Mayo 20-2-20 (HR n=86, Non-HR n=216, UC n=106), IMWG 2020 (HR n=43, Non-HR n=217, UC n=148), Any model (HR n=173, Non-HR n=168, UC n=67), Total N=408 Progressed to MM, n: AQUILA 80/3/62; Mayo 50/55/40; IMWG 25/55/65; Any model 72/45/27; Total 145 Progressed to MM and had organ damage at the time of MM, n (%): AQUILA: HR 52(65.0), Non-HR 1(33.3), UC 33(53.2) Mayo 20-2-20: HR 32(64.0), Non-HR 32(58.2), UC 22(55.0) IMWG 2020: HR 14(56.0), Non-HR 35(63.6), UC 36(56.9) Any model: HR 46(63.9), Non-HR 27(60.0), UC 13(46.4) Total: 86(59.3) Type of organ damage, %: Bone pain - AQUILA: 38.5/100/24.2; Mayo: 28.1/37.5/36.4; IMWG: 35.7/37.1/29.7; Any model: 32.6/37.0/30.8; Total: 33.7 Bone fracture - AQUILA: 25.0/0/27.3; Mayo: 25.0/27.3/27.3; IMWG: 21.4/22.9/29.7; Any model: 21.7/22.2/46.2; Total: 25.6 Renal failure - AQUILA: 5.8/0/6.1; Mayo: 6.3/3.1/9.1; IMWG: 0/2.9/10.8; Any model: 8.7/0/7.7; Total: 5.8 Hypercalcemia - AQUILA: 5.8/0/3.0; Mayo: 3.1/0/13.6; IMWG: 0/0/10.8; Any model: 4.3/0/15.4; Total: 4.7 Anemia - AQUILA: 36.5/0/51.5; Mayo: 40.6/37.5/50.0; IMWG: 50.0/37.1/43.2; Any model: 43.5/53.8/41.9; Total: 41.9 Other - AQUILA: 28.8/0/21.2; Mayo: 21.9/25.0/27.3; IMWG: 21.4/25.7/22.0; Any model: 23.9/33.3/15.4; Total: 25.6 Key Takeaway The findings from this large, real-world study of European patients with SMM reinforce the need for enhanced monitoring strategies and improved risk stratification to identify patients with high-risk SMM that could benefit from early treatment intervention Conclusions - Regardless of how high-risk SMM was defined, >57% of progressions to MM were CRAB events, and >56% of progressions to MM included organ damage, most commonly bone pain, bone fractures, and anemia - Patients with high-risk SMM were monitored more frequently than non-high-risk patients, yet the incidence of organ damage at MM diagnosis remained high - These data suggest that active monitoring alone is insufficient in preventing organ damage in the majority of patients with high-risk SMM; highlighting the unmet need for early treatment intervention

[Slide 1 - CONSORT diagram, SeaLAND trial (Selinexor-Lenalidomide vs Lenalidomide maintenance in NDMM), BJHaem] Randomised part: assessed n=175, excluded 41 (inclusion)/8 (declined), randomized n=126. Safety run-in: assessed n=20, excluded 3 (ineligible)/1 (no ASCT), n=16. Allocated to lenalidomide maintenance n=64; allocated to selinexor+lenalidomide maintenance n=78 (62 randomized + 16 safety run-in). Reasons for discontinuation - Lenalidomide arm (n=64): study closure 45, disease progression 13, withdrawal of consent 1, investigator decision 2, adverse event 2, other 1. Sel+Len arm (n=78): study closure 39, disease progression 14, withdrawal of consent 13, investigator decision 6, adverse event 3, intercurrent illness 1, non-compliance 1, other 1. [Slide 2 - Figure 2: Progression-free survival, BJHaem] Panel A - KM curves: Len alone vs Sel+Len as maintenance post-ASCT. Number at risk at months 0/6/12/18/24/30/36/42/48: Len 64/57/44/35/26/19/9/2/0; Sel+Len 78/64/54/43/29/21/13/9/5. Panel B - subgroup forest plot (HR, 95% CI), Sel+Len vs Len alone: Overall 21/78 (26.9%) vs 14/64 (21.9%), HR 1.22 (0.62-2.41). Age <65: 12/52 (23.1%) vs 10/40 (25.0%), HR 0.84 (0.36-1.95). Age >=65: 9/26 (34.6%) vs 4/24 (16.7%), HR 2.33 (0.71-7.68). R-ISS Stage I: 4/24 (16.7%) vs 4/24 (16.7%), HR 1.16 (0.29-4.64). Stage II: 6/33 (18.2%) vs 10/37 (27.0%), HR 0.77 (0.28-2.11). Risk Low: 18/62 (29.0%) vs 9/49 (18.4%), HR 1.67 (0.75-3.71). High: 3/16 (18.8%) vs 5/15 (33.3%), HR 0.47 (0.11-2.01). del17p None: 19/65 (29.2%) vs 13/58 (22.4%), HR 1.22 (0.60-2.46). del17p present: 2/13 (15.4%) vs 1/6 (16.7%), HR 1.07 (0.10-11.89).

International Myeloma Society New 2nd line treatments according to patient characteristics Columns: Tec +Ig (MajesTec 9) | Tec Dara +Ig (MajesTec 3) | Tal Dara (MonumenTal 3) | Tal Dara Pom (MonumenTal 3) | Mezi Kd (Successor 2) | [Press release only:] Tal Pom (MonumenTal 6) | Tec Tal +Ig (MonumenTal 6) | Elra +Ig (MagnetisMM 5) CD38 refractory: Yes | Possibly* | Possibly* | Possibly* | Yes | Yes | Yes | ? High risk: Possibly | Yes | Possibly | Possibly | Yes | Yes | Yes | ? EMD: Possibly | Possibly | Possibly | Possibly | Possibly | Possibly | Yes | ? Frail: Yes | Caution | Possibly | Caution | Yes | Possibly | Yes | ? High Infection risk: Possibly | Caution | Yes | Caution | Possibly | Possibly | Caution | ? Tolerability: Yes | Yes | Possibly | Possibly | Yes | Possibly | Caution | ? [Note: row-to-cell alignment approximate due to photo angle] Yes = Preferred / data supports; Possibly = Conditional / Limited data; Caution = possible but caution required Tec = Teclistamab | Tal = Talquetamab | Elra = Elranatamab | Dara = Daratumumab | Pom = Pomalidomide | Ig = Immunoglobulin | Mezi = mezigdomide | K = carilzomib | *based on TRIMM-2 data Based on trial eligibility + published efficacy/safety signals for BCMA & GPRC5D bispecifics. Table represents personal opinion and based on limited evidence! Costa LJ, et al. N Engl J Med. 2026;394(8):739-752, Touzeau C, et al. N Engl J Med. 2026, Mina R, et al. N Engl J Med. 2026, Dimopoulos et al., The Lancet 2026, Johnson & Johnson. press release, 23 July 2026, Pfizer. press release, 29 April 2026.

Phase 3 evidence has moved inmune therapy into early relapse [Table: STUDY | COMPARISON | forest plot (FAVOURS EXPERIMENTAL / FAVOURS CONTROL) | PFS HR | OS HR] CARTITUDE-4 | Cilta-cel vs DPd/PVd | PFS HR 0.29 | OS HR 0.55 DREAMM-7 | BVd vs DVd | PFS HR 0.41 | OS HR 0.58 DREAMM-8 | BPd vs PVd | PFS HR 0.52 | OS HR 0.77 MajesTEC-3 | Tec-Dara vs DPd/DVd | PFS HR 0.17 | OS HR 0.46 MonumenTAL-3 | Tal-DP vs DPd | PFS HR 0.28 | OS HR 0.47 (MonumenTAL-3) | Tal-D vs DPd | PFS HR 0.33 | OS HR 0.51 MajesTEC-9 | Teclistamab vs PVd/Kd | PFS HR 0.29 | OS HR 0.60 MonumenTAL-6 | Tal-Tec vs EPd/PVd | PFS HR 0.11 | OS HR NR (MonumenTAL-6) | Tal-Pom vs EPd/PVd | PFS HR 0.27 | OS HR NR CERVINO | Etentamig vs investigator's choice | PFS HR 0.40 | OS HR NR RANDOMIZED PHASE 3 TRIALS WITHOUT PUBLIC NUMERICAL HRs MagnetisMM-5 | Elranatamab +/- Dara vs DPd | Positive top-line result; no numerical data reported [Forest plot x-axis tick labels largely illegible]

Mayo Clinic / 23rd International Myeloma Society Annual Meeting, Glasgow, Scotland. September 2026. [Poster Discussion badge] Christoph Schaefers, MD Presents: PA-130 Dasatinib and quercetin in patients with relapsed, refractory multiple myeloma receiving CAR-T therapy: First report from the safety run-in cohort of the phase II DART trial Session: Poster discussion. Room: Alsh, SEC. Thursday, September 24, 12:00-12:50 PM

[Poster PA-279, IMS 23rd Annual Meeting - Lesion-Level PET/CT Analysis Identifies Anatomic and Biologic Predictors of Progression After BCMA CAR T-Cell Therapy in Relapsed/Refractory Multiple Myeloma. A. Dreyfuss et al.] Analyzed 102 consecutive triple-class-exposed R/R myeloma patients treated with cilta-cel 2018-2024. Four PET/CT timepoints: pre-apheresis (80 scans, 307 lesions), pre-LDC (98, 336), day 30 post-CAR T (102, 134), first POD (102, 178). Individual lesions longitudinally followed for metabolic response and POD (CR/PR/SD/PD by SUV). Multivariable Cox regression evaluated associations with PFS and OS. Time from diagnosis to CAR T median 5.9yr (0.6-20.0); age at infusion median 66 (42-86); penta-class exposed 65(65%); high-risk cytogenetics 57(57%); prior anti-BCMA therapy 20(20%). Results: Worse PFS/OS with increased number of radiologic sites and EM/PM involvement pre-LDC. Lesion-level analysis shows higher rate of lesion progression in EM/PM sites vs bony sites. Worse PFS with incomplete radiologic response at day 30; lesion-level analysis shows highest rates of progression in EM sites with stable disease at day 30. Conclusions: Identified lesion-level characteristics associated with dominant resistance niches, including EM/PM sites, pre-existing lesions, and incompletely responding sites of disease. Findings suggest lesion-directed or systemic escalation strategies for select patients may improve CAR T outcomes.

[MILESTONE (MRD-guided ASCT deferral trial) - Cohort A study schema] Induction (DVRd x6 cycles) -> MRD1 assessment -> if MRD1<10^-5: Arm A1 (DVRd x3 cycles consolidation) or Arm A3 -> MRD2 -> Maintenance (DR, yearly MRD, until PD or sustained 12mo MRD<10^-5) -> Sustained MRD<10^-5 leads to MRD-SURE (treatment-free observation and MRD surveillance). If MRD1>=10^-5: Arm A2 (AHCT consolidation) -> MRD2 -> Maintenance -> MRD-SURE. DaraVRd regimen: Daratumumab 1800 SQ days 1,8,15,22 (C1-2; days 1,15 C3-6; day 1 C>6); Bortezomib 1.3mg/m2 Days 1,8,15,22; Lenalidomide 25mg Days 1-21; Dexamethasone 40mg PO Days 1,8,15,22. [Results - MRD flow through treatment phases] Cumulative MRD-neg CR rate <10^-3 85%; <10^-6 55%. Sustained MRD neg <10^-3 is 60% and <10^-6 is 35%. Sankey flow diagram tracking MRD depth categories (Induction -> ASCT -> Consolidation -> Maintenance -> Year 2 -> Ongoing Treatment). 100% MRD trackable; ITT MRD<10^-3 post-induction 25% (N=5); AHCT deferral rate 30%. [35 patients (n=20 deferred ASCT, 35% deferral rate per tweet), IFM 2020/MILESTONE trial] [Immune profiling slide - Higher EM:EMRA CD8 T cells ratio in Blood and Bone marrow is associated with lower MRD] Flow cytometry gating (CD8+T/CD4+T/DNT from T cells; CM/NAV from CD8+T via CCR7/CD45RA; EM/EMRA subsets via CD57/PD1). Heatmap of fold-change (SCN->M, C->M) across immune subsets in Blood and BM. Correlation plots: CD57negPD1negEM CD8T increase from SCN to M (Blood) vs log10(MRD_M+1): r=-0.7854, p=0.0004. Same metric Bone Marrow: r=-0.5268, p=0.0462. CD57negPD1negEM/EMRA CD8T ratio increase (Blood): r=-0.6275, p=0.0086. Same ratio (Bone Marrow): r=-0.7922, p=0.0008. "The exhausted-like population (CD11c-CD16+CD57-PD1+ EMRA-CD8+T) tracks with persistent disease. CD57+PD1- EM-CD8+T cells population is associated with lower MRD." MILESTONE trial N=20, 35% ASCT deferral rate. Immune profiling linking T-cell/monocyte/NK signatures to MRD depth.

[Poster PA-380, IMS 23rd Annual Meeting - Radiographic Response to Bridging Therapy is Predictive of Efficacy Post BCMA CAR T-cell Therapy in Relapsed/Refractory Multiple Myeloma. A. Dreyfuss et al.] Analyzed 102 RRMM triple-class-exposed patients treated with cilta-cel (2018-2024) with serial FDG PET/CT at 3 timepoints: pre-apheresis (80 scans, 307 lesions), pre-lymphodepleting chemo (98, 336 lesions), first progression of disease (102, 178 lesions). Time from diagnosis to CAR T median 5.9yr (0.6-20.0); age at infusion median 66 (42-86); penta-class exposed 65(65%); high-risk cytogenetics 57(57%). Bridging therapy: 50(49%) chemotherapy, 25(25%) debulking alkylator, 4(3.9%) bispecific antibody, 4(3.9%) radiotherapy. Bridging therapy type NOT associated with PFS/OS on Cox regression (p=0.200). PFS/OS by radiologic bridging response (CR/PR vs SD vs PD): significant association with PFS, not with biochemical bridging response alone. Table: Radiologic bridging response, N, Event N, HR (95% CI), p-value: CR/PR N=15 Event=4 (reference); SD N=10 Event=4, HR 2.46 (0.61-9.87); PD N=22 Event=18, HR 4.87 (1.62-14.7), p=0.006 overall. Extramedullary/paramedullary (EM/PM) sites less likely to respond to bridging; poor response to bridging associated with individual site POD. Conclusions: Radiographic bridging response significantly predicted CAR T efficacy, and pre-CAR T PET/CT provided prognostic stratification beyond biochemical response or bridging type. Routine PET/CT during bridging may identify high-risk patients and inform post-infusion salvage/consolidation strategies.

[GMMG-HD10/DSMM-XX/MajesTEC-5, presented by L Rasche at 23rd IMS Annual Meeting] Study design: TE-NDMM patients, aged 18-70, ECOG PS 0-2. Induction (6x28-day cycles): Tal-DR n=20 -> HDT+ASCT -> Maintenance (x18 cycles): Tec+D or Tal+D. Primary endpoints: AEs, SAEs. Secondary: MRD negativity (10^-5 via NGF), ORR, >=CR, >=VGPR, Stem cell yield. Induction dosing table: Tal SC (C1: SUD 0.01/0.06/0.4mg/kg Days 2,4,8 then 0.8mg/kg Day 15; C2: 0.8mg/kg Q4W Day1; C3-6: 0.8mg/kg Q4W Day1). D: 1800mg SC QW(C1)/QW(C2)/Q2W(C3-6). R: 25mg PO QD Days1-21 (C2-6). d: 20mg PO/IV twice weekly (C1-2). Baseline demographics (N=20): Median age 57.5 (45-69), >=65 5(25%), Male 15(75%), White 20(100%), ECOG<=1 18(90%)/2 2(10%), >=60% BMPCs 11(55%), >=1 soft tissue plasmacytoma 1(5%), ISS I 9(45%)/II 5(25%)/III 6(30%), High cytogenetic risk 6(30%). Disposition: Enrolled N=21, 1 did not receive treatment, received >=1 dose n=20, 1 discontinued due to PD during C3, completed induction n=19 (median follow-up 12.5mo, range 3.4-14.9), 19 underwent stem cell mobilization (18 underwent ASCT), 2 discontinued, continued to maintenance n=17. Stem cell mobilization/ASCT: underwent mobilization 19(95%), yield median 6.4 (1.4-54.0) x10^6 CD34/kg, transplanted 18(90%), engraftment neutrophils median 14d (3-67), platelets 14d (4-57). TEAEs (N=20): Any AE 20(100%)/Gr3-4 13(65%). Neutropenia 7(35%)/6(30%). Leukopenia 8(40%)/6(30%). Lymphopenia 5(25%)/5(25%). Thrombocytopenia 4(20%)/2(10%). Anemia 3(15%)/2(10%). Taste changes 16(80%)/-. Infections 14(70%)/4(20%). CRS 14(70%)/0. Hypogammaglobulinemia 11(55%)/1(5%). Rash AE 11(55%)/2(10%). Fatigue 9(45%)/1(5%). Decreased weight 8(40%)/3(15%). Nausea 7(35%)/0. Hypokalemia 7(35%)/0. No ICANS reported. 4 patients (20%) discontinued Tal due to TEAE. 2 deaths (1 trauma, 1 progressive disease). Infections: patients with >=1 TE infection 14(70%)/Gr3-4 4(20%). No opportunistic infections, no fatal infections, no discontinuations due to infection. 17(85%) received >=1 dose of IgG replacement. AEs of clinical interest: Taste changes 16(80%, mostly Gr1); Rash predominantly low grade; 4(20%) nail-related events; 3/20 patients had ataxia/balance disorders, all Gr1-2, all recovered/recovering. Disease burden: 95% of treated patients achieved 100% reduction in paraprotein production by post-ASCT/pre-maintenance visit. Response and MRD: Overall response 90% (18/20) - sCR 55%, CR 15%, VGPR 15%, PR 5% (>=CR 70%, >=VGPR 85%). MRD negativity (10^-5): 90% (18/20). "High rate of deep responses with 90% achieving MRD negativity post ASCT/prior to maintenance." Median follow-up 12.5 months, data cutoff March 2, 2026.

[Title slide - IFM 2022-01, NCT06353022, presented by Cyrille Touzeau] "Teclistamab plus Talquetamab consolidation in newly diagnosed myeloma patients with positive measurable residual disease after DVRd induction: interim analysis of the phase 2 study IFM 2022-01 (NCT06353022)." Authors: Cyrille Touzeau, Aurore Perrot, Gedeon Agbetsivi, Thomas Chalopin, Cyrille Hulin, et al., Philippe Moreau.

[Slide 1 - WGS study sample cohort, Holly Lee/Nizar Bahlis] A. Case summary: Total cohort n=84, split into anti-BCMA only (n=51: TCE n=42 [TECLI n=23, ELRA n=13, ABBV383 n=4, Unknown n=1], IDE-CEL n=7), anti-GPRC5D only (n=23: TCE n=22 [TALQ n=21, FORIM n=1], ARLO-CEL n=1), anti-BCMA+GPRC5D (n=4: ELRA-TALQ n=3, TALQ-TECLI n=1), anti-FCRL5 (n=6: CEVO n=6). B. Sample summary bar chart, pre/post samples by category: anti-BCMA only 27 pre/40 post; anti-GPRC5D only 9 pre/20 post; anti-BCMA+GPRC5D 4 pre/8 post; anti-FCRL5 5 pre/2 post. C. BCMA escape in all relapsed/refractory cases (cohort includes post-relapse n=26 and primary refractory n=13 patients) - per-patient PFS bar chart (months) colored by best response (sCR/CR/VGPR/PR/SD/PD/NA), with BCMA escape (filled circle) vs BCMA WT (open circle) markers; range from MM-35 (IDECEL, 33.8mo, VGPR) down to MM-210 (TECLI, 0.6mo, PD). Red dotted line at 8 months. [Slide 2 - Genomic TNFRSF17 events observed in 81% of cases with acquired resistance] A. N=55 patients treated with anti-BCMA (74 samples): Pre-therapy n=23 samples, escape 2/23 (9%): WT 19, gain 2, mono_del 1, biallelic_del 1. Primary ref (pre) n=7, escape 0/7 (0%): WT 7. Primary ref (post) n=13, escape 3/13 (23%): WT 8, gain 1, cn_LOH 1, mono_del 1, biallelic_del 2. Post-relapse n=31 samples, escape 25/31 (81%): WT 5, gain 1, mono_del_snv 8 (26%), biallelic_del 17 (55%). B. N=18 patients with paired samples (Pre vs Post), tracked by patient ID (MM-02, MM-124, MM-111, MM-209, MM-42, MM-74, MM-75, MM-78, MM-03, MM-10, MM-15, MM-17, MM-35, MM-04, MM-06, MM-112, MM-20, MM-210) - most WT pre-therapy convert to biallelic_del or mono_del_snv post-therapy. [Slide 3 - Copy number losses of TNFRSF17 on chr.16p post anti-BCMA therapy] A. Sample-level copy number heatmap (Pre-BCMA vs Post-BCMA) across chr16p region (0-36Mb), showing widespread copy-number loss (blue=CN1, dark red=CN0, maroon=CN4+) concentrated near the TNFRSF17 locus in Post-BCMA samples vs mostly neutral (grey=CN2) in Pre-BCMA samples. Bottom track: % samples with CN loss (blue) or biallelic loss (red) peaking sharply at ~55-60% right at the TNFRSF17 locus (chr16p13.13). B. Higher-resolution view (11,964,000-11,970,000 bp) across TNFRSF17 exons 1-3, showing single-hit (circle) and multi-hit (triangle) mutation events concentrated in Exon 1 and Exon 2 regions in Post-BCMA samples. [Slide 4 - Recurrent mutational hotspots within BCMA extracellular domain] A. BCMA protein sequence diagram (extracellular/transmembrane/intracellular domains) annotated with recurrent point mutations: p.Cys24Tyr, p.Gln25ter, p.Arg27Pro, p.Arg27Gln, p.Ser30del, p.Ser30Phe, p.Thr32Ser, p.Pro33Ser, p.Pro34del, p.Cys37Phe**, p.Asp15Glu, p.Tyr13Asp**, p.Gln7ter, p.Ser48ter, p.Gly61ter**, p.Ser67Phe (** = mutations also reported by Sudha et al. Blood Adv 2025). B. Lollipop plot of variant type (snv/del/ins/silent/splice site) by position across exon 1 (extracellular), exon 2 (transmembrane), exon 3 (cytoplasmic) domains, with variant clusters at p.Arg27Gly/Ter/Gln/Pro (n=6), p.Ser30del/Phe (n=4), p.Pro34del (n=8).

[Infection management with BsAbs and CAR-T, @paurotero, ClinicaNavarra] [Slide 2 - Infections rates over time in the CAR-T and BsAbs recipients, Real-world analysis]: Cumulative incidence of infection over time (days from treatment start) for BsAb vs CAR-T. At-risk/events tables through day 550+. "No differences in the incidence rate of severe infection in the first 100-days post CAR T or BsAbs. Higher incidence with BsAbs beyond 6-month." Sources: Nath K et al, Blood Cancer Journal (2024) 14:88; Mateos MV et al, LBA ASH 2025; NEJM Dec 2025; Touzeau C et al, NEJM 2026. [Slide 3 - Extended dosing intervals were associated with lower annualized infection rates, even with concomitant immunoglobulin replacement]: Overall annualized infection rates by dosing interval (Weekly/Biweekly/Monthly/Bimonthly) - All-grade: 6.08/3.54/2.93/2.25 infections per patient-year (p<0.001 weekly vs others). Severe: 0.81/0.39/0.29/0.10 (p<0.05). During IVIG supplementation - All-grade: 3.02/3.13/3.22/2.88. Severe: 0.44/0.31/0.32/0.13. Source: Smits F et al, Blood Cancer J. 2026 Feb 21;16(1):26. [Slide 4 - Can immune monitoring help identify patients at increased risk of infection?]: CD27+ B cells, CD27- NK cells, and the CD27-/CD27+ T-cell ratio independently predict infection risk throughout the disease course and can be monitored in peripheral blood. Four KM curves (time to infection probability vs months) each with P<.001: CD27+ B cells % (<0.1% vs >=0.1%), CD27- NK cells % (<2.8% vs >=2.8%), CD27-/CD27+ T cell ratio (<0.37 vs >=0.37), Immune score (<=1 vs >=2 risk factors) - all show higher-risk group reaching ~65-85% infection probability vs lower-risk group plateauing ~45-50%. Source: Zabaleta A et al, Blood April 2026;147(18).

[Slide 1 - Prophylaxis with dexamethasone, BOSS protocol (Bispecific Outpatient Step-Up Dosing)] Treatment: Elranatamab 12mg admin Step-up #1 Day 1 (Mon/Tue), Labs/clearance -> Day 2 Safety call, Daily Dex 12mg -> Elranatamab 32mg Step-up #2 Day 3 (Wed/Thurs) -> Day 4,5,6,7 Safety call Daily Dex 12mg -> Elranatamab 76mg Full dose Day 8 (Mon/Tue) -> Day 9,10 Safety call Daily Dex 12mg -> Discharge from BOSS program, continue usual care. Acute Care Plan: Fever>100.4 (any neurologic symptoms and/or any acute changes) -> PAGE BOSS Team (triage by covering RN/NP, escalate to MD on call) -> During the day: Urgent Add-On Outpatient -> Fever Resolution -> Discharge Home. After Hours/Weekends: Urgent Triage for Inpatient Bed -> No other indication for inpatient stay -> Discharge Home. Labs, ID workup +/- tocilizumab and supportive care. Source: Cirstea D et al., IMS 2026. [Slide 2 - BOSS: outpatient step-up changes the pattern of CRS without requiring routine admission] Sankey diagrams tracking CRS grade through Elranatamab step-up dosing for two BOSS cohorts: (n=26) BOSS cohort and (n=34) iSUD cohort, tracking No CRS / Grade 1 CRS / Grade 2 CRS counts across dosing stages (after SUD2, after SUD3, after Full Dose). BOSS n=26: No CRS(n=23), Grade 1 CRS(n=1), Grade 2 CRS(n=2), converging to No CRS(n=24) and Grade 1 CRS(n=2) after full dose. iSUD n=34: No CRS(n=19), Grade 1 CRS(n=12), Grade 2 CRS(n=3), converging to No CRS(n=34) after full dose. Source: Cirstea D et al., IMS 2026. 23rd International Myeloma Society Annual Meeting.

[Bispecifics create layered immune deficits that compound infection risk - Dr. Noopur Raje] Immunosuppression: T-cell activation, proliferation, & exhaustion. T-cell exhaustion: BsAb-TCEs can [expand regulatory T-cells] (Tregs), which can suppress pro[liferation of] effector T-cells; overall T-cell depletion [impairs] adaptive immunity. Cytopenias: Neutropenia, Lymphopenia, Anemia, Thrombocytopenia. Neutropenia note: MM patients commonly experience neutropenia, or abnormally low levels of white blood cells/neutrophils; this immune deficiency increases vulnerability to infections. Death of Mature B Cells (via BsAb-TCE) and Death of Plasma Cells (Malignant & Benign, via BsAb-TCE) shown as mechanism diagram. Hypogammaglobulinemia (HGG): Normal plasma cells depleted; Low serum IgG levels (immunoparesis). HGG note: MM patients commonly experience this secondary immunodeficiency caused by low serum IgG levels; HGG increases infection risk with encapsulated bacteria. BsAb-TCE = bispecific antibody T-cell engager. Sources: Raje N, et al. Blood Cancer J. 2023 Aug 1;13(1):116. Friedrich MJ, et al. Cancer Cell. 2023;41(4):711-725.e6. Lejeune M, et al. Front Immunol. 2020;11:762.

[Slide 1 - SCOPE-MM: Stratification of CAR-T Outcomes at Pre-Apheresis Evaluation in Multiple Myeloma, IMS26] Point-based weighted SCOPE-MM model: 1 point each for BMPC >=50%, history of EMD, serum LDH above ULN; 2 points each for high pre-apheresis CAR-HEMATOTOX score and prior BCMA exposure. Risk groups (partitioning model, HR/event/total pts): 0 points=low risk (0.61, 98/279, 35%); 1-2 points=intermediate risk (0.87 101/216 27%, 1.2 72/126 16%); >=3 points=high risk (1.8 112/164 20%, 3.8 19/20 2%). SCOPE-MM Low = 0 points, Intermediate = 1-2 points, High = >=3 points. High-Risk CAR-HEMATOTOX >=3 points. Source: Rejeski K, Cordas Dos Santos D, et al. IMS 2025, abstract OA-3. [Slide 2 - Predictors of Parkinsonism: Multivariable Analyses Across Studies] Mayo Clinic cohort (OR, 95% CI): Age >75 8.2 (1.8-34.9); 1-3 prior lines vs >=4 3.4 (0.9-14.5); BMPC >=20% 4.3 (1.0-21.9); Involved FLC >=20mg/dL 4.0 (1.0-19.7); Pre-infusion ferritin >=400microg/L 4.1 (1.0-17.5); ICANS 6.1 (1.2-26.9); IEC-HS 5.7 (0.8-29.4); High cumulative steroid use 10.7 (2.6-53.6); Tocilizumab use >=2 43.8 (2.5-768.4); Post-infusion peak ALC >=3.0x10^9/L 16.9 (3.0-317.2); Day 28 ferritin >=400microg/L 6.2 (1.4-44.4). U.S. MM Immunotherapy Consortium (Odds Ratio, 95% CI, P-value): Peak ALC >3000/uL vs <3000/uL 12.7 (5.1-34.8) P<0.001; Progression/Stable disease vs PR (response to bridging) 9.9 (2-180) P=0.03; No Bridging vs PR 0 (P>0.9); Lines of therapy 4+ vs 1-3 0.35 (0.1-1.1) P=0.0054. ALC threshold table: ALC >3000/uL vs <3000/uL - Parkinsonism 12% vs 1%, P<0.001. Sources: Lim K et al. Blood Cancer J 2026;16(1):18. Sidana et al. ASH 2025, abstract 1034. [Slide 3 - Immune Effector Cell-Associated Enterocolitis (IEC-EC) Incidence and Characterization in Cilta-cel-Treated Patients With RRMM in CARTITUDE Clinical Studies] Of 483 treated patients with RRMM, 6 (1.2%) developed IEC-EC; all 6 achieved complete response or better with cilta-cel. Cases managed with steroids (3/6) and/or immunosuppressants (cyclosporine). No cases of IEC parkinsonism observed; plasma cells largely absent from GI biopsies. Most IEC-EC cases (4/6) resolved, though two patients died from COVID-19/multi-organ failure. Longer persistence of CAR+ T-cell levels observed in IEC-EC patients vs those without GI symptoms (log10 CAR+ T cells 10^6/L, peak ~day 14-21 then IEC-EC group stays elevated through day 313 vs control returning near 0 by day 78-100). Patients with IEC-EC experienced prolonged depletion of IgA vs GI-symptom-free control group at comparable time (IEC-EC/Other GI stay near 0.05 g/L through day 1054, Control group recovers to ~0.27 g/L by day 997). GI biopsy histopathology (H&E, CD3, MUM-1, Camelid stains) showed plasma cell absence and signs of injury, but no epithelial damage associated with CAR-T cells. Source: Lin Y et al. ASCO 2026, abstract 7533. [Slide 4 - IEC-Enterocolitis Can Occur in Patients Without ALC Expansion Above Normal Range] Pie chart of toxicity categories (N=233 total): No Toxicity n=208 (89%); IEC-NP only n=12 (5%); IEC-EC only n=4 (2%); IEC-PKS only n=5 (2%); IEC-PKS+IEC-EC n=3 (1%); IEC-EC+IEC-NP n=2 (1%); IEC-PKS+IEC-NP n=1 (0%). ALC dynamics by group (days from CAR-T infusion) and biomarker box plots (ALC at LD, ALC-peak, AUC) comparing Control / IEC-EC / IEC-EC+DNT groups, with significance markers (*P<0.05, **P<0.01). Univariable/Multivariable analysis (Odds Ratio, P value univariable/multivariable): IMS/IMWG-HR 6.1 (1.2-30) P=0.03 / P=0.009; CAR-HEMATOTOX >=3 4.4 (1.04-17.7) P=0.04 / P=0.55; Tocilizumab >=1 dose 7.8 (1.6-38.8) P=0.006; CRS duration >=3 days 19.8 (2.4-162) P=0.0005 / P=0.009; Delayed Neurotoxicity 12.1 (3.0-48.8) P=0.001 / P=0.006. IEC-EC = IEC-enterocolitis, IEC-NP = IEC-nerve palsies, IEC-PKS = IEC-parkinsonism. Sources: Lim K et al. BCJ 2026. Fortuna et al, BCJ 14:180 (2024). Bar et al, IMS 2024 P-008. Banerjee et al, BCJ 15:112 (2025).

[Mass spectrometry in myeloma - Dr. Noemi Puig, IMS26] [Slide 1 - Serum MS and BM MRD]: 2x2 concordance framework: BM MRD- / Serum MS- = "Concordant deep response". BM MRD- / Serum MS+ = "Residual serum disease" (delayed clearance, especially early/IgG; patchy or extramedullary disease; BM sampling or analytical attribution). BM MRD+ / Serum MS- = "Residual marrow disease" (low- or non-secretory residual clone; serum below detection; trackability or sensitivity limitations). BM MRD+ / Serum MS+ = "Concordant residual disease." Interpretive framework based on paired serum MS/BM NGF datasets, Puig et al., Blood 2024. [Slide 2 - Harmonizing MS with the Current Response Framework]: Current response framework (Baseline -> PR >=50% down -> VGPR >=90% down or SPEP-/IFE+ -> CR IFE-) mapped against MS measurement range (Quantitative MS spanning PR through harmonization zone, then Detectable/non-quantifiable, then MS-). >=90% -> >=99% -> >=99.9% -> deeper response -> MS-. PR deemed "likely transferable"; VGPR/CR zone deemed the "Harmonization Zone" requiring defined clinically meaningful response thresholds. [Slide 3 - Conclusions]: 1) MS extends the measurable M-protein range: from low-level disease to deeper response and earlier re-emergence. 2) Clinical interpretation requires consideration of timing, isotype and longitudinal kinetics. 3) MS and BM MRD are often hierarchical rather than complementary, with BM MRD usually being more informative. 4) Integration into current response assessment requires harmonization and clinical validation.

[LMIC Breakfast Meeting - IMS LMIC Myeloma Survey 2026, 190 respondents, 32 countries] Q34 - "Across all pillars, rank the barriers to improving myeloma outcomes." % ranking each barrier 1st or 2nd (n=130): Drug access/affordability (mean rank 1.9) 80%; Diagnostics access (2.8) 54.6%; Clinical trial access (3.9) 26.9%; Transplant access (3.7) 24.6%; Data/registries (4.5) 10%; Supportive/palliative care (5.2) 2.3%; Workforce/training (6.1) 1.5%. Four ways to close the gap: Early Diagnosis - Shorten the path to the laboratory (public-sector diagnosis takes over two weeks for 56%, yet assays return in days; fast-track referral and marrow scheduling, track time to diagnosis). Risk Stratification - Regionalise cytogenetics (FISH routine for 23%, takes more than two weeks in half; hub-and-spoke FISH referral, turnaround target under 14 days). Implementation - Tier the recommendations (70% find guidelines only partly implementable; resource-stratified pathway, VRd backbone for all, pooled procurement for daratumumab/carfilzomib/pomalidomide). Measurement - Measure what is delivered (only 26% keep a structured database; add time from diagnosis to first dose; 114 respondents volunteered for follow-up LMIC network).

[Gut microbiome and CAR-T outcomes - Melody Smith, Stanford Medicine, Summary] Exposure to anaerobic-targeting antibiotics peri-CAR T-cell therapy is associated with shorter PFS and OS and a higher incidence of ICANS (data not shown). These data support antibiotic stewardship in CAR-T patients. Clinical cohorts demonstrate that specific bacterial species may enhance CAR-T outcomes. The intestinal microbiome has been shown to affect outcomes in recipients of CD19- and BCMA-targeted CAR T-cell therapy. However, preclinical models are needed to understand how the microbiome modulates clinical outcomes, including toxicity and efficacy.

Poster PA-365, IMS 23rd Annual Meeting & Exposition, September 23-26, 2026, Glasgow, Scotland Title: PROGNOSTIC IMPACT OF [18F] FDG PET/CT BEFORE AND AFTER BCMA-DIRECTED CAR T-CELL THERAPY IN RELAPSED/REFRACTORY MULTIPLE MYELOMA: A MULTICENTER ANALYSIS Authors: L-E. Tamariz-Amador; M. Talarico; M. Romera; A. Alfonso Pierola; ML. Fuente-Iglesias; JC. Ponce-Jarquin; M-T. San Miguel-Arregui; A. Urrutia; A. Valcarcel; C. Martin; S. Morales; M. Panizo-Inoges; M. Marcos-Jubilar; S. Huerga; S. Villar-Fernandez; J. Rifon; R. Lecumberri; F. Prosper; J. San-Miguel; J. Arbizu; L. Garcia-Belaustegui; C. Nanni; M-J. Garcia-Velloso; E. Zamagni; P. Rodriguez-Otero Affiliations: 1) Cancer Center Clinica Universidad de Navarra, Pamplona, Spain; 2) IRCCS Azienda Ospedaliero-Universitaria di Bologna, Istituto di Ematologia 'Seragnoli,' Bologna, Italy; 3) Dipartimento di Scienze Mediche e Chirurgiche, Universita di Bologna, Bologna, Italy; 4) Nuclear Medicine, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy; 5) Hospital Universitario de Navarra, Pamplona, Spain INTRODUCTION: Depth of response is one of the most relevant prognostic markers in multiple myeloma (MM), as highlighted by minimal residual disease (MRD) data. Functional imaging such as [18F] FDG PET/CT has gained importance for assessing disease burden and response. However, its prognostic value in relapsed/refractory MM (RRMM) treated with BCMA-directed CAR T-cell therapy remains unclear. AIM: Evaluate the prognostic utility of baseline and post-treatment [18F] FDG-PET/CT in relapsed/refractory MM (RRMM) treated with BCMA-directed CAR T-cell therapy. METHOD: We analyzed patients treated with BCMA-targeted CAR T-cell therapy (April 2018-January 2024) at Clinica Universidad de Navarra (Spain) and IRCCS Azienda Ospedaliero-Universitaria di Bologna (Italy). PET/CT scans at baseline, month 1, and month 3 post-infusion were assessed for focal lesions (FLs), paramedullary (PMD) and extramedullary disease (EMD), alongside metabolic tumor volume (MTV) and total lesion glycolysis (TLG). Response was evaluated using Deauville criteria. Optimal cut-off values with the highest discriminative prognostic value for baseline and month 1 MTV and TLG were determined using the maxstat algorithm (categorized as high vs. low b-MTV, b-TLG, MTV-1, and TLG-1). RESULTS: In total 75 patients received BCMA-targeted CAR T-cell therapy (61 Spanish & 14 Italian). ORR was 85%, with 45% achieving >=CR, and 57% MRD negativity. Median PFS and OS for the entire cohort was 12 (95% CI, 8.8-15.2) and 24 months (95% CI, 15.9-32.3), respectively. A positive baseline scan was associated with inferior PFS (11.5 months [95% CI, 7.8-15.3]) compared to having a negative study (18.9 months [95% CI, 12.5-25.4]) (p=0.03). Baseline EMD predicted shorter PFS (4.6 vs. 14.1 [95% CI, 10.4-17.7]; p=0.014). Baseline high MTV (>=810 mL) and TLG (>=1640 g) predicted inferior PFS (5.6 vs 15.6 months; HR 1.9; p=0.03) and OS (11.4 months vs. not reached; HR 3.0; p=0.002). At month 1, persistent hypermetabolic PMD trended toward shorter PFS (7.1 vs 16.0 months; HR 1.8; p=0.058), and significantly shorter OS: 11.7 vs 42.3 months (HR 2.4; p=0.02). Positive FLs in the absence of PMD and EMD had no significant impact on PFS (p=0.32). At month 3, persistent hypermetabolic PMD was associated with shorter PFS (6.9 months [95% CI, 0.0-14.9] vs 16 [95% CI, 10.8-21.2]; HR 3.6; p=0.006) and OS (7.8 months [95% CI, 6.8-8.8] vs 34.2 [95% CI, 10.9-57.1]; HR 3.7; p=0.01). Deauville score 4 FLs trended toward shorter PFS (8 vs 17 months) compared to Deauville score <4 Table: Total CAR T-cell infusions (n=75) - Age, median(range),years: 58(40-79). Number of previous lines: 1-3: 47(62.7%); >3: 28(37.3%). Triple-class refractory: Yes 51(68.0%); No 24(32.0%). Cytogenetics: Standard-risk 33(44.0%); High-risk 33(44.0%); Not available 9(12.0%). Paramedullary disease: Positive 41(54.6%); Negative 34(45.4%). Extramedullary disease: Positive 66(88.0%); Negative 9(12.0%) Figure 1 caption: Flow of patients according to PET scan results and clinical outcomes over time (Sankey diagram: Positive baseline PET 63 / Negative baseline PET 12 -> Positive month1 PET 41 / Negative month1 PET 24 -> Positive month3 PET 21 / Negative month3 PET 37, plus No PET at month3 assessment 10 and Death or progression during first 3 months) Figure 2 caption: Kaplan-Meier curves for PFS and OS by Baseline EMD, stratified MTV (>=810 mL) and stratified TLG (>=1640 g). Panels illustrate survival outcomes as follows: (A) PFS for patients with baseline EMD, (B) OS for patients with baseline EMD, (C) PFS stratified by MTV at baseline, (D) OS stratified by baseline MTV, (E) PFS stratified according to TLG at baseline, and (F) OS stratified by baseline TLG. Figure 3 caption: Kaplan-Meier curves for PFS and OS by PMD 3 months after therapy. Panels illustrate survival outcomes as follows: (A) PFS for patients with PMD 3 months after therapy and (B) OS for patients with PMD 3 months after therapy. CONCLUSIONS: Baseline tumor burden by PET/CT, including high MTV/TLG and EMD, was associated with inferior outcomes after BCMA-directed CAR T-cell therapy. Early assessment at month 1 appeared suboptimal, as persistent FLs uptake without concurrent EMD/PMD was not associated with worse survival. In contrast, month 3 PET/CT provided the strongest prognostic discrimination, supporting this timepoint as the preferred assessment for post-CAR T response evaluation. ACKNOWLEDGEMENT: To all the patients and their families. CONTACT INFORMATION: Luis-Esteban Tamariz-Amador, LTAMARIZ@UNAV.ES, Cancer Center Clinica Universidad de Navarra, Pamplona, Spain. @EstebanTamariz Footer: PA-365 | MRD and Biomarkers | Luis-Esteban Tamariz-Amador

LBA-20 23rd Annual Meeting & Exposition, September 23-26, 2026, Glasgow, Scotland Second CAR T-cell infusion after prior CAR T-cell failure in multiple myeloma Jianlin HU1, Ning AN1, Yuhan BAO1, Qiuxia YU1, Xinran WANG1, Yuying WANG1, Yang GAO1, Shangyun SUN1, Peiling ZHANG1, Di WANG1 and Chunrui LI1* 1 Department of Hematology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China INTRODUCTION: Relapse after CAR T-cell therapy remains a major challenge in relapsed/refractory multiple myeloma (RRMM). Evidence guiding a second CAR T-cell infusion is limited, particularly regarding target selection and the clinical relevance of in vivo expansion. AIM: To evaluate outcomes after a second CAR T-cell infusion and explore associations of target strategy and CAR T-cell expansion kinetics with efficacy and toxicity. METHOD: This retrospective single-center study included 23 patients with RRMM who received a second CAR T-cell infusion after prior CAR T-cell failure between January 2018 and April 2026. CAR T-cell products targeted BCMA, GPRC5D, or both. Responses were assessed according to IMWG criteria. PFS2, OS, and DOR were estimated using Kaplan-Meier methods. Expansion was characterized by Cmax, Tmax, and AUC0-28. Comparative analyses were exploratory. RESULTS: Among 20 response-evaluable patients, the ORR was 80.0%, including CR/sCR in 30.0%. Median DOR was 9.3 months, median PFS2 7.2 months, and median OS 32.5 months. Among 16 patients evaluable after both infusions, 7 (43.8%) achieved an equal or deeper response with the second infusion. CR/sCR occurred in 5 of 12 patients receiving same-target retreatment and none of the 8 who switched targets (P = 0.029), although target strategy was not associated with PFS2 or OS. Cmax and AUC0-28 were higher in responders than in nonresponders (P = 0.037 and 0.046). Higher Cmax was also associated with early grade >=3 thrombocytopenia and anemia (P = 0.024 for both). Later Tmax was associated with a lower risk of progression or death (HR per day, 0.893; P = 0.028). CRS occurred in 91.3%, including grade >=3 CRS in 13.0%. No ICANS or treatment-related deaths were observed. Figure panels: A: Median PFS2: 7.2; 95% CI, 3.4-24.6 (KM curve, Overall 23, at-risk: 23, 8, 5, 2, 1) B: Median OS: 32.5; 95% CI, 12.4-NR (KM curve, Overall 23, at-risk: 23, 10, 6, 4, 1) C: Peak CAR-T expansion (Cmax, copies/ug DNA) >=PR vs <PR, P = 0.037 D: CAR-T exposure (AUC0-28, copies-day/ug DNA) >=PR vs <PR, P = 0.046 E: CAR-T expansion (copies/ug DNA) over time (days 0,7,14,21,28) by >=PR vs <PR F: Treatment-related toxicities (%): Cardiotoxicity 4/23 (17.4%), Hepatotoxicity 6/23 (26.1%), Early grade >=3 thrombocytopenia 19/23 (82.6%), Early grade >=3 neutropenia 22/23 (95.7%), Early grade >=3 anemia 19/23 (82.6%), Grade >=3 CRS 3/23 (13.0%), Infection 17/23 (73.9%) G: Peak CAR-T expansion (Cmax) by early grade >=3 thrombocytopenia/anemia (Yes vs No), P = 0.024 Figure caption: Clinical outcomes, CAR T-cell expansion, and safety after the second infusion. (A-B) PFS2 and OS. (C-E) CAR T-cell expansion according to response (Cmax, n = 19; AUC0-28, n = 18). (F) Treatment-related toxicities. (G) Cmax according to early grade >=3 thrombocytopenia or anemia. Boxes show the median and interquartile range (IQR), whiskers show the range, and dots represent individual patients. Panel E shows median and IQR. CONCLUSIONS: A second CAR T-cell infusion produced meaningful responses in selected patients after prior CAR T-cell failure. Same-target retreatment was associated with a higher CR/sCR rate, but not longer PFS2 or OS. Greater early expansion was associated with both response and severe cytopenias, whereas later Tmax was associated with a lower risk of progression or death. These findings warrant validation in larger prospective studies. ACKNOWLEDGEMENT: Funding: NSFC 82170223 and Hubei NSF 2024AFD421 (Chunrui Li). DISCLOSURES: Nothing to disclose. Corresponding author: cunrui5650@hust.edu.cn

Live audience poll slide: "Which of these tests do you currently perform prior to initiating treatment in candidates for BCMA-targeted CAR-T or T-cell engager therapy?" A. Targeted sequencing of TNFRSF17 - 0% B. Soluble BCMA analysis in peripheral blood - 21% C. Next-generation sequencing (NGS) using a panel that includes TNFRSF17 and GPRC5D - 13% D. Only sequencing - 4% E. None of the above - 62%

Tumor assessment approaches Blood (Peripheral blood tests) - Measures tumor products (e.g., M-protein, free light chains) - Widely available Advantages: - Less invasive - Cheaper - Can be repeated multiple times - Easily accessible Disadvantages: - Indirect measure of tumor burden (rather than the tumor itself) Bone Marrow (Aspiration and biopsy) - Direct assessment of plasma cells - Common site of myeloma involvement Advantages: - Gets directly to the tumor at its common site Disadvantages: - Invasive - Painful - Can be patchy (disease may be missed due to focal involvement) - Expensive Imaging (e.g., whole-body CT, MRI, PET-CT) - Detects bone lesions and extramedullary disease - Provides whole-body assessment Advantages: - Less invasive than bone marrow biopsy - Can assess disease outside the marrow (e.g., focal bone lesions, extramedullary sites) Disadvantages: - Often expensive - May not fully characterize active tumor (e.g., difficulty distinguishing active vs. inactive lesions or treatment-related changes) - Less sensitive for diffuse marrow involvement compared with marrow biopsy International Myeloma Society 23rd International Myeloma Society Annual Meeting 6 Response assessment - Circulating PCs Bar chart (% with cPC+/cPC- by response category): CR: cPC+ 1.3%, cPC- remainder (n cPC+ = 3, n cPC- = 236) PR: cPC+ 12.1%, cPC- remainder (n cPC+ = 42, n cPC- = 305) <PR: cPC+ 54.3%, cPC- remainder (n cPC+ = 19, n cPC- = 16) Survival curve (Survival Probability % vs Time, two lines: cPC+,PR and cPC-,PR): - cPC-,PR (black line): starts ~100%, declines gradually, crosses 50% around ~130 on time axis, ends ~48% - cPC+,PR (green line): starts ~100%, declines faster, crosses 50% around ~50 on time axis, plateaus near ~30% after ~130 X-axis: Time 0-150; Y-axis: Survival Probability 0-100% International Myeloma Society 23rd International Myeloma Society Annual Meeting 23 Bohra et al, submitted Monitoring EMD A. SNVs = 935 Venn diagram of three circles: BM-T, EM-S, and 315 (labeled region) - BM-T only: 60 - EM-S only: 190 - BM-T ∩ EM-S: 186 - BM-T ∩ EM-S ∩ 315 (center): 255 - BM-T ∩ 315: 13 - EM-S ∩ 315: 92 - 315 only: 139 Imaging thumbnail with arrows labeled EM-S -315 and BM-T Right panel: "Fractional abundance of NRAS Q61H in PL, BM and EM-tissue" - line/scatter chart, dual y-axis (Fractional abundance 0-60 left; Serum-Free Light Chains 0-1500 right), x-axis dated Nov 2016 - Oct 2017 Legend: ctDNA (red circle), EM-tissue (blue square), BM-trephine (triangle), SFLC (kappa) (dashed line) Treatment timeline annotations along top: Rd, PAN-Vd, PCd, DLI, DLI, P, KPd-Dara "EM progression" bracket shown across March-October 2017 region Data points show ctDNA and SFLC(kappa) rising over the observation period, with EM-tissue point plotted near end (~October 2017) at high fractional abundance (~48-50) International Myeloma Society 23rd International Myeloma Society Annual Meeting 24 Int. J. Mol. Sci. 2018, 19(7), 1858 Future directions - Blood is increasingly becoming the source for detecting the presence and measurement of clonal plasma cells - Sensitivity for detection of intact cells (MFC, CyTOF, Spectral Flow), monoclonal protein (mass spec), tumor DNA, and other tumor derived products continues to increase - Currently marrow and imaging adds to blood assessments - complementary - Correlations should not be just cross sectional - we need to ask if repeated measurements in blood replace other modalities? - In future is likely that we will be doing less of marrow and imaging - but use blood-based assessments to decide when to effectively use the other modalities International Myeloma Society 23rd International Myeloma Society Annual Meeting 28

Photo of an auditorium screen showing slide "Future directions": - Blood is increasingly becoming the source for detecting the presence and measurement of clonal plasma cells - Sensitivity for detection of intact cells (MFC, CyTOF, Spectral Flow), monoclonal protein (mass spec), tumor DNA, and other tumor derived products continues to increase - Currently marrow and imaging adds to blood assessments - complementary - Correlations should not be just cross sectional - we need to ask if repeated measurements in blood replace other modalities? - In future is likely that we will be doing less of marrow and imaging - but use blood-based assessments to decide when to effectively use the other modalities International Myeloma Society 23rd International Myeloma Society Annual Meeting 28

Detecting/measuring the monoclonal process Central plasma cells branch to: Intact Plasma Cells (Bone Marrow; Plasmacytoma; Peripheral Blood [Flow Cytometry] CD138 vs CD38 plot showing Plasma cells CD38++CD138+ gate); Secreted by PC (Immunoglobulins [Ig]: IgG,IgA,IgM,IgD,IgE; Immunoglobulin Free Light Chains [FLC]: Kappa, Lambda; Lactate Dehydrogenase [LDH]; Cytokines: IL-6,IL-1b,TNF-a,IL-10; Chemokines: CXCL12[SDF-1a], CCL4[MIP-1b], CCL3[MIP-1a], CCL5[RANTES]); Shed by PC (Beta 2 Microglobulin [b2M]; Syndecan-1[CD138] shed ectodomain, shed CD138/syndecan-1 ectodomain with glycosaminoglycan chains; BCMA shed ectodomain, BCMA ectodomain [soluble]); Plasma cell components (Circulating Tumor DNA [ctDNA]; Extracellular Vesicles [EVs]) 23rd International Myeloma Society Annual Meeting, slide 5 Tumor assessment approaches Blood (Peripheral blood tests): Measures tumor products (e.g., M-protein, free light chains); Widely available. Advantages: Less invasive, Cheaper, Can be repeated multiple times, Easily accessible. Disadvantages: Indirect measure of tumor burden (rather than the tumor itself). Bone Marrow (Aspiration and biopsy): Direct assessment of plasma cells; Common site of myeloma involvement. Advantages: Gets directly to the tumor at its common site. Disadvantages: Invasive, Painful, Can be patchy (disease may be missed due to focal involvement), Expensive. Imaging (e.g., whole-body CT, MRI, PET-CT): Detects bone lesions and extramedullary disease; Provides whole-body assessment. Advantages: Less invasive than bone marrow biopsy, Can assess disease outside the marrow (e.g., focal bone lesions, extramedullary sites). Disadvantages: Often expensive, May not fully characterize active tumor (e.g., difficulty distinguishing active vs. inactive lesions or treatment-related changes), Less sensitive for diffuse marrow involvement compared with marrow biopsy. 23rd International Myeloma Society Annual Meeting, slide 6 Response assessment - Circulating PCs Left bar chart (% with cPC+/cPC- by response category): CR: cPC+ 3 (1.3%), cPC- 236 PR: cPC+ 42 (12.1%), cPC- 305 <PR: cPC+ 19 (54.3%), cPC- 16 Right Kaplan-Meier survival curve (Survival Probability vs Time), groups cPC+,PR (green) and cPC-,PR (gray): cPC+,PR line crosses 50% survival around Time~48; cPC-,PR line crosses 50% survival around Time~132; both curves plateau near end (cPC+,PR ~27%; cPC-,PR ~46%) Citation: Bohra et al, submitted 23rd International Myeloma Society Annual Meeting, slide 23

Photo collage of a masterclass talk titled 'Immune Markers of Disease Activity' (presenter Samir Parekh, MD) at the IMS 23rd Annual Meeting, with an audience shot and 5 content panels: Panel 1 - 'SHAJAM: AI Myeloma platform' - workflow diagram: Fragmented clinical data (Clinical notes, Pathology, FISH, Genomics, Imaging, Labs, Transcriptomics) -> SHAJAM (source-linked analysis layer) -> Cohort/Structured patient table -> source-linked, analysis-ready output. Right-side text steps: Open each clinical note by hand -> Semantic search across 9.21M note chunks -> Email the bioinformatics team and wait -> Run real statistics on live rows in chat -> FISH & MRD in PDFs -> Extracted with evidence from notes & pathology -> A cohort question takes weeks -> Built, verified & exported in minutes. (23rd International Myeloma Society Annual Meeting, slide 5) Panel 2 - 'Independent validation by Shajam' - Kaplan-Meier plot: Composite event-free survival by baseline sBCMA threshold (471.75 ng/mL); >471.75 ng/mL (blue) vs <=471.75 ng/mL (red); P<0.01; n=201 teclistamab patients; adjusted hazard ratio of 3.92 with 95% CI; 12-month event-free survival of 69.8% vs [value cut off] threshold. Citation: Lagana and Sh[cut off] Panel 3 - 'aS100A8/A9 mAb salvages CAR-T cytotoxicity' - two box plots (% Cytotoxicity) for HD derived CAR-T Cells and MM PT derived CAR-T Cells, each comparing CAR-T Cells alone vs +S100A8/A9 vs +S100A8/A9+aS100A8/A9 mAb, with ** (p<0.01) significance brackets. HD derived: ~73% vs ~46% vs ~83%. MM PT derived: ~78% vs ~61% vs ~81%. Panel 4 - 'sBCMA guides therapeutic selection' - diagram: High detectable sBCMA -> sBCMA >=400ng/mL -> Target alternate antigen (anti-GPRC5D TCE, anti-BCMA TCE); No detectable sBCMA -> (Copy number loss / Extracellular domain mutation, TNFRSF17 mutation, BCMA antigen escape) -> Target alternate antigen; Detectable sBCMA -> sBCMA <=400ng/mL -> Treat with anti-BCMA bispecific -> T cell immunomodulation -> Tumor lysis. Caption: 'Measuring biomarker for treatment selection'. Small citation text (partially legible): Lee H, Durante et al... soluble BCMA and ... cell factors ... BCMA-targeting T ... myeloma. Blood 202...; PMID ... Panel 5 - 'Conclusions' (with speaker photo): Soluble BCMA <500ng/L is associated with better outcomes. High S100A8/9 (Calprotectin) is associated with inferior outcomes. Higher baseline ALC >0.9k/uL associated with improved outcomes with TCEs. Prospective studies testing a composite index to help select patients are necessary.

sBCMA guides therapeutic selection Three pathway rows, each with illustrated cell diagrams and arrows leading to treatment decision boxes: Row 1 - High detectable sBCMA: High tumor burden (down E:T ratio) -> sBCMA >= 400 ng/mL -> [feeds into] Target alternate antigen box: anti-BCMA TCE and anti-GPRC5D TCE illustration showing BCMA and GPRC5D receptors on tumor cell surface. Row 2 - No detectable sBCMA: TNFRSF17 mutation illustration -> Copy number loss / Extracellular domain mutation box (labeled "BCMA antigen escape") -> [feeds into] Target alternate antigen box (same as above). Row 3 - Detectable sBCMA: sBCMA <= 400ng/mL illustration -> "Treat with anti-BCMA bispecific" -> T cell immunomodulation box: T cells illustration, anti-BCMA TCE, leading to Tumor lysis. Reference (right margin): Lee H, Durante et al.. Impact of soluble BCMA and non-T-cell factors on refractoriness to BCMA-targeting T-cell engagers in multiple myeloma. Blood. 2024 Dec 19;144(25):2637-2651. doi: 10.1182/blood.2024026212. PMID: 39321344; PMCID: PMC11738017. International Myeloma Society Bottom banner: Measuring biomarker for treatment selection S100A8/A9 (Alarmins) effect myeloid derived immune suppression 1. Alarmins are part of the DAMP (Damage Associated Molecular Pattern) family released in response to tissue injury. 2. Secreted by Myeloid cells, mesenchymal cells, Neutrophils in inflammatory states. 3. Tumor cells (HSCs, AML, MM) can also secrete S100A8/A9 4. Transgenic Mouse models of high S100A8/A9 induce MDS (List, JCI 2013) [Diagram]: Bone marrow -> BMC with "Chronically elevated S100A8/A9" arrow into Lymph node -> MDSCs (myeloid-derived suppressor cells) -> branch to T cell responses (inhibited, shown with blocked arrow) and to T helper cell balance: Treg, Treg, TH17 (seesaw balance icon) Abbreviations implied: DAMP, MDS, MDSCs, HSCs, AML, MM Reference: Von Wulffen, et al. Cell Rep 2023 International Myeloma Society S100A8/A9 levels predict response to T cell directed therapy Left: Rapid Progressors N=8, Responding Pts N=15. Timeline: Pre Infusion -> CAR T -> Post Infusion; Peripheral Blood sample collection at Baseline and Month 1. Kaplan-Meier "Probability of Progression" curve (x-axis Days Post CART 0-1500, y-axis 0-100): Responders (blue) vs Rapid Progressors (red). Log Rank p<0.001. Right top: box plot Month_1, y-axis S100A8/A9 pg/mL 0-3000: Rapid Progressors n=7 (median ~750-800, dots up to ~2600) vs Responding Patients n=15 (median ~200-300, mostly under 600, one dot ~1100). Significance marker ** shown between groups. Bottom-left: Drug key Tal (green dot) Tec (purple dot). Diagram: Sample Collection (Baseline) -> Bispecific Antibody -> Sample Collection (During Therapy). Kaplan-Meier "Probability of Progression" curve, x-axis Days 0-900, y-axis 0%-100%: S100 High (n=9, 8 events, red) vs S100 Low (n=10, 6 events, blue). Log Rank p<0.001. Below, same-style curve labeled Log-rank p=0.0137, timepoint = During Bispecific Tx, marks = censored. Bottom-right: box plot "During Bispecific Tx", y-axis S100A8/A9 pg/mL log scale 1,000-100,000, Wilcoxon p=0.015: Rapid Progressors n=7 (lower, mostly 300-1000, Tal=green dots, Tec=purple dots) vs Responding Patients n=15 (higher, up to ~10,000+). International Myeloma Society Bottom banner: Measuring biomarker for treatment selection ALC trajectories after starting BiAb Left: "Class-Specific ALC Trajectories" plot, y-axis Log-Transformed ALC Value -3 to 1, x-axis Days From Start of BsAb Treatment 0-150+. Three classes: A1 (n=46) red, A2 (n=106) blue, A3 (n=170) green, each with predicted (dotted) and observed (solid) lines and confidence intervals. Middle: PFS Kaplan-Meier curves, x-axis Months from Start of BsAb 0-50, y-axis Proportion Progression Free and Alive 0.00-1.00: A2 (blue), A1 (red), A3 (green). p<0.0001. Number at risk table: A2(blue) 106,39,25,16,10,3; A1(red) 46,3,2,1,0,0; A3(green) 168,85,45,24,18,8 at months 0,10,20,30,40,50. Right: OS Kaplan-Meier curves, x-axis Months from Start of BsAb 0-50, y-axis Survival Probability 0.00-1.00: A1(blue in legend order shown as A1,A2,A3 - blue/red/green), p<0.0001. Number at risk table: 106,69,44,33,19,11; 46,18,6,4,1,1; 169,122,79,48,37,21 at months 0,10,20,30,40,50. Summary text: A1 (persistent lymphopenia, PFS 1.1 mos, OS 4.6 mos, ORR 32%) A2 (stable ALC, PFS 6.47 mos, OS 38.1 mos, ORR 67%) A3 (rising ALC, PFS 22.87 mos, OS N/A, ORR 83%) International Myeloma Society

Conclusions Soluble BCMA <500ng/L is associated with better outcomes High S100A8/9 (Calprotectin) is associated with inferior outcomes Higher baseline ALC >0.9k/ul associated with improved outcomes with TCEs Prospective studies testing a composite index to help select patients are necessary

Cell-free DNA (cfDNA) Diagram: blood vessel showing circulating tumour cell, healthy tissue, inflamed tissue, phagocyte, red blood cells, endothelial cells, chromosomes/mutations (XXX), apoptosis/necrosis insets, blood plasma or serum sample containing ctDNA (test tube) Legend: Healthy cell, Phagocyte, Tumour cell, Mutation, Red blood cell, Endothelial cell, Chromosome Advantages: - Allows for molecular profiling where BM samples are sub-optimal - Has the potential to overcome spatial limitations and longitudinally barriers of tumor biopsies Crowley et al., Nature Reviews Clinical Oncology, 2013 Millions of myeloma needles in the cell-free DNA haystack [Image] Haystack labeled with magnet "Cell-free DNA sequencing" pointing to a pile of needles How do we identify them? Cancer features that be detected from cfDNA 1. Somatic Mutations - Detect point mutations, small insertions/deletions, and other sequence variants. [Sanger trace: A G C T T G C T G] Applications: Diagnosis, prognostication, treatment selection, monitoring 2. Copy Number Alterations - Identify gene amplifications and deletions across the genome. [copy number plot, 0-4, showing gain (green) and loss (red) segments] Applications: Prognosis, therapy selection, resistance mechanisms 3. Methylation Patterns - Detect aberrant DNA methylation changes at specific regions. [Unmethylated CG CG CG CG vs Methylated CG CG CG CG diagram] Applications: Early detection, tissue of origin, prognosis 4. Fragmentomics - Analyze cfDNA fragment size profiles and genomic positioning. [Fragment count vs Fragment size (bp) 50-250 curve, peak ~150bp] Applications: Early detection, tissue of origin, prognosis 5. Structural Variants & Rearrangements - Detect gene fusions, translocations, and other structural alterations. [Gene A - Gene B - Gene A-B Fusion diagram] Applications: Therapy selection, resistance mechanisms Center diagram: "cfDNA in Blood - Contains tumor-derived DNA fragments" with tumor shedding DNA into bloodstream Clinical Utility: Early Detection | Prognosis & Risk Stratification | Therapy Selection & Resistance | Monitoring & Response Assessment | Minimal Residual Disease (MRD) | Surveillance & Recurrence Detection Image generated by ChatGPT Summary - Longitudinal cfWGS guided by personalized mutation catalogs from high disease burden BM detects residual tumor DNA at ultra-low levels - This approach tracked disease dynamics - Achieved strong concordance with clinical BM MRD assays - Predicted PFS - Contributes orthogonal information including genome-wide mutations, CNAs, fragmentomic features and subclonal evolution that protein and flow-based assays cannot - 30X cell-free DNA WGS is emerging as a tool for MRD surveillance, complementing BM MRD and can be followed by "upgrading" to 150X to profile relapsing subclones (inform resistance & treat early?) [Bottom right chart] Circulating tumour DNA frequency (0.001% to 100%) vs disease stage: MGUS -> SMM -> MM -> Treatment -> MRD -> Relapse Labels: "Multiple cancer early detection assays", "150X WGS to pin-point cancer drivers and generate mutation list for tumour-informed MRD", "30X monitoring molecular residual disease", ">150X 'upgrade' to pin-point cancer drivers", "Cure" (green dashed line), "MRD" (blue dashed line), "Relapse/Refractory" (red line up) Pugh TJ. Seminars in Hematology 55 (2018) 38-40

But plasma NGS approaches thus far have shown poor sensitivity [Left chart] Tumour Data table comparing cfDNA Data vs Tumour Data: Tumour Data: 51 | 195 cfDNA Data 53: 49 (teal) | 4 (yellow) -> 92% cfDNA Data 193: 2 (orange) | 191 -> 99% Column totals: 96% | 98% (Kis et al., Nature Communications 2017) [Bottom left chart] "Distribution of mutations in new-diagnosed and relapsed/refractory patients" - bar chart, Y axis "Number of mutations" 0-20, legend: Mutations (PL), Mutations (BM), Mutations (BM + PL) (Mithraprabhu et al., Leukemia 2017) [Middle chart] Gene alteration heatmap across samples/timepoints (2017, 2021, 2025, 2019) for genes: KRAS, NRAS, ATM, ACTG1, BRAF, ERBB4, FGFR3, HIST1H1E, IDH1, IRF4, LTB, B1, TP53, ZFHX4, ABL1 (9p34.12), CDK6 (7q21.2), E2F3 (6p22.3), IL6R (1q21.3), NOTCH1 (9p34.12), SOX4 (6p22.3), ARID1B (6p25.3), CDKN1B (12p13.1), CDKN2C (1p32.3), FAM46C (1p12), RPL5 (1p21.1), TP53 (17p13.1), TRAF3 (14q32.32), RB1 (13q14.2) Legend: Hotspot mutation (black), Missense mutation (green), Nonsense mutation (purple), Not detected but powered (grey), Not powered (white/blank), Gain (red), Loss (blue) (Manier et al., Nature Communications 2018) [Right box] Poor sensitivity when compared to BM MRD3 Plasma cfDNA VDJ NGS (clonoseq) Specificity 90.1% 94.9% PPV Sensitivity 33.9% NPV 21.7% 3Wiedmeier-Nutor et al, Blood Cancer J. 2026;16(1):66 30-50X cell-free DNA whole genome sequencing enables sensitive detection of cancer by testing 1,000s of mutations simultaneously Three 3D surface plots (Detection probability vs Tumor fraction vs Unique genomic equivalents): Single-mutation capture: 0.1% at 10,000X 10-mutation capture: 0.01% at 10,000X 10,000-mutation capture: 0.001% at 30X Below: fingerprint icon progression: Small gene panels -> Bespoke targeted assays -> Whole genome Zviran et al. Nat Med. 2020 Jul;26(7):1114-1124 [Slide number 8] Tumour-informed mutation detection rates decrease after initial treatment, stay down in non-relapsed patients & increase before relapse Two panel scatter/line plot, Y axis "Cumulative VAF (Z)" (0 to >1500), X axis "Weeks Since Baseline" (0 to >250): Left panel: "Relapse within 180 days of blood draw (n = 11 patients; 41 samples)" - red lines/dots trending upward before relapse ("within 180 days of relapse"), black lines trending down Right panel: "No relapse within 180 days of blood draw (n = 33 patients; 105 samples)" - lines trending down and staying near 0 (">180 days to relapse") Abelman et al. medRxiv 2025.10.24.25338566 Summary - Longitudinal cfWGS guided by personalized mutation catalogs from high disease burden BM detects residual tumor DNA at ultra-low levels - This approach tracked disease dynamics - Achieved strong concordance with clinical BM MRD assays - Predicted PFS - Contributes orthogonal information including genome-wide mutations, CNAs, fragmentomic features and subclonal evolution that protein and flow-based assays cannot - 30X cell-free DNA WGS is emerging as a tool for MRD surveillance, complementing BM MRD and can be followed by "upgrading" to 150X to profile relapsing subclones (inform resistance & treat early?) [Bottom right chart] Circulating tumour DNA frequency (0.001% to 100%) vs disease stage: MGUS -> SMM -> MM -> Treatment -> MRD -> Relapse Labels: "Multiple cancer early detection assays" (low %), "150X WGS to pin-point cancer drivers and generate mutation list for tumour-informed MRD" (peak ~10-100%), "30X monitoring molecular residual disease", ">150X 'upgrade' to pin-point cancer drivers", "Cure" (green dashed line down), "MRD" (blue dashed line), "Relapse/Refractory" (red line up) Pugh TJ. Seminars in Hematology 55 (2018) 38-40

Mass spectrometry approaches to M-protein detection Table: INTACT PROTEIN MS | CLONOTYPIC MS MASS-FIX . EXENT | EasyM . M-InSight Measures the molecular mass of the M-protein | Measures patient-specific clonotypic peptides Prior knowledge of the M-protein not required | Prior characterization of the M-protein required Diagnosis / screening [check] | Diagnosis / screening Monitoring [check] | Monitoring [check][check] (International Myeloma Society logo) Serum MS and BM MRD SERUM MS- SERUM MS+ BM MRD- Concordant deep response | Residual serum disease: Delayed clearance-especially early/IgG . patchy or extramedullary disease . BM sampling or analytical attribution BM MRD+ Residual marrow disease: Low- or non-secretory residual clone . serum below detection . trackability or sensitivity limitations | Concordant residual disease Interchangeable, complementary, hierarchical or independent? Footnote: Interpretive framework based on paired serum MS / BM NGF datasets. Puig et al., Blood 2024. (International Myeloma Society logo) PFS according to paired MRD results across NGF, NGS and MS at prespecified CASSIOPEIA landmarks 5 Kaplan-Meier PFS curves (Y axis PFS %, X axis Time in months), each with 4 groups NGF+/MS+, NGF-/MS-, NGF+/MS-, NGF-/MS+ with n(events) counts and significance brackets: Day28 combined NGF/MS: NGF+/MS+ 93(59); NGF-/MS- 12(0); NGF+/MS- 10(3); NGF-/MS+ 30(11) Day100 combined NGF/MS: NGF+/MS+ 53(36); NGF-/MS- 63(24); NGF+/MS- 15(5); NGF-/MS+ 25(17) Week 25 combined NGF/MS (arrow-highlighted): NGF+/MS+ 45(38); NGF-/MS- 90(28); NGF+/MS- 15(10); NGF-/MS+ 25(14) Week 52 combined NGF/MS: NGF+/MS+ 30(28); NGF-/MS- 83(23); NGF+/MS- 9(5); NGF-/MS+ 7(3) Week 105 combined NGF/MS: NGF+/MS+ 20(15); NGF-/MS- 90(14); NGF+/MS- 8(3); NGF-/MS+ 7(6) Box: 'Complementary [not equal] Hierarchical - BM MRD often refines an MS result; the incremental value of MS after high-quality BM MRD is less consistently demonstrated' Citation: Dejoie T, et al. Blood. 2026 Aug 17:blood.20260337 (International Myeloma Society logo) Harmonizing MS with the Current Response Framework CURRENT RESPONSE FRAMEWORK: Baseline -> PR >=50% down -> VGPR >=90% down or SPEP-/IFE+ -> CR IFE- MS MEASUREMENT RANGE: QUANTITATIVE MS (spans PR through into VGPR/CR range) -> Detectable/non-quantifiable -> MS- >=90% -> >=99% -> >=99.9% -> deeper response -> MS- PR: likely transferable | HARMONIZATION ZONE Measure and define clinically meaningful response thresholds (International Myeloma Society logo)

Mass spectrometry approaches to M-protein detection Table: | INTACT PROTEIN MS | CLONOTYPIC MS Assay names | MASS-FIX · EXENT | EasyM · M-InSight Measures | Measures the molecular mass of the M-protein | Measures patient-specific clonotypic peptides Prior knowledge | Prior knowledge of the M-protein not required | Prior characterization of the M-protein required Diagnosis / screening | Yes (check) | Diagnosis / screening (no checkmark shown) Monitoring | Yes (single check) | Monitoring (double check) International Myeloma Society MS Extends the Observable Disease Trajectory in Both Directions SCREENING / EARLY DETECTION: - Low-level clone, Small M-protein, Earlier detection - Additional diagnostic window DEEPENING RESPONSE: - VGPR, ≥99% ↓, ≥99.9% ↓ - Additional depth of response NADIR / MS- (center box) DISEASE RE-EMERGENCE: - Reappearance, Quantifiable rise, Rising clone - Additional pre-progression window - CONVENTIONAL PD (below, right side) International Myeloma Society Higher sensitivity for M-protein detection PROMISE (Population screening, n=7,622): 6% → 13% MGUS detected, SPEP/IFE → MS 43% of MS-detected clones were very low level (<0.2 g/L; MGIP) OLMSTED COUNTY (Previously screening-negative, n=226): 10.6% → 50.0% → 65.9% IFE → MALDI-TOF → miRAMM Monoclonal proteins detectable ~10 years before clinical diagnosis miRAMM: LC and high-resolution MS (LC) El-Khoury et al. Lancet Haematol 2022 · Murray et al. Blood Cancer J 2019 International Myeloma Society Serum MS and BM MRD 2x2 grid: | SERUM MS- | SERUM MS+ BM MRD- | Concordant deep response | Residual serum disease - Delayed clearance-especially early/IgG · patchy or extramedullary disease · BM sampling or analytical attribution BM MRD+ | Residual marrow disease - Low- or non-secretory residual clone · serum below detection · trackability or sensitivity limitations | Concordant residual disease Interpretive framework based on paired serum MS / BM NGF datasets. Puig et al., Blood 2024. International Myeloma Society

Circulating cells: recapitulate marrow transcriptome and genome Molecular profiling: DNA and RNA sequencing [illustration: blood draw -> circulating tumor cells] Similar genetic makeup [panel D heatmap]: Copy number scale 0 1 2 3 >=4; markers FISH + WGS (black) and WGS only (green). Rows labeled CTCs / BMPCs and patient IDs CTF001-CTF056 (approx.) across columns 1-19+ representing genomic regions; row annotation "gain(1q) not found by FISH" pointing to relevant rows. Reference: Dutta, Alberge et al., Can Disc 2023. PMID 36477267 Similar transcriptome and clonal structure [UMAP plots]: individual patient UMAPs labeled MM_1 through MM_7 (Bone marrow vs Peripheral blood dots), n=65,819 cells. Bar charts to the right for MM_1-MM_7 showing subclone proportions (Subclone 1, 2, 3, 4) split by BM and PB for each patient. References: Fokkema, Bertamini et al., Blood 2026. PMID 41411148; Lightbody et al., Nat Cancer 2025. PMID 40781193 Abbreviations: CTC, circulating tumor cell International Myeloma Society 23rd International Myeloma Society Annual Meeting 11 Circulating cells: higher levels prognostic CTC burden measure: flow cytometry [scatter plot CD38 vs CD45 showing two populations] CTCs: a gradient of risk (European CTC Consortium) Kaplan-Meier PFS curves by CTC burden category, x-axis Time from diagnosis (months) 0-96+, y-axis PFS (probability) 0.00-1.00: >2400 newly-diagnosed myeloma + 118 PCL (young/old, trial/real world, mainly triplets) Categories (best to worst prognosis, colored lines): <0.002% (blue) 0.002 to 0.02% (blue/purple) 0.02 to 0.2% (green) 0.2 to 2% (orange) 2 to 5% (dark red/maroon) >=5% (PCL) (dark red, steepest decline) Abbreviations: CTC, circulating tumor cell; PCL, plasma cell leukemia Reference: Bertamini et al., JCO 2026 in press International Myeloma Society 23rd International Myeloma Society Annual Meeting 12 Combining CTCs with other biomarkers Left panel: 2025 IMS/IMWG risk Kaplan-Meier PFS curves (x-axis Time from diagnosis, months, 0-96+; y-axis PFS probability 0.00-1.00), four groups: CTC-low - CGS std-risk (light blue, best outcome) CTC-low - CGS high-risk (dark blue) CTC-high - CGS std-risk (pink/salmon) CTC-high - CGS high-risk (dark red, worst outcome) Log-rank p<0.0001 Abbreviations: CTC, circulating tumor cell; CGS, consensus genomic staging; std, standard; SS, staging system. CTC-low <0.02%, CTC-high >=0.02% Right panel: Blood-based staging (CTC-SS) CTC log10 levels, ISS, LDH Kaplan-Meier PFS curves, three groups: CTC-SS I (blue, best) CTC-SS II (orange) CTC-SS III (dark red, worst) p<0.0001 References: Bertamini et al., JCO 2026 in press; D'Agostino et al., under revision 2026 European CTC Consortium International Myeloma Society 23rd International Myeloma Society Annual Meeting 15 More CTCs reflect tumor burden and complex genetics Left: "Why more CTCs = worse prognosis?" Kaplan-Meier PFS curve (same multi-tier CTC-burden curve as prior slide, x-axis Time from diagnosis months 0-96+, y-axis PFS probability 0.00-1.00). Reference: Bertamini et al., JCO accepted 2026 Right: three linked concept circles feeding into "Genetic alterations" box: Cell proliferation (dividing cell icon) BM tumor burden (clustered plasma cell icon) Genetic alterations (DNA icon) -> leads to: - TP53 mutations/del17p, amp1q - MAF translocations t(14;16), t(14;20), t(8;14) - APOBEC mutagenesis, chromothripsis Abbreviations: CTC, circulating tumor cell; BM, bone marrow References: Fokkema, Bertamini et al., Blood 2026. PMID 41411148; Garces, Diamond, Sevcikova et al., Hemasphere 2025. PMID 40995465 International Myeloma Society 23rd International Myeloma Society Annual Meeting 16

Circulating cells in myeloma Diagram: blood draw -> circulating tumor cells -> two detection method branches: Microscopy (H&E morphology, Slide-based immunofluorescence) and Flow cytometry (2-4 colors flow cytometry, 6-8 colors flow cytometry; small CD38 vs CD45 flow plot shown). Bar chart "Detection rate (% patients with CTCs)" vs "Sensitivity": H&E morphology[1-2]: 4% | Sensitivity 10^-3 Slide-based immunofluorescence[3]: 35% | Sensitivity 10^-4 2-4 colors flow cytometry[4]: 60% | Sensitivity 10^-5 6-8 colors flow cytometry[5-8]: 90% | Sensitivity 10^-6 Abbreviations: CTC, circulating tumor cell; MM, multiple myeloma; NGF, next-generation flow cytometry; H&E, hematoxylin and eosin staining; FC, flow cytometry 1. Gluzinski and Reicheinstein, Wiener Klinische Wochenschrift 1906. 2. Granell et al., Haematologica 2017. 3. Witzig et al., Blood 1996. 4. Nowakowski et al., Blood 2005. 5. Dingli et al., Blood 2006. 6. Bertamini et al., JCO 2022. 7.Garces et al., JCO 2022. 8. Jelinek et al., JCO 2022. 9. Kostopoulos et al., JCO 2023. International Myeloma Society - 23rd International Myeloma Society Annual Meeting [Slide 9] Circulating cells: higher levels prognostic Left: "CTC burden measure: flow cytometry" - small CD38 vs CD45 flow plot; blood draw diagram. Abbreviations: CTC, circulating tumor cell; PCL, plasma cell leukemia Right, labeled "CTCs: a gradient of risk" with "European CTC Consortium" logo: Kaplan-Meier chart, y-axis "PFS (probability)" 0.00-1.00, x-axis "Time from diagnosis (months)" 0-96+. Annotation: ">2400 newly-diagnosed myeloma +118 PCL (young/old, trial/real world, mainly triplets)" Risk-stratum legend (color gradient, best to worst PFS): <0.002%, 0.002 to 0.02%, 0.02 to 0.2%, 0.2 to 2%, 2 to 5%, >=5% (PCL) [arrow indicates decreasing PFS with increasing CTC%] Citation: Bertamini et al., JCO 2026 in press International Myeloma Society Annual Meeting [Slide 12] More CTCs reflect tumor burden and complex genetics Left box: "Why more CTCs = worse prognosis?" with the same Kaplan-Meier PFS-by-CTC-burden curve as prior slide (y-axis PFS probability 0-1.00, x-axis Time from diagnosis (months) 0-96+). Citation: Bertamini et al., JCO accepted 2026. Abbreviations: CTC, circulating tumor cell; BM, bone marrow Right: three-circle schematic linked to the CTC concept - "Cell proliferation", "BM tumor burden", "Genetic alterations" -> arrows to: TP53 mutations/del17p, amp1q; MAF translocations t(14;16), t(14;20), t(8;14); APOBEC mutagenesis, chromothripsis References: Fokkema, Bertamini et al., Blood 2026. PMID 41411148 Garces, Diamond, Sevcikova et al., Hemasphere 2025. PMID 40995465 International Myeloma Society Annual Meeting [Slide 16] Future myeloma treatment "Risk adapted" schematic: patient population split into three risk groups (blue/orange/red figures): - Low-risk: less intensive, fixed duration, response adapted (icon: toy dart gun) - Intermediate-risk: multidrug, intermediate intensity, response adapted (icon: question mark) - High-risk: intensive multidrug, long duration, MRD neg sustained (icon: explosion/bomb) Arrows point to a schematic survival curve (y-axis "Survival" 0-100, x-axis "Time") showing three lines labeled Low risk (top), Intermediate risk (middle), High risk (bottom) - illustrative/conceptual curve, not showing specific numeric data points. Black banner: Better diagnostics for better treatment and outcome Abbreviations: MRD, minimal residual disease International Myeloma Society Annual Meeting [Slide 18]

MagnetisMM-3: Infections in Patients on Elranatamab With or Without Ig Replacement Therapy RRMM refractory to ≥1 of each of the following: PI, IMiD and anti-CD38 mAb Any grade (N=187) bar chart - Monthly exposure-adjusted infection rate (95% CI) by category: All infections: rate with 0.22, without 0.36; cases with 96, without 402 Bacterial: rate with 0.04, without 0.09; cases with 19, without 106 Fungal: rate with 0.02, without 0.03; cases with 8, without 31 Viral: rate with 0.07, without 0.10; cases with 31, without 115 Pathogen unspecified: rate with 0.09, without 0.13; cases with 38, without 150 Grade ≥3 (N=187) bar chart: All infections: rate with 0.05, without 0.14; cases with 22, without 152 Bacterial: rate with 0.02, without 0.05; cases with 9, without 56 Fungal: rate with 0.00, without 0.01; cases with 2, without 7 Viral: rate with 0.02, without 0.05; cases with 9, without 59 Pathogen unspecified: rate with 0.00, without 0.03; cases with 2, without 30 In MagnetisMM-3, prophylactic administration of SCIg or IVIg was recommended for patients with an IgG ≤400 mg/dL Lower monthly exposure-adjusted infection rates were observed in patients with vs without Ig replacement (0.22 vs 0.36) Similar trends were observed across infection types Data cutoff was 16 April 2023. Median follow-up not reported. CI, confidence interval; Ig, immunoglobulin; IMiD, immunomodulatory drug; IV, intravenous; mAb, monoclonal antibody; PI, proteasome inhibitor; RRMM, relapsed and/or refractory multiple myeloma; SC, subcutaneous. 1. Lelue X, et al. Abstract P-286. Poster presented at: 2023 IMS Annual Meeting; Athens, Greece. 2. Lesokhin AM, et al. Nat Med. 2023;29:2259-67. Figures reproduced from Lelue X, et al. Abstract P-286. Poster presented at 2023 IMS Annual Meeting; Athens, Greece. 3. ClinicalTrials.gov. MagnetisMM-3 Study Protocol. Accessed 21 September 2026. This presentation is intended for healthcare professionals only. (slide 89) The Multidisciplinary Team Is Involved Throughout BsAb Therapy 1. INITIAL EVALUATION: Evaluate patient for eligibility for outpatient step-up dosing administration 2. PATIENT AND CARE PARTNER EDUCATION: Educate patient and care partner on monitoring for CRS and ICANS during early treatment phase 3. TREATMENT ADMINISTRATION: Ensure correct dosing and route, beginning with C1D1 and extending throughout therapy 4. LONG-TERM MANAGEMENT: Monitor for, and help mitigate, potential AEs throughout the treatment course The multidisciplinary team is a crucial element of successful BsAb administration throughout the treatment process AE, adverse event; BCMA, B-cell maturation antigen; BsAb, bispecific antibody; C, Cycle; CRS, cytokine release syndrome; D, Day; ICANS, immune effector cell-associated neurotoxicity syndrome. 1. Bellevier C, et al. Clin J Oncol Nurs. 2025;29:E7-15. 2. Myeloma UK HCP Hub. Consensus framework for the safe delivery of bispecific antibodies in multiple myeloma. Accessed 24 August 2026. 3. International Myeloma Foundation Nurse Leadership Board. Bispecific antibody therapy: Exploring efficacy, safety, and transition of care management. Accessed 25 August 2026. 4. Garfall AL, et al. Front Oncol. 2025;15:1630146. This presentation is intended for healthcare professionals only. (slide 84)

Instead Conclusion: How to reduce cost pressure, provide cure and productive investment Strongest policy case combines timely treatment access with support that preserves function and employment 1. Accelerate access: Reduce delays to effective treatment and use managed-entry or outcomes-based agreements to balance access with affordability. 2. Preserve productivity: Integrate rehabilitation, occupational support and flexible work pathways for patients during and after treatment. 3. Support caregivers: Protect caregiver employment through flexible arrangements. 4. Measure full-value: Include tax revenues, social transfers and indirect losses when making decision about reimbursement and budget allocation Modelled opportunity 2024-2030: Annual savings ≈ 9M Eu per country 1. Diagnose earlier - Primary care awarness; referral triggers and standard minimum test sets 2. Make risk stratification routine: Reliable access to standard risk stratification methodology 3. Close the reimbursement gap: Prioritize first-line access for best outcomes 4. Use the regional networks eg. Balkan Myeloma Registry, EMN, IMS T-MAP: Expand registry coverage, shared standards and cross-country evidence generation Figure 8: Barriers to clinical trials for multiple myeloma in the CE&B region - Administrative barriers and workforce limitations - Lack of access to approved treatments/patients may not fulfil eligibility criteria - Small number of patients for recruitment disincentivises pharmaceutical companies - Lack of awareness and stigma regarding clinical trials among patients - Lack of, or perverse, incentives for physicians International Myeloma Society

Instead Conclusion: How to reduce cost pressure, provide cure and productive investment Strongest policy case combines timely treatment access with support that preserves function and employment 1. Accelerate access: Reduce delays to effective treatment and use managed-entry or outcomes-based agreements to balance access with affordability. 2. Preserve productivity: Integrate rehabilitation, occupational support and flexible work pathways for patients during and after treatment. 3. Support caregivers: Protect caregiver employment through flexible arrangements. 4. Measure full-value: Include tax revenues, social transfers and indirect losses when making decision about reimbursement and budget allocation Modelled opportunity 2024-2030: Annual savings ≈ 9M Eu per country 1. Diagnose earlier - Primary care awarness; referral triggers and standard minimum test sets 2. Make risk stratification routine: Reliable access to standard risk stratification methodology 3. Close the reimbursement gap: Prioritize first-line access for best outcomes 4. Use the regional networks eg. Balkan Myeloma Registry, EMN, IMS T-MAP: Expand registry coverage, shared standards and cross-country evidence generation Figure 8: Barriers to clinical trials for multiple myeloma in the CE&B region - Administrative barriers and workforce limitations - Lack of access to approved treatments/patients may not fulfil eligibility criteria - Small number of patients for recruitment disincentivises pharmaceutical companies - Lack of awareness and stigma regarding clinical trials among patients - Lack of, or perverse, incentives for physicians International Myeloma Society

Clinical Trials: Unequal Access to Innovation Research infrastructure is concentrated on a few countries and major urban centers Table (Pais/Country | Total | CAR-T Cells | Biespecificos/Belantamab | CELMoDs | Anti-CD38 | Otros): Brasil | 43 | 4 | 21 | 4 | 5 | 9 Argentina | 19 | 2 | 9 | 4 | 2 | 1 Mexico | 12 | 0 | 7 | 2 | 1 | 0 Chile | 7 | 0 | 3 | 2 | 2 | 0 Colombia | 2 | 0 | 0 | 2 | 0 | 0 - Clinical trials can provide access to innovative therapies. - Sustainable research infrastructure and referral networks are essential, not optional - Brazil has the highest number of MM clinical trials in LATAM, (40+), reflecting its most advanced research infrastructure and high enrollment rate. - Mexico and Argentina: also have a significant number of trials, benefiting from strong healthcare systems and collaboration with international pharmaceutical companies. Generally, 10-20 trials. - Chile, Colombia, Peru, etc.: Have smaller numbers of trials, often focused on urban centers with advanced medical facilities. Fewer than 10 trials per country, with some having only 1-5 trials. International Myeloma Society ClinicalTrials.gov

23rd Annual MEETING & EXPOSITION - International Myeloma Society Current (GMT+1) 07:03 AM LMIC Breakfast Session 7:00 AM - 8:00 AM Angelo Maiolino, Shaji Kumar, Uday Yanam... (Moderator) Shaji Kumar - Welcome and objectives - 7:00 AM - 7:05 AM Jelena Bila - The real-world burden of multiple myeloma in... - 7:05 AM - 7:20 AM Dorotea Fantl - The real-world burden of multiple myeloma in... - 7:05 AM - 7:20 AM Uday Yanamandra - IMS LMIC Survey: identifying the major gaps in... - 7:20 AM - 7:40 AM Angelo Maiolino - IMS LMIC Survey: identifying the major gaps in... - 7:20 AM - 7:40 AM

[Slide: Conclusions - GEM Frailty/GAH Study, IMS 2026] Conclusions - The SFI appears to overestimate frailty in a relevant proportion of TIE patients, potentially restricting access to quadruplet therapy. - The concordance between SFI and GAH for identifying patients potentially eligible for quadruplet was modest about 45%. - The GAH scale may help identify different fitness profiles. In particular, GAH-ultra-fit patients appears to have a more favourable benefit-risk profile with DKRd, with lower unacceptable toxicity and higher MRD negativity, resulting in longer TTF. - These exploratory findings support further prospective evaluation of the GAH scale for patient selection, particularly in the context of quadruplet regimens and emerging immunotherapy-based approaches. Presented at International Myeloma Society 23rd Annual Meeting & Exposition. Affiliations: Hospital Universitario de Salamanca (Hematologia), CIC, IBSAL, Sacyl, Junta de Castilla y Leon, Universidad de Salamanca.

Conclusion Three-column layout: 1. Which therapy? 2026: Lenalidomide + daratumumab [arrow down] 2030: Immunotherapy approaches? 2. How long for? 2026: To progression but..... - low threshold for stopping after 2-3 years in event of toxicity if no high risk lesions and MRD -ve - higher threshold for stopping/try to continue if any high risk lesions or MRD +ve [arrow down / clock icon] 2030: Immunotherapy approaches likely to be time limited 3. Personalisation? 2026: Double hit / IMWG-IMS high risk patients need intensive consolidation and multi-agent maintenance approaches [arrow down] 2030: More information on older patients and time limited therapy Footer: ICR - Images created using BioRender. The ROYAL MARSDEN NHS Foundation Trust

Conclusions (UKE Hamburg / Universitatsklinikum Hamburg-Eppendorf, slide 20) - Quadruplet induction is currently approved for all NDMM patients - Quadruplet induction in the setting of consolidation and (combination) maintenance/continuous treatment allows the highest rates of MRD and sustained MRD negative responses which translates into the current most favorable PFS and OS outcomes - Sustained MRD negativity is the imminent prerequisite for stopping treatment and a potential curative approach - Achieving MRD negative response after induction allows to question the role of ASCT in standard-risk patients - Strategies including quadruplet induction show the lowest rates of functional high-risk disease - Carfilzomib based quadruplets show favorable MRD responses especially in high-risk patients - For comorbid, old and/or frail patients, triplet or doublet treatment should be considered - Whenever available and feasible, quadruplet treatment should be initiated in any NDMM patient

Getting the right treatment to the right patient: Using diagnostics to personalise care and maximise treatment value? Enhanced Risk Assessment at 1st relapse (GEP risk* + IMS-IMWG (NGS) + WB-MRI) + Personalised MRD tracking (Bone marrow NGS MRD + WB-MRI Imaging MRD) = Joint decision making / Choice essential! -> -> e.g. for High-Risk MM: Continued treatment* -> e.g. for Standard Risk + MRD-neg: Treatment Stop* optional, supported by Enhanced remission monitoring (PB Mass-Spec) Treatment selection e.g. options (from Enhanced Risk Assessment + Personalised MRD tracking, plus Patient factors/preference): Option TCE+CD38(1); Option CAR-T(2); Option ADC triplet; Option Trad triplet, etc. Footnote citations: 1 Banerjee R, et al, EHA 2026. 2 Costa L, et al, ASH 2025. *Kaiser M et al, JCO 2023; Leypoldt L JCO 2024; Beer S et al, Blood 2026; Kaiser M et al, Lancet Onc 2026; Kumar S et al, NEJM 2026 Bottom banner: 'UK patient-centric diagnostic pathway in development by UKMS. Presently accessible: MyTrack Myeloma, others...' (UK Myeloma Society logo)

CAR-T, Bispecific or ADC? Table: Aspect | ADC (Belantamab) | CAR-T (cilta-cel) | BCMA or GPRC5D-BsAb Indication | RRMM after >=1 prior LoT (PI + IMiD exposed) | RRMM after >=1 prior LoT (PI + IMiD; often lenalidomide-refractory) | RRMM after >=1 prior LoT (PI + IMiD) Efficacy | Improved outcomes vs DVd/PVd | Marked efficacy benefit vs standard of care DPd or PVd | Clinically meaningful efficacy benefit vs standard of care with the best HR so far reported Key safety | Ocular toxicity, thrombocytopenia | CRS, ICANS, infections, delayed neurotoxicity | Infections requiring prophylaxis and IVIG; On target off tumor toxicity for GPRC5D Eligibility | Few formal contraindications | Required normal function of vital organs | Few formal contraindications Treatment burden | Regular ophthalmologic assessments | Lymphodepletion and close post-infusion monitoring | Clinical monitoring focused on infectious risk Administration | Intravenous, short infusion time, no premed | Single infusion at certified centers | Subcutaneous with SUD and premed: Monthly admon in the f/u for most of them CELMoDs offer convenient, non-T-cell-redirecting options, oral administration [IBSA logo] What about sequencing: Crucial challenge What does the sequencing evidence support? BCMA CAR-T -> GPRC5D Best prospective evidence - MonumenTAL-1 prior T-cell redirection cohort - After prior CAR-T: ORR 72.9%; median PFS 12.3 months - After prior BCMA bispecific: ORR 52.2%; median PFS 4.1 months GPRC5D -> BCMA Feasible, but mainly retrospective Small published series and registries - support activity of BCMA CAR-T or bispecific therapy - A 10-patient series reported median PFS of 5.5 months Could belamaf rescue BCMA-TT or viceversa? - Different mechanism: ADC activity does not depend on T-cell fitness - Potentially useful after BCMA CAR-T or bispecific failure if BCMA remains expressed or viceversa Practical sequencing principles - Switch target after early or refractory failure: GPRC5D after BCMA has the clearest clinical support - Avoid immediate T-cell redirection when possible: Real-world series associate intervals >=6 months with better outcomes - Identify the mechanism of resistance involved in the relapse/progression of the disease: Consider antigen retention, T-cell fitness, disease tempo and infection burden Evidence remains incomplete Other challenges When to start treatment? - The earlier the use of these therapies -> The better - Should we use them at biochemical PD or clinical PD -> Biochemical PD is better but... - Would it be possible to start in case of relapse from MRD-ve? Schedule of treatment - Should we move to treatments of fixed duration, especially with Bispecific mAbs? - Fixed duration or duration adapted to the MRD? - Would it be possible to implement maintenance after CAR-T for maintaining the MRD-ve - Is it possible to extend the interval between doses to improve the safety profile?

To Declare Cure Requires convergence of three domains When All Three Align -> Credible Declaration of Cure. Domain | Requirement Statistical | No excess mortality Clinical | Long-term Treatment-free Disease-free survival Biological | No detectable malignant clone International Myeloma Society Proposed Definition of Cure Patient could be newly-diagnosed or with relapsed myeloma Patient in sustained Complete Remission (CR) by standard criteria MRD negative at 10^-6 sensitivity by either NGS or NGF MRD negativity Sustained for 5 years WITHOUT any therapy. (Irrespective of length of previous sustained MRD negativity on therapy) - There should be at least 4 MRD assessments in 5 years, last one being at 5 years; all negative and no positive test in between, if performed. Negative by functional imaging (PET/CT or DWI WB MRI) - At least at the beginning of 5 years period and at the end of 5 years period with no intervening positivity, if performed. (slide 28, International Myeloma Society) Timing of relapse after stopping therapy Time to Progression after stopping therapy (KM curve, TTP probability, 0-96 months) Stopping Treatment: Patients with sustained MRD-negativity (≥24 mo), N=265 stopped treatment at varying timepoints. Landmark analysis for all patients from time of stopping therapy, with median f/u 3 yrs after holding Tx Number at risk (All): 265 210 159 123 82 55 42 24 12 (at 0,12,24,36,48,60,72,84,96 months) First 5 years progression risk: 20.8% After 5 years progression risk: 3.0% Interpretation: Relapse risk after stopping therapy plateaus at ~5 years - supporting the ≥5 year off therapy rule. Courtesy Sundar Jagannath - Unpublished (slide 22)

Photo of poster board "LBA-02", 23rd IMS Annual Meeting: LBA-02 Early Indian Real-World Experience with Imported Talquetamab in Heavily Pretreated Plasma Cell Neoplasms: A Single-Centre Case Series Mehak Trehan, Paritosh Garg, Anusha Swaminathan, Nikhil M Kumar, Rahul Bhargava Fortis Memorial Research Institute, Gurugram, India Frequent recorded responses accompanied substantial infection and mucocutaneous toxicity in this 10-patient case series 80% Recorded response ≥PR (8/10; exact 95% CI 44.4-97.5%) 60% Recorded complete response (6/10 patients) 50% Grade ≥3 infection (5/10 patients) Study design & cohort: Retrospective descriptive review of 10 adults treated with imported talquetamab at an Indian tertiary centre. Clinicians recorded response using IMWG criteria, with no independent adjudication. [1] Table - Characteristic | Value: Age, median (range): 62 (46-78) years Myeloma / plasma cell leukaemia: 9 / 1 Prior lines, median (range): 5 (4-6) Prior BCMA-directed exposure: 6/10 (60%) Teclistamab-refractory disease: 5/10 (50%) Extramedullary involvement: 3/10 (30%) Recorded adverse events (table): Event | Patients CRS, all grade 1: 7/10 (70%) ICANS: 0/10 Any-grade infection: 7/10 (70%) Grade ≥3 infection: 5/10 (50%) CMV reactivation: 4/10 (40%) Grade 3 cytopenia: 6/10 (60%) Taste/oral changes: 9/10 (90%) Skin changes / nail changes: 6/10 each (60%) One pulmonary tuberculosis event was recorded. No deaths were recorded during the available observation window. Event categories may overlap. Recorded response categories (bar chart, counts out of 10): CR 6 (60%), VGPR 1 (10%), PR 1 (10%), PD 1 (10%), Persistent/no response 1 (10%) Counts are recorded response categories, not independently adjudicated ORR. One PCL patient had VGPR. Among 3 patients [with] EMD, 2 had CR and 1 had persistent/no response. Delivery & interpretation: [text partially obscured by presenter] received inpatient step-up and outpatient continuation, monthly antiviral/Pneumocystis prophylaxis and vaccination. [Small sample] size, variable observation and nonadjudicated outcomes limit [interpret]ation. [No pre]specified clinical cutoff or survival estimates were available. Nail changes during treatment: photo of hands showing nail changes in a patient from this series. Photograph reproduced with permission. Mucocutaneous effects: Taste/oral effects affected 9/10 patients. Skin and nail changes each affected 6/10. These grouped records do not establish the grade or timing of individual findings. Neurologic observation: Two dizziness/ataxia events did not meet ICANS criteria. [Footnote references partially cut off]: IMWG response and MRD criteria. Lancet Oncol. 2016;17:e328-e346. [...] ASTCT consensus grading. Biol Blood Marrow Transplant. 2019;25:625-638. Ethics: FMRI IEC/2026/OAS/027 (28 July 2026). Response; VGPR, very good partial response; PR, partial response; PD, progressive disease; PCL, plasma cell leukaemia; BCMA, B-cell maturation antigen; EMD, extramedullary disease; CMV, cytomegalovirus; IVIG, intravenous immunoglobulin; CRS, cytokine release syndrome; ICANS, immune effector cell-associated neurotoxicity syndrome. Disclosures: Nothing to disclose. Funding: No specific funding.

Where we come from: ASCT vs the same drugs Bar chart - Absolute median PFS gain with ASCT (months): IFM 90 (27 vs 18): +9 EMN02 (56.7 vs 41.9): +14.8 IFM 2009 (47 vs 35): +12 DETERMINATION (67 vs 46): +21 FORTE (99 vs 70): +29 Labels: median PFS with ASCT vs without (months); FORTE = KRd-ASCT vs KRd12, no transplant Attal 1996; Cavo 2020; Richardson 2022; Gay 2021; FORTE 8-y update, EHA 2025 (median PFS 99 vs 70 mo, HR 0.69); Chakraborty, Cancer 2022 The control arm never changed: Chemotherapy -> VMP -> VRd -> KRd: always proteasome inhibitor ± IMiD No anti-CD38, no immunotherapy in any no-ASCT arm Benefit grows with induction depth: +9 months (IFM 90) to +29 months (FORTE, 8-year follow-up) OS gain mostly in high-risk cytogenetics: HR 0.66 vs 0.90 in standard risk (4 RCTs, n = 2,959) Banner: ASCT was tested against the same mechanisms of action, never against a new one International Myeloma Society 23rd International Myeloma Society Annual Meeting 2 Who might still need an alkylator? Table: Profile | Why an alkylator might still matter | Evidence today t(11;14) | Overexpresses cyclin D1. BCL2 dependence -> cells more "primed" for apoptosis -> hypothesis: more sensitive to pro-apoptotic stress including melphalan | Hypothesis? MIDAS small numbers t(11;14): n=32 Arm A, n=46 Arm B Very high risk (biallelic TP53, ≥2 CGS criteria, PCL) | Proliferative disease: cytotoxic pressure may complement T-cell engagement | Tandem ASCT case by case, no robust data Extramedullary disease, high tumor burden | Rapid debulking alongside cell therapy (bridging, CRS risk) | Hypothesis? Antigen escape (BCMA loss) | Argues first for dual targeting (BCMA × GPRC5D, BCMA/CD19) rather than an alkylator | Being tested in the multi-target trials T-cell fitness, marrow microenvironment | Exhausted or immune-cold marrow: T-cell engagement may not suffice; Mel may modify the niche | Hypothesis: biomarker to test? The new trials need prespecified biology-based subgroups and translational cohorts: MRD kinetics, T-cell fitness, target expression International Myeloma Society 23rd International Myeloma Society Annual Meeting 7 Perspectives: which patient needs what? Three columns: Standard risk, MRD-negative | Grey zone (e.g., t(11;14), MRD-positive) | High / very high risk Today (all three columns): Anti-CD38-VRd -> Mel 200 + ASCT -> Dara-R Tomorrow (if the trials are positive): - Standard risk, MRD-negative: Anti-CD38-VRd -> T-cell redirection (bispecific or CAR-T), no Mel 200 - Grey zone: Open question: is an alkylator still needed? Biomarker-driven trials - High / very high risk: ASCT + T-cell redirection (e.g., ASCT -> anti-BCMA CAR-T) Banner: Today: ASCT for all eligible patients. Tomorrow: not "ASCT or not", but which patient needs which. International Myeloma Society 23rd International Myeloma Society Annual Meeting 8

Proposed Definition of Cure - Patient could be newly-diagnosed or with relapsed myeloma - Patient in sustained Complete Remission (CR) by standard criteria - MRD negative at 10^-6 sensitivity by either NGS or NGF - MRD negativity Sustained for 5 years WITHOUT any therapy. (Irrespective of length of previous sustained MRD negativity on therapy) - There should be at least 4 MRD assessments in 5 years, last one being at 5 years; all negative and no positive test in between, if performed. - Negative by functional imaging (PET/CT or DWI WB MRI) - At least at the beginning of 5 years period and at the end of 5 years period with no intervening positivity, if performed. International Myeloma Society [Slide number 28] Cure Fraction - Accumulated data suggests that a cure fraction is observed following various therapies and long-term follow up - Average calculated cure ranges between 5-30% - These numbers are with therapies utilized 10 years ago. We would predict a higher number with the current modern therapies - The cure fraction varies based on - Type of therapy - Age - Myeloma Risk categories International Myeloma Society [Slide number 29]

Table: Clinical trial participation by country (columns: Country | Number of Trials | Total Enrollment | Population 2025), with red arrows highlighting select rows (United States, Canada, France, Germany, China): United States: 337 | 17,962.0 | 341,814,000 [arrow highlighted] Israel: 6 | 400.0 | 9,932,000 Norway: 3 | 215.0 | 5,550,000 Denmark: 3 | 60.0 | 5,973,000 Greece: 5 | 219.0 | 10,246,000 Switzerland: 4 | 411.0 | 8,968,000 Canada: 17 | 1498.0 | 41,288,000 [arrow highlighted] Netherlands: 6 | 878.0 | 17,942,000 France: 19 | 3001.0 | 68,374,000 [arrow highlighted] Sweden: 3 | 83.0 | 10,658,000 Ireland: 1 | 16.0 | 5,281,000 Spain: 9 | 884.0 | 48,592,000 Japan: 20 | 484.0 | 123,753,000 Australia: 4 | 363.0 | 27,184,000 Germany: 11 | 2239.0 | 84,593,000 [arrow highlighted] United Kingdom: 8 | 1005.0 | 68,598,000 Austria: 1 | 200.0 | 9,159,000 Belgium: 1 | 23.0 | 11,738,000 China: 102 | 6282.0 | 1,408,112,000 [arrow highlighted] Italy: 4 | 199.0 | 58,784,000 Slide number 64. Small inset video/photo of a speaker at a podium (International Myeloma Society branding) in the top-right corner. International Myeloma Society

International Myeloma Society RWE: Cilta-cel outcomes in earlier versus later lines German nationwide registry of standard-of-care cilta-cel, 2022-2025 OUTCOME | EARLIER LINE (1-3 prior lines, n = 177) | LATER LINE (>3 prior lines, n = 429) 12-month PFS: 79% (95% CI, 71-87) | 70% (95% CI, 65-76) Overall response rate: 91% | 88% Complete response: 63% | 54% Non-ICANS neurotoxicity: 3% | 8% Maintained CR was associated with 100% PFS at 12 months in both cohorts Gagelmann N, et al. J Hematol Oncol. 2026;19:46. doi:10.1186/s13045-026-01806-6 Nonrandomized registry analysis. The investigators did not perform formal comparisons between cohorts.

International Myeloma Society Global heterogeneity CAR-T access (incl lymphoma and myeloma) blood — Plenary Paper IMMUNOBIOLOGY AND IMMUNOTHERAPY Global access to commercial CAR T-cell therapies: a cross-sectional study of health technology assessment across the G20 countries Alex T. Ga, William B. Feldman, Martin F. Kaiser, Kai Rejeski, Gloria Iacoboni, Gaurav Narula, Jason Yongsheng Chan, Michael J. Dickinson, Aaron S. Kesselheim, and Edward R. Scheffer Cliff [affiliations partially illegible: Department of Medicine, UMass Chan Medical School, Worcester, MA; Program on Regulation, Therapeutics, and Law, Division of Pharmacoepidemiology and Pharmacoeconomics, ... Brigham and Women's Hospital, Harvard Medical School, Boston, MA; Division of Genetics and Epidemiology, The Institute of Cancer Research, London, United Kingdom; Adult Bone Marrow Transplant ...] [Diagram] HTA-cited barriers to CAR T access/analysis: • Single lump-sum payment for infusion; payment regardless of clinical outcome → Greater sensitivity to price/market share which in turn leads to uncertainty in projected economic impact • Single-arm studies → Naive comparisons limit analyses of comparative effectiveness, safety, quality of life, etc • Immature survival data with small n → Extrapolation of OS and additional assumptions are required when estimating ICERs • Increased incidence of grade ≥ 3 AEs (eg, CRS, ICANS) and hospitalizations → Additional costs for treatment of adverse events are often underestimated • Other (ethical, logistical, organizational issues) → Manufacturing time, administration in specialized centers, patient costs, and transportation are not effectively addressed by HTA models [World map with number of approved CAR T products per country: US 18, Canada 12, UK/EU region 5/14/7/9/14/9/10, Japan 11, China 2, Australia 7, Brazil 0, others shown 4; color scale 0-18] ICR The Institute of Cancer Research — Presented by: Martin Kaiser, MD, FRCP, FRCPath @MyMKaiser Content of this presentation is property of the author. Permission required for use International Myeloma Society Getting the right treatment to the right patient: Using diagnostics to personalise care and maximise treatment value? [Flow diagram] GEP risk* + IMS-IMWG (NGS) + WB-MRI → Enhanced Risk Assessment at 1st relapse Bone marrow NGS MRD + WB-MRI Imaging MRD → Personalised MRD tracking Enhanced Risk Assessment + Personalised MRD tracking = Joint decision making — Choice essential! Patient factors/preference → Treatment selection, e.g.: Option: TCE+CD38¹ | Option: CAR-T² | Option: ADC triplet | Option: Trad triplet — etc → e.g. for High-Risk MM: Continued treatment* → e.g. for Standard Risk + MRD-neg: Treatment Stop* optional Enhanced remission monitoring PB Mass-Spec 1 Banerjee R, et al, EHA 2026 2 Costa L, et al, ASH 2025 *Kaiser M et al, JCO 2023; Leypoldt L JCO 2024; Beer S et al, Blood 2026; Kaiser M et al, Lancet Onc 2026; Kumar S et al, NEJM 2026 UK patient-centric diagnostic pathway in development by UKMS Presently accessible: MyTrack Myeloma, others... UK Myeloma Society [International] Myeloma Society Summary • Limited access to new treatments can have multiple reasons • In high-income countries, not all reasons are linked to treatment cost • Practice-relevant data is increasingly important in context of HTA • Diagnostics can play an important part in defining treatment value moving forward - 'right treatment for the right patient' • Joint efforts and collaboration (academic trial groups and networks) can be valuable for access to best treatments

[Slide 1 of 2 in composite photo] MRD state and relapse after sMRD24 Cumulative Incidence of Relapse by MRD State Y-axis: Relapse probability (0.00–1.00); X-axis: Months from first MRD negativity (0–96) NGS-neg or LOD0: aHR (95% CI) Reference NGS-LOD+ or NGF-neg: 8.02 (4.25–15.10) p < 0.0001 Callout: MRD confirmation as NGS-neg or LOD0 even once significantly decreases relapse risk after sMRD24 Number at risk (number censored): Neg/LOD0: 280 (0) | 280 (0) | 280 (2) | 242 (41) | 189 (98) | 120 (153) | 74 (191) | 48 (214) | 21 (242) LOD+/NGF: 150 (0) | 150 (0) | 150 (0) | 138 (4) | 116 (16) | 91 (32) | 57 (49) | 43 (63) | 17 (88) 16 --- [Slide 2 of 2 in composite photo] Flow vs NGS sensitivity Flow-neg samples with simultaneous NGS (N=232) [Pie chart:] NGS-neg 47% | NGS-pos 16% | LOD+ 22% | LOD0 15% [Table:] | Relapse | Relapse or MRD Resurg NGS-neg or LOD0: 0.7% | 1.4% NGS-pos or LOD+: 7% | 16% 23rd International Myeloma Society Annual Meeting | 12 International Myeloma Society

International Myeloma Society 23rd Annual MEETING & EXPOSITION — September 23-26, 2026 · Glasgow, Scotland Five-Year Survival Outcomes among Patients with Multiple Myeloma in Tanzania, A Multicentre Retrospective Study LBA-05 J.NYAMATAGA¹, M.YONAZI², A. KILUNGU³, L.MBAGA⁴, H.IDDY⁵ and C.CHAMBA¹ 1 Muhimbili University of Health and Allied Sciences (MUHAS), Dar es Salaam, Tanzania. 2 Muhimbili National Hospital (MNH), Dar es Salaam, Tanzania 3 Mbeya Zonal Referral Hospital (MZRH), Mbeya, Tanzania, 4 Kilimanjaro Christian Medical Centre (KCMC), Kilimanjaro, Tanzania 5 Ocean Road Cancer Institute (ORCI), Dar Es Salaam, Tanzania INTRODUCTION Multiple myeloma (MM) remains a major cause of cancer-related morbidity and mortality worldwide. While survival has significantly improved in high-income countries following the introduction of modern novel therapies and autologous stem cell transplantation, long-term outcomes in sub-Saharan Africa remain poorly characterized with limited data available. AIM To evaluate five-year overall survival (OS) and identify patient-, disease-, and treatment-related factors associated with mortality among adult MM patients treated across tertiary referral hospitals in Tanzania. METHOD We conducted a multicenter retrospective cohort study across four tertiary referral facilities in Tanzania (Muhimbili National Hospital [MNH], Kilimanjaro Christian Medical Centre [KCMC], Mbeya Regional Referral Hospital [MZRH], and Benjamin Mkapa Hospital [BMH]). Ethical approval No. MUHAS-REC-05-2025-2853 was obtained from the MUHAS Institutional Review Board (IRB), alongside data collection permits from all participating study sites. Adult patients aged 18 years and above diagnosed with MM as per International Myeloma Working Group (IMWG) criteria between 2012 and 2025 were included. Overall survival was estimated using the Kaplan-Meier method and log-rank tests. Univariable and multivariable Cox proportional hazards regression models were used to identify independent predictors of mortality. RESULTS Out of 13,004 records screened, 504 fulfilled the eligibility criteria and were included in the final analysis. The median age was 58 years (IQR: 51-65), 55.2% were male, and 61.7% had health insurance. Overall, the 5-year OS rate was 20.1% (95% CI: 13.5-24.5%) with a overall median survival time of 13.9 months (95% CI: 11.0-16.9). Patients treated with bortezomib-based regimens demonstrated a significantly higher 5-year OS compared to those receiving melphalan or other traditional regimens (21.9% vs. 15.1%, log-rank p = 0.006; median survival 19.3 vs. 13.3 months, p < 0.001). Completing ≥ 8 chemotherapy cycles was associated with marked survival improvements compared to <8 cycles (5-year OS: 41.7% vs. 12.2%, log-rank p < 0.001; median survival 46.3 vs. 10.9 months). In multivariable Cox regression, independent predictors of increased mortality included male sex (aHR = 1.56, 95% CI: 1.12-2.17, p = 0.009), lower educational attainment (primary vs. College: aHR = 2.06, 95% CI: 1.29-3.30, p = 0.003), anemia (HB < 10 g/dl): aHR = 2.09, 95% CI: 1.40-3.12, p < 0.001). Concurrent antiviral administration was protective against mortality (aHR = 0.45, 95% CI: 0.23-0.92, p = 0.028). Table 1: Baseline Socio-Demographic and Clinical Characteristics (N=504) Age Group: <60 years 275 (54.6%), ≥ 60 years 229 (45.4%) Sex: Male 278 (55.2%), Female 226 (44.8%) Education: No formal education 16 (3.2%), Primary level 276 (54.8%), Secondary level 110 (21.8%), College / University 102 (20.2%) Health Insurance: Insured 311 (61.7%), Uninsured 193 (38.3%) Common Symptoms: Low back pain 295 (58.5%), Generalized weakness 199 (39.5%), Bone pain 196 (38.9%), Anemia 131 (26.0%), Fatigue 97 (19.2%) Laboratory Parameters: Hemoglobin <10g/dL 341 (69.5%), Serum Creatinine (Elevated) 317 (65.5%), Serum Albumin (Low) 238 (64.0%), beta_2-Microglobulin (Elevated) 298 (97.4%), High M-protein 317 (90.1%) ECOG Performance Status: 0-1 309 (62.8%), ≥ 2 183 (37.2%) Figure 3: Kaplan-Meier Survival Curve showing Five-Year Overall Survival of study participants — Five-Year OS=20.1%, 95% CI: 13.5-24.5 Figure 4: Five-Year Overall Survival by ECOG Performance status at diagnosis — ECOG <2 OS=23.9% (95% CI: 17.2-30.4); ECOG ≥2 OS=8.3% (95% CI: 3.5-15.6); Log rank test P = 0.001 Figure 6: Five-Year Overall Survival by Initial Chemotherapy regimen received — Bortezomib Group Five-Year OS: 21.9% (95% CI: 14.5-30.3); Melphalan & other Five-Year OS: 15.1% (95% CI: 9.5-21.8); Log rank test P = 0.006 Figure 5: Five-Year Overall Survival by number of chemotherapy cycles received — Five-Year OS in Cycle <8: 12.2% (95% CI: 7.5-18.1); Five-Year OS in Cycle 8+: 41.7% (95% CI: 22.8-60.3); Log rank test P <0.001 CONCLUSIONS Long-term survival among MM patients in Tanzania remains low, largely driven by late presentation, socioeconomic disparities, and inconsistent treatment continuation. Expanding access to proteasome inhibitors (bortezomib), improving supportive care, ensuring insurance coverage, and establishing national registries are vital strategies to optimize outcomes. REFERENCES • Mohty M, Terpos E, Mateos MV, et al. Multiple Myeloma Treatment in Real-world Clinical Practice: Results of a Prospective, Multinational, Noninterventional Study. Clin Lymphoma Myeloma Leuk. 2018;18(10):e401-e419. • Manyega KM, Lotodo TC, Oduor MA, et al. Retrospective Analysis of Presentation, Treatment, and Outcomes of Multiple Myeloma at a Large Public Referral Hospital in Eldoret, Kenya. JCO Glob Oncol. 2021;(7):391-399. • Desai K, Iqbal S, Rather M, Thirumaran R. Has the Survival in Patients with Multiple Myeloma Changed with the Advent of Latest Treatments? Blood. 2023;142(Suppl 1):6655. • Raza FM, Mkwizu EW, Leak SA, Sadiq AM, Lyamuya FS, Chamba NG, et al. Factors Associated With Survival Among Patients With Multiple Myeloma in Northeastern Tanzania. JCO Glob Oncol. 2024;(10):e2300388. • Rajkumar SV, Dimopoulos MA, Palumbo A, et al. International Myeloma Working Group updated criteria for the diagnosis of multiple myeloma. Lancet Oncol. 2014;15(12):e538-e548. ACKNOWLEDGEMENT MUHAS; Mentors; Study centres focal persons CONTACT INFORMATION Email: josephkurati@gmail.com

[Slide 1 of 2 in composite photo] Flow Between BCMA and GPRC5D directed therapy Flow between BCMA-, GPRC5D-, and FcRH5-directed immunotherapy exposures across successive lines of immune-effector-cell (IEC) therapy True Dual Exposure cohort (N = 151) Columns: 1st IEC exposure | 2nd IEC exposure | 3rd IEC exposure | 4th+ IEC exposure [Sankey diagram; legible node labels:] No further exposure 44 (29%) [after 3rd IEC]; No further exposure 117 (77%); 4th+ (has more) [count partially legible: 3]4 (23%) Legend: BCMA | GPRC5D | FcRH5 | Combo (≥2 targets, same line) | No further exposure | 4th+ (has more) Footnote: Each patient's immunotherapy exposures are ordered by the sequence in which they were administered (bridging/non-IEC chemotherapy regimens are excluded from the sequence). "Combo" denotes a single regimen/line targeting more than one of BCMA, GPRC5D, or FcRH5 simultaneously, counted regardless of whether exposures were sequential. Band width is proportional to patient count. International Myeloma Society | 7 --- [Slide 2 of 2 in composite photo] Short PFS after Dual Exposure 4.3 mo Median PFS (95% CI 3.8–5.5) 64% 3-month PFS 38% 6-month PFS 17% 12-month PFS [Kaplan-Meier curve: Event-free probability (0–100%) vs Months from dual exposure (0–30)] Number at risk: All 116 | 38 | 15 | 9 | 4 | 1

PA-125 International Myeloma Society - 23rd Annual MEETING & EXPOSITION - September 23-26, 2026 . Glasgow, Scotland Cytomegalovirus Reactivation after BCMA CAR-T Cell Therapy in Relapsed/Refractory Multiple Myeloma: Incidence, Risk Factors, and Clinical Outcomes D. Shah*, H. Hashmi*, A. Mammadzadeh, S. Rajeeve, T. Farzana, R. Firestone, E. Jurgens, F. Maura, A. Lesokhin, C. Tan, G. Shah, N. Korde, H. Landau, M. Merz, M. Scordo, H. Hassoun, K. Maclachlan, U. Shah, M. Hultcrantz, A. Derkach, D. Nemirovsky, S. Giralt, S. Mailankody, S.Z. Usmani, Z. Shahid Memorial Sloan Kettering Cancer Center, New York, NY, USA *Contributed equally. Corresponding author: Z Shahid ABSTRACT # 2434740 BACKGROUND - Infectious complications are a leading cause of morbidity and mortality after BCMA-directed CAR-T therapy for relapsed/refractory multiple myeloma (RRMM). - CMV reactivation is well characterized after allogeneic stem cell transplant, where it independently predicts inferior survival; however, its epidemiology, risk factors and prognostic impact after BCMA CAR-T remain poorly defined. - Lymphodepleting conditioning, immunosuppressive treatment of CRS and ICANS, and prolonged B-cell aplasia with hypogammaglobulinemia uniquely predispose BCMA CAR-T recipients to viral reactivation. - CMV risk is established for bispecific antibodies in RRMM, but CAR-T reports remain limited to small, heterogeneous cohorts without multivariable analysis. OBJECTIVES To characterize, in a large single-center RRMM cohort: - Incidence and timing of CMV detection and CMV reactivation after BCMA CAR-T - Baseline and treatment-related risk factors for reactivation - Impact of CMV reactivation on overall (OS) and progression-free survival (PFS) - Patterns of CMV-directed therapy METHODS Design. IRB-approved retrospective cohort of 351 consecutive RRMM patients who received BCMA-directed CAR-T at Memorial Sloan Kettering Cancer Center between March 2018 and January 2026 and had >=1 post CAR T-cell infusion serum CMV DNA test. Definitions. CMV detection -> CMV DNA below the quantitation threshold (<34.5 IU/mL); CMV reactivation -> CMV DNA >=34.5 IU/mL. Clinically significant CMV infection (csCMVi) -> DNAemia requiring anti-CMV therapy, or CMV end-organ disease. Statistics. Cumulative incidence of CMV endpoints estimated with death as a competing risk; groups compared by Gray's test. Cause-specific Cox models estimated hazard ratios, with corticosteroid exposure, tocilizumab exposure and CMV reactivation modeled as time-dependent covariates; OS and PFS by Kaplan-Meier. CRS and ICANS subgroup analyses landmarked at 1 month post-infusion. Analyses in R v4.5.1. RESULTS - Of 351 patients, 19 (5.4%) developed CMV reactivation (>=34.5 IU/mL) and 9 (2.6%) detection below quantitation threshold. - Reactivation occurred early, at a median of 26 days after CAR T-cell therapy infusion (range 9 to 265 days), with a 12-month cumulative incidence of 5.5%. - Clinically significant CMV infection occurred in 12 patients (3.4%) and CMV end-organ disease in 4 (1.1%); 10 patients received CMV-directed therapy. - Corticosteroid (HR 34.4) and tocilizumab (HR 60.0) exposure for CAR-T toxicity were the strongest predictors of reactivation (both p <0.001) - Baseline IgG >=500 mg/dL was protective (HR 0.31, p = 0.037); any-grade ICANS identified a higher-risk group (7.3% vs 1.5% at 12 months, p=0.037). - CMV reactivation was not significantly associated with overall survival (HR 1.83, p=0.2) or progression-free survival (HR 1.41, p=0.3). Table 1. Baseline characteristics (N=351) Age at infusion, median (range), y: 65 (37-86) Age >=70 years: 113 (32%) Prior lines of therapy, median (range): 5 (1-14) Late CAR-T (>=4 prior lines): 291 (83%) Prior anti-BCMA therapy: 44 (13%) Prior stem cell transplant: 303 (86%) High-risk cytogenetics: 162 (46%) ISS stage III: 57 (19%) Cilta-cel: 223 (64%) IgG, median (range), mg/dL: 549 (43-6,770) IgG <500 mg/dL: 108 (45%) ALC, median (range), x10^9/L: 0.70 (0.10-3.19) ALC <0.2 x10^9/L: 7 (2.2%) Any-grade CRS: 270 (77%) Any-grade ICANS: 31 (8.8%) Corticosteroids for CRS/ICANS: 19 (5.4%) Tocilizumab for CRS/ICANS: 26 (7.4%) Percentages of N=351 unless noted. [footnote partially illegible] Figure 1. Cumulative incidence of CMV detection and reactivation CMV reactivation (>=34.5 IU/mL): 19 events - 3.7% (1 mo), 4.6%, 4.6%, 4.9% (6 mo), 5.5% (12 mo) CMV detection (<34.5 IU/mL): 9 events - 0.85%, 1.7%, 2.0%, 2.6% (6 mo), 2.6% (12 mo) Estimates at 1, 2, 3, 6 and 12 months after CAR-T infusion, with death as a competing risk Table 2. CMV Endpoints [Endpoint | n (%) | 12-month cumulative incidence] CMV detection (<34.5 IU/mL): 9 (2.6%) | 2.6% (1.3-4.7) CMV reactivation (>=34.5 IU/mL): 19 (5.4%) | 5.5% (3.4-8.3) Clinically significant CMV infection: 12 (3.4%) | 3.4% CMV end-organ disease: 4 (1.1%) | 1.2% Received CMV-directed therapy: 10 (2.8%) Median time to detection 70 days (range 12-330); to reactivation 26 days (range 9-265). End-organ disease: confirmed colitis (n=1) and retinitis (n=1), suspected colitis (n=1) and gastritis (n=1). [remainder of footnote illegible] 12-month cumulative incidence of reactivation was 2.5% (1.0-5.2) with any-grade CRS vs 0% without (p = 0.2) [partially illegible] Table 3. Survival Outcomes [Outcome | Estimate (95% CI)] Median OS: 47 months 12-month OS: 82% (78-86) 24-month OS: 68% (63-74) Median PFS: 14 mo (12-17) 12-month PFS: 54% (49-60) 24-month PFS: 40% (34-46) OS by CMV reactivation: HR 1.83 (0.85-3.96), p=0.2 PFS by CMV reactivation: HR 1.41 (0.77-2.61), p=0.3 CMV reactivation modeled as a time-dependent covariate. Median follow-up 25 months (21-26) by reverse Kaplan-Meier. Figure 2. Cause-specific hazard ratios for CMV reactivation [Characteristic | HR (95% CI) | p] BASELINE FACTORS (univariable) Age >=70 y (vs <70 y): 2.33 (0.95-5.74) | 0.067 Prior lines of therapy (per line): 1.02 (0.84-1.24) | 0.9 Late (4+ LOT) (vs early 1-3 LOT): 1.15 (0.33-3.94) | 0.8 Prior anti-BCMA therapy (vs none): 2.95 (1.06-8.20) | 0.060 Non-cilta-cel product (vs cilta-cel): 0.73 (0.26-2.03) | 0.5 ISS stage II (vs stage I): 0.74 (0.20-2.74) | 0.8 ISS stage III (vs stage I): 1.28 (0.39-4.16) | 0.8 IgG >=500 mg/dL (vs <500 mg/dL): 0.31 (0.10-1.00) | 0.037 TOXICITY MANAGEMENT (time-dependent) Corticosteroids for CRS / ICANS: 34.4 (13.8-85.7) | <0.001 Tocilizumab for CRS / ICANS: 60.0 (19.9-181) | <0.001 Squares are point estimates, bars 95% confidence intervals. Corticosteroid and tocilizumab exposure were modeled as time-dependent covariates. Figure 3. CMV reactivation by ICANS status No ICANS (n = 288) vs Any-grade ICANS (n = 28); Gray's test p = 0.037 1 mo: 0.69% vs 3.6%; 3 mo: 0.69% vs 3.6%; 6 mo: 1.1% vs 3.6%; 12 mo: 1.5% vs 7.3% Months from 1 month post-CAR-T infusion (landmark) Figure 4. CMV viral load trends in clinically significant CMV infection 10 patients who received CMV-directed treatment, from reactivation through day +100 (up to five measurements each). Gray upward triangles mark treatment start, black downward triangles treatment end. Points below the dotted line are under the 34.5 IU/mL quantitation threshold and are not exact values. DISCUSSION - 12-month CMV reactivation rate of 5.5% is lower than rates reported after allogeneic stem cell transplantation (20-70%) but higher than rates observed after autologous transplantation (<5%), reflecting distinct immunosuppressive profile of BCMA CAR-T therapy. - Immunosuppression from combined corticosteroid and tocilizumab use support targeted CMV monitoring, particularly in ones with prior ICANS. - Baseline hypogammaglobulinemia (IgG <500 mg/dL) prior to infusion was associated with CMV reactivation - CMV reactivation was not significantly associated with OS or PFS, though immunological modulation from CMV reactivation or myelosuppressive effects of antiviral therapy could indirectly impair CAR-T expansion and durability, contributing to worse efficacy outcomes over longer follow-up. - Findings from these study support risk-stratified CMV surveillance for patients receiving steroids and/or tocilizumab, particularly those with baseline hypogammaglobulinemia, and the need for prospective study of IVIG or letermovir prophylaxis. CONCLUSIONS - CMV reactivation after BCMA-directed CAR-T for RRMM is relatively low (5.4%), occurs early (within a month of CAR-T infusion), and rarely progresses to clinically significant infection or end-organ disease. - Immunosuppression for CRS and ICANS increases reactivation risk, with corticosteroid and tocilizumab exposure contributing to it. - Low baseline IgG increased risk, while CAR-T product, prior lines and disease stage did not. - Reactivation was not significantly associated with OS or PFS. REFERENCES 1. Brudno JN, Kochenderfer JN. Blood Rev. 2019;34:45-55. 2. Hakki M, et al. Bone Marrow Transplant. 2021;56:1644-1663. 3. Ljungman P, et al. Clin Infect Dis. 2017;64:87-91. 4. San-Miguel J, et al. N Engl J Med. 2023;389:335-347. 5. Jurgens E, et al. Blood Cancer J. 2025;15(1):189. 6. Vora SB, et al. Clin Infect Dis. 2021;73:e2098-e2106. CONTACT & ACKNOWLEDGEMENTS Zainab Shahid, MD, Memorial Sloan Kettering Cancer Center, New York, NY ([illegible]@mskcc.org). We thank the patients and families who participated and the MSK Myeloma and Cellular Therapy teams. Footer: PA-125 - Cellular and T cell engager Immunotherapy (real life studies and sequencing)

23rd Annual Meeting & Exposition — September 23-26, 2026 · Glasgow, Scotland Ciltacabtagene Autoleucel as Early Versus Late Line of Therapy for Patients With Relapsed/Refractory Multiple Myeloma: A Real-World Analysis A. Mammadzadeh, MD1 · H. Hashmi, MD1 · T. Sebastian, MD1 · S. Rajeeve, MD1 · T. Farzana, MPH1 · R. Firestone, MD, PhD1 · E. Jurgens, MD1 · K. C. Miller, MD1 · F. Maura, MD1 · A. Lesokhin, MD1 · C. Tan, MD1 · G. Shah, MD1 · N. Korde, MD1 · H. Landau, MD1 · M. Scordo, MD1 · H. Hassoun, MD1 · K. Maclachlan, MBBCh, PhD1 · U. A. Shah, MD, MS1 · M. Hultcrantz, MD, PhD1 · A. Derkach, PhD2 · D. Nemirovsky, MS2 · S. Giralt, MD1 · S. Mailankody, MBBS1 · S. Z. Usmani, MD, MBA, FACP, FASCO1 · M. Merz, MD1 1 Myeloma Service, Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, NY · 2 Department of Epidemiology & Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY Memorial Sloan Kettering Cancer Center — Abstract # 2436404 INTRODUCTION Ciltacabtagene autoleucel (cilta-cel) produces deep and durable responses in relapsed/refractory multiple myeloma (RRMM). Real-world evidence on whether earlier administration improves survival remains limited. AIM Primary: Compare progression-free survival (PFS) and overall survival (OS) after early (1-3 prior LOT) versus late (>=4 prior LOT) cilta-cel. Secondary: Compare efficacy and toxicity and identify independent baseline predictors of survival. METHODS Design: Single-center retrospective cohort of RRMM patients treated with commercial cilta-cel. Groups: Early, 1-3 prior LOT (n=50); late, >=4 prior LOT (n=173). Endpoints: PFS, OS, response, and Day +90 safety: CRS, ICANS, NINT, cytopenias, infections, and NRM. Analysis: Cox models evaluated survival predictors. Inverse probability weighting addressed cytogenetics, EMD, functionally high-risk status, and disease burden. CONCLUSIONS • Timing alone was not independently associated with OS or PFS. • Functionally high-risk disease predicted inferior OS. • High-risk cytogenetics and EMD predicted inferior PFS. • Prior BCMA exposure predicted shorter PFS and OS, supporting earlier cilta-cel use. CONTACT Aytaj Mammadzadeh, MD — mammada@mskcc.org Abbreviations: BCMA, B-cell maturation antigen; CI, confidence interval; CR, complete response; CRS, cytokine release syndrome; EMD, extramedullary disease; FHR, functionally high-risk; HR, hazard ratio; ICANS, immune effector cell-associated neurotoxicity syndrome; LOT, lines of therapy; NRM, non-relapse mortality; NR, not reached; ORR, overall response rate. RESULTS — Patient flow and outcomes Identified in MSKCC REDCap (n=351) → Received other CAR T-cell products (n=128) / Treated with cilta-cel (n=223) → Early: 1-3 prior LOT (n=50) / Late: >=4 prior LOT (n=173) → BCMA-naive (n=143) / BCMA-exposed (n=30) Table 1. Demographics and baseline disease characteristics Characteristic | Overall N=223 | Early LOT N=50 | Late LOT N=173 | p-value Age at infusion, median (range), y | 65 (38-86) | 65 (43-80) | 65 (38-86) | 0.5 Age >=70 y, n (%) | 72 (32%) | 16 (32%) | 56 (32%) | >0.9 Prior LOT, median (range) | 4 (1-12) | 3 (1-3) | 5 (4-12) | <0.001 Prior anti-BCMA, n (%) | 30 (13%) | 0 (0%) | 30 (17%) | 0.002 Functionally high-risk, n (%) | 121/208 (58%) | 25/45 (56%) | 96/163 (59%) | 0.7 High disease burden, n (%) | 26 (12%) | 4 (8%) | 22 (13%) | 0.4 High-risk cytogenetics, n (%) | 108/222 (49%) | 25/49 (51%) | 83/173 (48%) | 0.7 EMD: none | 162 (73%) | 37 (74%) | 125 (73%) EMD: paraskeletal | 18 (8.1%) | 3 (6.0%) | 15 (8.7%) EMD: extraosseous | 42 (19%) | 10 (20%) | 32 (19%) | >0.9 Key findings: • Earlier patients had fewer prior LOT • Prior BCMA exposure occurred only in the late group Table 2. Efficacy by timing of cilta-cel Outcome | Overall | Early LOT | Late LOT | p-value ORR, n (%) | 211 (95%) | 49 (98%) | 162 (94%) | 0.3 CR or better, n (%) | 150 (67%) | 37 (74%) | 113 (65%) | 0.9 Median PFS, mo (95% CI) | 33 (25-NR) | NR | 31 (17-NR) 12-mo PFS, % (95% CI) | 71% (65-78) | 86% (75-97) | 68% (61-76) | 0.06 Median OS, mo (95% CI) | 51 (42-NR) | 51 (—) | NR (42-NR) 12-mo OS, % (95% CI) | 87% (83-92) | 87% (77-98) | 87% (82-93) | 0.7 12-mo NRM, % (95% CI) | 5.5% (2.9-9.3) | 7.3% (1.8-18) | 5.0% (2.3-9.2) | 0.7 Key findings: • Response rates and 12-month OS were similar • PFS favored earlier use numerically (p=0.06) Table 3. Toxicity by timing of cilta-cel Safety outcome through Day +90 | Overall | Early LOT | Late LOT | p-value CRS, any grade, n (%) | 172 (77%) | 43 (86%) | 129 (75%) | 0.09 CRS, grade >=3, n (%) | 2 (0.9%) | 0 (0%) | 2 (1.2%) | >0.9 ICANS, any grade, n (%) | 14 (6.3%) | 3 (6.0%) | 11 (6.4%) | >0.9 ICANS, grade >=3, n (%) | 6 (2.7%) | 0 (0%) | 6 (3.5%) | 0.3 Non-ICANS neurotoxicity, n (%) | 22 (9.9%) | 5 (10%) | 17 (9.8%) | >0.9 D+90 cytopenia, any grade, n (%) | 126 (59%) | 29 (59%) | 97 (59%) | >0.9 D+90 cytopenia, grade >=3, n (%) | 36 (17%) | 10 (20%) | 26 (16%) | 0.4 D+90 infection, any grade, n (%) | 95 (43%) | 20 (40%) | 75 (43%) | 0.7 D+90 infection, grade >=3, n (%) | 35 (16%) | 5 (10%) | 30 (17%) | 0.2 Key findings: • No statistically significant toxicity differences • CRS was numerically higher in the early group (p=0.09) SURVIVAL ANALYSES A. PFS by timing: p=0.06 B. OS by timing: p=0.70 C. PFS by prior BCMA exposure: p<0.001 D. OS by prior BCMA exposure: p=0.005 Table 4. Multivariable models for OS and PFS Characteristic | N | OS Events | OS HR | OS 95% CI | OS p | PFS Events | PFS HR | PFS 95% CI | PFS p CAR-T timing: Early (1-3 LOT) | 44 | 5 | — | — | 0.8 | 8 | — | — | 0.14 Late (4+ LOT) | 134 | 21 | 1.17 | 0.43, 3.21 | | 49 | 1.71 | 0.80, 3.64 Age >=70 years: No | 122 | 17 | — | — | 0.5 | 41 | — | — | 0.7 Yes | 56 | 9 | 1.31 | 0.56, 3.09 | | 16 | 0.88 | 0.48, 1.61 Functionally high-risk disease: No | 77 | 6 | — | — | 0.001 | 20 | — | — | 0.2 Yes | 101 | 20 | 3.82 | 1.67, 8.74 | | 37 | 1.49 | 0.85, 2.61 High disease burden: No | 160 | 24 | — | — | 0.5 | 50 | — | — | 0.8 Yes | 18 | 2 | 0.63 | 0.15, 2.74 | | 7 | 1.13 | 0.50, 2.55 High-risk cytogenetics: No | 89 | 9 | — | — | 0.4 | 19 | — | — | 0.035 Yes | 89 | 17 | 1.47 | 0.64, 3.34 | | 38 | 1.98 | 1.04, 3.80 Extramedullary disease: No | 132 | 14 | — | — | 0.12 | 35 | — | — | 0.046 Paraskeletal only | 12 | 2 | 1.54 | 0.31, 7.63 | | 5 | 2.41 | 0.90, 6.46 Yes | 34 | 10 | 2.53 | 1.08, 5.96 | | 17 | 1.98 | 1.08, 3.63 Timing was not independently associated with OS or PFS. Functionally high-risk disease predicted inferior OS; high-risk cytogenetics and EMD predicted inferior PFS.

International Myeloma Society — 23rd Annual Meeting & Exposition — September 23-26, 2026 · Glasgow, Scotland Prior BCMA Exposure in Patients with Relapsed/Refractory Multiple Myeloma Treated with Teclistamab and Elranatamab from the IMWG Immunotherapy Database CR. Tan1, H. Mian2, CY. Huang3, M. Dave4, H. Einsele5, T. Steinbrunn5, R. Hajek6, S.Z. Usmani1, R. Parrondo7, S. Chhabra8, E. Kastritis9, J. Martinez-Lopez10, L. Garderet11, C. Nagarajan12, WJ. Chng13, A. Tso14, M. Htut15, S. Bal16, J. Du17, K. Kim18, T. Martin3, Y. Lin19, R. Popat20 [affiliations fine print: 1. Myeloma Service, Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, NY 2. Department of Oncology, McMaster University, Hamilton, ON, Canada 3. UCSF Helen Diller Family Comprehensive Cancer Center, San Francisco, CA 4. International Myeloma Foundation 5. Department of Internal Medicine II, University Hospital Wurzburg, Wurzburg, Germany 6. Department of Hematooncology, University of Ostrava, Ostrava, Czech Republic 7. Division of Hematology and Medical Oncology, Department of Medicine, Mayo Clinic, Jacksonville, FL 8. Division of Hematology and Medical Oncology, Department of Medicine, Mayo Clinic, Phoenix, AZ 9. Department of Clinical Therapeutics, National and Kapodistrian University of Athens, Athens, Greece 10. Departamento de Hematologia, Hospital 12 de Octubre, Complutense University, CNIO, Madrid, Spain 11. Service Hematologie, Hopital Universitaire Pitie-Salpetriere, Paris, France 12. Department of Haematology, Singapore General Hospital and National Cancer Centre Singapore 13. Department of Hematology and Oncology, Department of Medicine, University of Alabama at Birmingham [partially illegible] 14. [illegible] 15. City of Hope, Duarte, CA 16. [illegible] 17. Shanghai Changzheng Hospital, China 18. Samsung Medical Center, Seoul, South Korea / Sungkyunkwan University 19. Division of Hematology-Oncology and Blood and Marrow Transplantation Program, Mayo Clinic, Rochester, Minnesota 20. Department of Haematology, University College London Hospital, London, United Kingdom] PA-364 — International Myeloma Foundation · International Myeloma Working Group INTRODUCTION • BCMA-targeted therapies, including CAR T-cell therapy (CART), bispecific antibodies (bsAb), and antibody drug conjugates (ADC) have changed the treatment landscape of RRMM. • However, patients still relapse with limited subsequent treatment options. • Data on outcomes with teclistamab (tec) and elranatamab (elra) after prior BCMA-directed therapy is limited due to small cohorts. • In MajesTEC-1, 40 patients with prior BCMA-directed therapy (CART and ADC) subsequently received tec with ORR of 53%, median (m) duration of response (DOR) of 14.8 months, mPFS of 4.5 months, and mOS of 15.5 months. AIM • The International Myeloma Working Group (IMWG) Immunotherapy Working Group aimed to conduct a real-world analysis evaluating outcomes associated with teclistamab and elranatamab in a real-world cohort of patients treated with prior BCMA-directed agents. • Data was analyzed from the International Myeloma Working Group (IMWG) Immunotherapy Database METHOD • This is an international multi-center retrospective study including patients with RRMM treated with teclistamab and elranatamab. • Data was collected through the IMWG Immunotherapy Database from 9 countries (US, UK, Greece, Spain, Czech Republic, Germany, Singapore, Canada, and France). • Data was collected until 4/24/2026. • If patients received both teclistamab and elranatamab, outcome and safety data from the last bsAb received was included in the analysis. • For the BCMA-exposed group, subgroups based on prior BCMA-directed agent(s) received include the following: CART, ADC, bsAb, ADC + CART, and >=2 BCMA agents with last agent being bsAb. • Responses were determined by providers at the corresponding treatment centers according to the IMWG Uniform Response Criteria. • Patient characteristics were summarized by proportion or median (interquartile range [IQR] or range), as appropriate. • Kaplan-Meier methods and log-rank tests were used to summary and compare ORR, PFS, and OS. RESULTS • 451 patients (pts) with RRMM treated with tec or elra were included in this analysis. • 353 pts were in the tec-treated group: 241 without prior BCMA (BCMA-naive) and 112 with prior BCMA exposure (BCMA-exposed). • 98 pts were in the elra-treated group: 74 BCMA-naive and 24 BCMA-exposed. • 53/126 (42%) BCMA-exposed pts received a BCMA-directed agent immediately prior to tec or elra. • For 115 BCMA-exposed patients with available data, ORR for last prior BCMA-directed agent was 64% (>=VGPR 44%). • Median time from last dose of prior BCMA-directed therapy to commencing tec or elra was 8.3 months (range, 0.5-77.1). • Median follow-up was 11.8 months (range, 0.2-42.1). • For the entire cohort, mPFS was 9.6 months (95%CI, 7.4-13.2), and mOS was 26.1 months (95%CI, 19.8-NR). • ORR with tec or elra based on prior BCMA agent(s) received was as follows: ADC 58%, CART 41%, bsAb alone 33%, ADC + CART 56%, >=2 agents including bsAb 25%. Table 1. Baseline Characteristics (Overall N=451 | BCMA-naive N=315 | BCMA-exposed N=136) Median age, yr (range): 67 (32-95) | 68 (32-95) | 66 (33-88); Age >=75 yr, n (%): 99 (23) | 81 (27) | 18 (13) Male, n (%): 233 (52) | 151 (48) | 82 (60) Race, n (%): White 308 (83) | 209 (83.9) | 99 (81.8); Black 30 (8.1) | 20 (8.0) | 10 (8.3); Asian 27 (7.3) | 18 (7.2) | 9 (7.4); Other 5 (1.4) | 2 (0.8) | 3 (2.5); Unknown 81 | 66 | 15 ECOG — n/N (%): 0: 42/283 (14.8) | 28/202 (13.9) | 14/81 (17.3); 1: 144/283 (50.9) | 101/202 (50) | 43/81 (53.1); >=2: 97/283 (34.3) | 73/202 (36.1) | 24/81 (29.6) EMD, n/N (%): 96/300 (32) | 58/203 (28.6) | 38/97 (39.2) HRCA, n/N (%): 226/408 (55.4) | 150/280 (53.6) | 76/128 (59.4) ISS stage, n/N (%): I 107/317 (33.8) | 75/230 (32.6) | 32/87 (36.8); II 107/317 (33.8) | 77/230 (33.5) | 30/87 (34.5); III 103/317 (32.5) | 78/230 (33.9) | 25/87 (33.9) [some cells partially illegible] Median prior LOT (IQR): 6 (5-8) | 5 (5-6) | 8 (7-11) Refractory status, n/N (%): Triple class refractory 299/366 (81.7) | 185/236 (78.4) | 114/130 (87.7); Penta drug refractory 166/364 (45.6) | 89/236 (37.7) | 77/128 (60.2) Figure 1. Best Response to Bispecific Antibody Therapy: All patients (N=431) ORR 64% (PR 26.2%, VGPR 29.9%, CR/sCR 7.7%); BCMA-naive (N=301) ORR 71% (PR 27.2%, VGPR 35.3%, CR/sCR 8.3%); BCMA-exposed (N=130) ORR 48% (PR 23.8%, VGPR 17.7%, CR/sCR 6.2%) Fig 2. PFS on tec or elra based on prior BCMA exposure: median PFS no prior BCMA 16.1 mo vs prior BCMA 2.7 mo; HR, 1.95 (95%CI, 1.52-2.49); P<0.001 [CI of medians partially illegible] Fig 5. OS on tec or elra based on prior BCMA exposure: HR, 1.84 (95%CI, 1.39-2.45); P<0.001 [medians partially illegible] Fig 3. PFS on tec or elra based on prior BCMA agent(s): P=0.49 Fig 6. OS on tec or elra based on prior BCMA agent(s): P=0.15 Fig 4. PFS by time from last BCMA agent to tec or elra (< 9 months vs 9+ months): HR, 0.61 (95%CI, 0.40-0.93); P=0.02 Fig 8. OS by time from last BCMA agent to tec or elra: HR, 0.66 (95%CI, 0.42-1.05); P=0.08 [partially illegible] CONCLUSIONS • Outcomes for patients receiving teclistamab or elranatamab after prior BCMA exposure is inferior compared to BCMA-naive patients, particularly in pts with prior BCMA bsAb although numbers are small. • PFS was significantly better in patients with >=9-month interval between last BCMA agent and tec or elra. • There was a trend toward improved OS in patients with >=9-month interval between last BCMA agent and tec or elra. REFERENCES 1. Moreau P, Garfall AL, van de Donk NWCJ, Nahi H, San-Miguel JF, Oriol A, et al. Teclistamab in Relapsed or Refractory Multiple Myeloma. New England Journal of Medicine. 2022;387(6):495-505. 2. Lesokhin AM, Tomasson MH, Arnulf B, Bahlis NJ, Miles Prince H, Niesvizky R, et al. Elranatamab in relapsed or refractory multiple myeloma: phase 2 MagnetisMM-3 trial results. Nature Medicine. 2023. 3. San-Miguel J, Dhakal B, Yong K, Spencer A, Anguille S, Mateos M-V, et al. Cilta-cel or Standard Care in Lenalidomide-Refractory Multiple Myeloma. New England Journal of Medicine. 2023;389(4):335-47. 4. Rodriguez-Otero P, Ailawadhi S, Arnulf B, Patel K, Cavo M, Nooka AK, et al. Ide-cel or Standard Regimens in Relapsed and Refractory Multiple Myeloma. New England Journal of Medicine. 2023;388(11):1002-14. 5. Touzeau C, Mina R, Quach H, Hungria V, Bhutani D, Chen W, et al. Teclistamab in Multiple Myeloma with One to Three Previous Lines of Therapy. New England Journal of Medicine. [year illegible] 6. Kumar S, Paiva B, Anderson KC, Durie B, Landgren O, Moreau P, et al. International Myeloma Working Group consensus criteria for response and minimal residual disease assessment in multiple myeloma. Lancet Oncol. 2016;17(8):e328-e46. ACKNOWLEDGEMENT We would like to thank the patients and their family and caregivers. We would like to acknowledge the investigators of the IMWG Immunotherapy Working Group Committee and their clinical and research team. Additionally, we would like to acknowledge the International Myeloma Foundation for its support of this project and the IMWG Immunotherapy Database. CONTACT INFORMATION Carlyn Tan: tanc4@mskcc.org Rakesh Popat: rakesh.popat@ucl.ac.uk

PA-059 MAYO CLINIC BRAIN FDG-PET AS AN IMAGING BIOMARKER OF CAR-T MOVEMENT AND NEUROCOGNITIVE TREATMENT-EMERGENT TOXICITY Ryan P Coburn1, Kenneth J C Lim2, Gemeng Zhang1, Nur Dizdar3, Derek R Johnson3, Melinda Tan2, Christoph Schaefers2, Hugo Botha1, Leland Barnard1, Anastasia Zekeridou1, Yi Lin2, Michel Toledano1 1Mayo Clinic Department of Neurology, 2Mayo Clinic Division of Hematology, Department of Medicine, 3Mayo Clinic Department of Radiology OBJECTIVE Establish brain fluorodeoxyglucose positron emission tomography (FDG-PET) as an imaging biomarker of chimeric antigen receptor T-cell therapy (CAR-T)-associated movement and neurocognitive treatment-emergent toxicity (MNT). BACKGROUND • MNT is a neurocognitive and hypokinetic movement disorder occurring after BCMA-targeting CAR-T therapy for multiple myeloma1. • Diagnosis is challenging due to phenotypic heterogeneity. • Described risk factors include pre-treatment myeloma burden and post-treatment high absolute lymphocyte count (ALC), but no established biomarkers exist2. • Prior FDG-PET descriptions report frontal striatal hypometabolism3-4. METHODS Cohort • 13 MNT cases identified from prospective IEC program compliance database (N=330) • 39 neurologically intact, high-ALC (>3x10^9/L) myeloma patients served as comparison cohort Imaging • Pre-/post-treatment whole-body FDG-PET processed via institutional pipeline to generate age-matched ROI Z-scores5 • MIMneuro (GE Healthcare) used as commercial comparator to assess generalizability Statistical Analysis • Linear mixed-effects models: ROI Z-scores (pre- vs. post-treatment)6 • Fixed effects: time x ROI x group interaction; subject as random intercept (repeated measures) • FWER correction for multiple comparisons (alpha=0.05)7 RESULTS Brain FDG-PET Z-Score Map Figure 1: A) Tri-planar Z-score map of a representative MNT case generated using our institutional processing method (pre-treatment (left) and post-treatment (right)). There is notable post-treatment basal ganglia hypometabolism (red arrows). Color scale: Warm colors represent an increased z-score, cool colors represent a decreased Z-score in comparison to a normative database. B) Biplanar Z-score map from the same MNT case generated using MIMneuro (pre-treatment (left) and post-treatment (right)). Similarly, notable post-treatment basal ganglia hypometabolism is appreciated (orange arrows). Color scale: Color gradient from light blue (Z-score -1.5) to magenta (Z-score -4) represents decreasing metabolism in comparison to a normative database. FDG-PET Metabolic Change Post CAR-T Therapy Institutional Pipeline — Region of Interest | Delta Z-Score | p-value: Right Putamen -2.14, 5.13e-06; Left Putamen -2.09, 1.08e-05; Right Pallidum -1.90, 1.51e-04; Left Pallidum -1.86, 2.92e-04; Right Caudate -1.78, 7.36e-04; Left Caudate -1.72, 1.59e-03 MIMneuro Pipeline — Region of Interest | Delta Z-Score | p-value: Left Putamen -2.03, 9.66e-08; Right Putamen -1.87, 1.92e-06; Right Caudate -1.43, 2.60e-03; Left Caudate -1.32, 1.04e-02; Right Pallidum -1.23, 3.28e-02 Table 2: Regional FDG-PET Z-score changes from pre- to post-treatment for basal ganglia regions that remained significant after family-wise error rate (FWER) correction in the MNT group using both imaging pipelines. Several frontal ROIs also demonstrated significant post-treatment metabolic decreases (not shown in above table). Abbreviations: FWER, Family-Wise Error Rate; ROI, Region of Interest; MNT, Movement and Neurocognitive Toxicity CONCLUSIONS • Post-CAR-T MNT patients show significant metabolic decrease in basal ganglia and frontal regions on FDG-PET, replicated across two processing pipelines — including MIMneuro, a widely available clinical tool • This regional hypometabolism pattern is consistent with clinical presentation • Findings are distinct from ICANS, which shows a characteristic orbitofrontal, dorsolateral frontal, and anterior cingulate cortex hypometabolism pattern8 • Basal ganglia damage and motor and frontal-striatal circuit disruption offer a unifying mechanism for the motor and cognitive dysfunction seen in MNT9 • FDG-PET is a candidate biomarker for post-CAR-T MNT, pending additional validation REFERENCES [fine print, largely illegible] (c)2026 Mayo Foundation for Medical Education and Research

To Declare Cure Requires convergence of three domains When All Three Align → Credible Declaration of Cure. Domain | Requirement Statistical | No excess mortality Biological | No detectable malignant clone Clinical | Long-term Treatment-free Disease-free survival International Myeloma Society Proposed Definition of Cure • Patient could be newly-diagnosed or with relapsed myeloma • Patient in sustained Complete Remission (CR) by standard criteria • MRD negative at 10⁻⁶ sensitivity by either NGS or NGF • MRD negativity Sustained for 5 years WITHOUT any therapy. (Irrespective of length of previous sustained MRD negativity on therapy) • There should be at least 4 MRD assessments in 5 years, last one being at 5 years; all negative and no positive test in between, if performed. • Negative by functional imaging (PET/CT or DWI WB MRI) • At least at the beginning of 5 years period and at the end of 5 years period with no intervening positivity, if performed. International Myeloma Society — 28

International Myeloma Society Proposed Definition of Cure - Patient could be newly-diagnosed or with relapsed myeloma - Patient in sustained Complete Remission (CR) by standard criteria - MRD negative at 10-6 sensitivity by either NGS or NGF - MRD negativity Sustained for 5 years WITHOUT any therapy. (Irrespective of length of previous sustained MRD negativity on therapy) - There should be at least 4 MRD assessments in 5 years, last one being at 5 years; all negative and no positive test in between, if performed. - Negative by functional imaging (PET/CT or DWI WB MRI) - At least at the beginning of 5 years period and at the end of 5 years period with no intervening positivity, if performed. International Myeloma Society — slide 28 International Myeloma Society Cure Fraction - Accumulated data suggests that a cure fraction is observed following various therapies and long-term follow up - Average calculated cure ranges between 5-30% - These numbers are with therapies utilized 10 years ago. We would predict a higher number with the current modern therapies - The cure fraction varies based on - Type of therapy - Age - Myeloma Risk categories International Myeloma Society — slide 29

International Myeloma Society Outline 1. Which therapy? - Single agent or more - Novel approaches 2. How long for? - Duration of therapy - Toxicity management 3. Personalisation? - Risk stratification - Patient preferences and quality of life ICR Images created using BioRender The ROYAL MARSDEN NHS Foundation Trust [Podium branding: International Myeloma Society 23rd Annual Meeting & Exposition]

Delphi Methods A modified 2-round Delphi panel Evidence review: HTA and literature inputs → Steering Committee: 4 UK experts; statement review → Round 1: 41 statements + 1 opened ended question → Controlled feedback: Aggregate results and comments → Round 2: 14 revised statements → Final synthesis: Consensus recommendations Multidisciplinary UK Expert Panel • The expert panel met pre-defined eligibility, participants were prioritised according to standardised expertise grading and geography • 29 HCPs completed ≥1 round including: 10 Haematologists (34.5%), 8 Clinical Nurse Specialists (27.6%), 8 Specialist Dietitians (27.6%) and 3 Haematology Pharmacists (10.3%) • All worked in teaching hospitals; most based in tertiary centres across the four UK nations Consensus was pre-defined and graded ≥75% agreement or disagreement required for consensus U 100% | A 91–99% | B 81–90% | C 75–80% Presented by R. Popat at the International Myeloma Society (IMS) 2026 Congress; September 23–26, 2026; Glasgow, United Kingdom

UKE Hamburg Conclusions - Quadruplet induction is currently approved for all NDMM patients - Quadruplet induction in the setting of consolidation and (combination) maintenance/continuous treatment allows the highest rates of MRD and sustained MRD negative responses which translates into the current most favorable PFS and OS outcomes - Sustained MRD negativity is the imminent prerequisite for stopping treatment and a potential curative approach - Achieving MRD negative response after induction allows to question the role of ASCT in standard-risk patients - Strategies including quadruplet induction show the lowest rates of functional high-risk disease - Carfilzomib based quadruplets show favorable MRD responses especially in high-risk patients - For comorbid, old and/or frail patients, triplet or doublet treatment should be considered - Whenever available and feasible, quadruplet treatment should be initiated in any NDMM patient Universitaeres Cancer Center HAMBURG Universitaetsklinikum Hamburg-Eppend[orf] [International] Myeloma Society Perspectives: which patient needs what? Columns: Standard risk, MRD-negative | Grey zone (e.g., t(11;14), MRD-positive) | High / very high risk Today: - Standard risk, MRD-negative: Anti-CD38-VRd -> Mel 200 + ASCT -> Dara-R - Grey zone: Anti-CD38-VRd -> Mel 200 + ASCT -> Dara-R - High / very high risk: Anti-CD38-VRd -> Mel 200 + ASCT -> Dara-R Tomorrow (if the trials are positive): - Standard risk, MRD-negative: Anti-CD38-VRd -> T-cell redirection (bispecific or CAR-T), no Mel 200 - Grey zone: Open question: is an alkylator still needed? Biomarker-driven trials - High / very high risk: ASCT + T-cell redirection (e.g., ASCT -> anti-BCMA CAR-T) Today: ASCT for all eligible patients. Tomorrow: not "ASCT or not", but which patient needs which. International Myeloma Society 23rd International Myeloma Society Annual Meeting 8 International Myeloma Society Conclusion 1. Which therapy? 2026: Lenalidomide + daratumumab 2030: Immunotherapy approaches? 2. How long for? 2026: To progression but..... - low threshold for stopping after 2-3 years in event of toxicity if no high risk lesions and MRD -ve - higher threshold for stopping/ try to continue if any high risk lesions or MRD +ve 2030: Immunotherapy approaches likely to be time limited 3. Personalisation? 2026: Double hit / IMWG-IMS high risk patients need intensive consolidation and multi-agent maintenance approaches 2030: More information on older patients and time limited therapy ICR Images created using BioRender The ROYAL MARSDEN NHS Foundation Trust

International Myeloma Society 3. Frailty impact on maintenance • Studies in transplant ineligible patients to date have continued treatment to progression • Ongoing studies are asking key questions in the context of current standard of care: Can we stop dara? Canadian Cancer Trials Group MY.13 Study NDMM TI, DRd x 18 cycles, >=PR → Rd to PD | DRd to PD Can we stop if MRD neg? iFIT UK-MRA Myeloma XVIII NDMM TI, DRd x 6 cycles, MRD neg → DR to PD | DR x 18 cycles Can we stop in all? HOVON-174 FABULOUS NDMM TI, DRd x 12 cycles, >=PR → DRd to PD | stop (restart before PD) ICR | CIs: Hira Mian, Charlotte Pawlyn/Gordon Cook, Sonja Zweegman | The Royal Marsden NHS Foundation Trust

Perspectives: which patient needs what? Columns: Standard risk, MRD-negative | Grey zone (e.g., t(11;14), MRD-positive) | High / very high risk Today: • Standard risk, MRD-negative: Anti-CD38-VRd → Mel 200 + ASCT → Dara-R • Grey zone: Anti-CD38-VRd → Mel 200 + ASCT → Dara-R • High / very high risk: Anti-CD38-VRd → Mel 200 + ASCT → Dara-R Tomorrow (if the trials are positive): • Standard risk, MRD-negative: Anti-CD38-VRd → T-cell redirection (bispecific or CAR-T), no Mel 200 • Grey zone: Open question: is an alkylator still needed? Biomarker-driven trials • High / very high risk: ASCT + T-cell redirection (e.g., ASCT → anti-BCMA CAR-T) Today: ASCT for all eligible patients. Tomorrow: not "ASCT or not", but which patient needs which. International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 8 Who might still need an alkylator? Profile | Why an alkylator might still matter | Evidence today t(11;14) | Overexpresses cyclin D1. BCL2 dependence -> cells more "primed" for apoptosis -> hypothesis: more sensitive to pro-apoptotic stress including melphalan | Hypothesis ? MIDAS small numbers t(11;14): n=32 Arm A, n=46 Arm B Very high risk (biallelic TP53, >= 2 CGS criteria, PCL) | Proliferative disease: cytotoxic pressure may complement T-cell engagement | Tandem ASCT case by case, no robust data Extramedullary disease, high tumor burden | Rapid debulking alongside cell therapy (bridging, CRS risk) | Hypothesis ? Antigen escape (BCMA loss) | Argues first for dual targeting (BCMA x GPRC5D, BCMA/CD19) rather than an alkylator | Being tested in the multi-target trials T-cell fitness, marrow microenvironment | Exhausted or immune-cold marrow: T-cell engagement may not suffice; Mel may modify the niche | Hypothesis: biomarker to test ? The new trials need prespecified biology-based subgroups and translational cohorts: MRD kinetics, T-cell fitness, target expression International Myeloma Society — 23rd International Myeloma Society Annual Meeting — 7

Memorial Sloan Kettering Cancer Center Treatment naïve GPRC5D negative multiple myeloma with t(11;14)(CCND1;IGH) Juan-Jose Garces1*, Holly Lee1,2*, Ross S Firestone1*, Ozgur Can Eren3*, Bachisio Zicchedu1, Christine Riedhammer4, Jan Frenking5, Kevin Herrera1, Kylee H Maclachlan1, Tjar Buchwald4, Irina Linkov3, Katherine Lopez3, Jesús Gutiérrez Abril1, Sarun Sereewattanawoot1, Eric Jurgens1, Kevin Han1, Carlyn Tan1, Eric Smith6, Paola Neri2, Benjamin Barwick7, Vikas Gupta7, Lawrence Boise7, Ahmet Dogan3, Issam Hamadeh1, Hermann Einsele4, Marc Raab5, Leo Rasche4, Michael J Kluk3#, Nizar Bahlis2#, Karlo Perica1#, Sham Mailankody1#, Saad Z. Usmani1# and Francesco Maura1* Introduction GPRC5D is one of the main actionable markers for immunotherapy development, particularly CAR T-cells and T-cell engagers. These have shown remarkable efficacy in relapsed/refractory multiple myeloma (RRMM), with response rates of up to 70-80%. However, a subset of treatment-naïve patients exhibits primary refractoriness due to the absence of GPRC5D expression, representing an uncharacterized population with distinct biological features. Methods - Exploratory dataset -> 549 newly-diagnosed MM patients (NDMM, CoMMpass) 1. RNAseq -> identify patients with low GPRC5D expression + functional characterization 2. WGS/WES -> genomic characterization - New immunohistochemistry (IHC) scoring system -> GPRC5D expression levels & %CD138+ expressing cells - 42 patients under Talquetamab (Tal) - Clinical validation dataset, - 138 patients treated with Tal - 30 patients under BCL-2 inhibitor (Venetoclax) - 26 RRMM patients with anti-GPRC5D therapies - NCT04555551 + NCT05431608 Conclusions 1. Identification of ~10% of NDMM patients with no GPRC5D expression -> no response to anti-GPRC5D CART/BsAb therapies. - This proportion increases in RRMM cohorts 2. Strong association of GPRC5D negativity with t(11;14), B-cell-like profile, and a genomically indolent MM -> potential refractoriness to anti-GPRC5D therapy. - But particularly sensitive to BCL-2 inhibitors. 3. New and straightforward IHC score to evaluate GPRC5D expression -> reproducible and cost-effective way to identify patients refractory to anti-GPRC5D therapies GPRC5D IHC expression is a robust biomarker to predict primary refractory patients to GPRC5D-directed therapies and to optimize treatment allocation. Results A) [box plots] Expression (Log +1), GPRC5D vs BCMA, p<0.0001 B) CCND1 translocation: GPRC5D (Log+1), WT vs CCND1, p<0.0001 C) MM molecular classification: GPRC5D (Log+1), CD1 vs CD2 vs Others, p<0.0001 / p<0.0001 D) Talquetamab treated RRMM: PFS, WT vs t(11;14)(CCND1;IGH), p = 0.25; at risk: 112 30 9 4 2 / 25 4 2 2 0 E) Heatmap rows: GPRC5D status, NDS2 translocation, B-cell-like, CCND1 translocation, GPRC5D RNA expression, MM molecular classification, Genomic classification. Legends: GPRC5D status (WT, Gain, Loss); GPRC5D RNA exp. (High, Low); MM molecular class. (CD1, CD2, HY, LB, MF, MS, PR); IRMMa genomic class. (CCND1_Complex_Cytog., CCND1_Simple, GainAmp1q_Del13q, HRD_Complex_Cytog., HRD_Gains, HRD_RAS, MAF_and/or_HyperAPOBEC, Multiple_Losses, NSD2_Del13q, NSD2_GainAmp1q_Del13q, NSD2_HRD, Simple); Scaled expression (4 to -4) Figure 1. Low GPRC5D characterization in NDMM (CoMMpass, n = 549). (a) GPRC5D and TNFRSF17 gene expression levels. (b) Patients harboring t(11;14) exhibit lower GPRC5D expression. (c-d) Among patients with t(11;14), the UAMS molecular expression group CD2 (c) and the CCND1 simple genomics subgroup (d) show lower GPRC5D expression. (e) Heatmap with expression of key genes included in the B-cell-like MM signature. WT, wild type. [Figure 2 IHC table] A) H&E / CD138 / GPRC5D; IHC score 0: Staining completely absent or barely perceptible = Negative; 1+: Weak staining of <=40% of CD138+ cells = Low; Weak staining of >40% of CD138+ cells = Intermed.; 2+ / 3+: Strong staining (2-3+) = High. Interpretation columns: Unlikely to benefit from GPRC5D-directed therapy / Highly likely to benefit from GPRC5D-directed therapy B) Talquetamab: PFS, p < 0.0001; at risk (Days): 18 0 0 0 / 24 13 6 2 C) GPRC5D-CART: PFS, p = 0.00017; GPRC5D Neg/Low vs GPRC5D Intermediate/High; at risk (Months): 22 16 10 8 5 2 1 / 4 0 0 0 0 0 0 Figure 2. MM patients with "Negative/Low" staining are refractory and have poor outcomes with Tal. (a) IHC-based scoring system to classify GPRC5D expression semi-quantitatively. (b-c) MM patients treated with Tal or CART, stratified by pre-treatment GPRC5D IHC score. 1. Maura. Nat Gen 2026 (in press) 2. Gupta et al., Blood 2021 3. Lee et al., Nature Medicine 2023 garcesj@mskcc.org / mauraf@mskcc.org

Frailty can improve and deteriorate [Sankey diagram of frailty category transitions across Baseline, Year 1, Year 2, Year 3] Legend: Nonfrail | Prefrail | Mildly Frail | Moderately Frail | Severely Frail | Deceased Baseline: Nonfrail 4.6% | Prefrail 24.8% | Mildly Frail 32.0% | Moderately Frail 22.8% | Severely Frail [illegible] Year 1: Nonfrail 3.5% | Prefrail 15.0% | Mildly Frail 21.1% | Moderately Frail 17.8% | Severely Frail [illegible] | Deceased 25.0% Year 2: Nonfrail 5.7% | Prefrail 17.8% | Mildly Frail 17.[illegible]% | Moderately Frail 10.8% | Severely Frail [illegible] | Deceased 39.7% Year 3: Nonfrail 6.7% | Prefrail 14.8% | Mildly Frail 13.0% | Moderately Frail 7.3% | Severely Frail [illegible] | Deceased 53.1% Mian et al. 2023 n = 4,617 93% changed frail category *those with 3[-]year data n= 2165 International Myeloma Society 20 [F]railty changed - but WHY? HOVON FITNESS Mian et al DYNAFIT VULNERABILITY: Reduced physiological reserve and resilience to stressors [Disease burden circle] - Disease burden - Anaemia - Renal dysfunction - Bone disease / pain [Treatment circle] - Toxicity - Infection / hospitalisation - Neuropathy - Steroids - Treatment intensity [Co-morbidity] - Cardiovascular disease - Respiratory disease - Diabetes - Other chronic conditions - Cumulative comorbidity burden [Social support] - Living situation - Carer support - Social isolation - Financial concerns - Access to care [Physical function] - Strength - Mobility - Balance - Physical activity - Sarcopenia [Mental health] - Cognitive impairment - Delirium risk - Depression / anxiety - Psychological distress [Polypharmacy] - Polypharmacy - Potentially inappropriate medications - Drug interactions - Treatment-related effects [Nutrition] - Weight loss - Protein / energy intake - Malnutrition - Sarcopenia Smits et al. 2023, Mian et al. 2023, Cook et al. 2023, Zhang et al. 2[cut off] International Myeloma Societ[y]

Helper and cytotoxic T cells cooperate to kill infected and cancerous cells through immunological synapses [Schematic: Helper T cell — IL-2 → Cytotoxic T cell (Immune attack) — Regulatory T cell, each forming synapses with target cells] International Myeloma Society — BioRender

International Myeloma Society Annual Number of Multiple Myeloma Clinical Trials by Funding Source High-volume funders Legend: Industry (red) | Other (Academic/Hospital/Other) (blue) Y-axis: Number of Clinical Trials (0 to 60) X-axis: Year, 2006 to 2026 [Line chart: Industry line rises from ~2 in 2006 to a peak of ~63 around 2025, with a circled data point at ~46 and a sharp drop toward 2026 highlighted by a downward red arrow. Other (Academic/Hospital/Other) line rises from ~0 in 2006 to ~24 by 2024, then also drops sharply toward 2026. Exact per-year values not printed] 63

Attitudes towards the use of MRD as an early endpoint in MM MRD-negative status is the preferred endpoint in clinical trials for NDMM HCP [stacked bar: majority Agree/Strongly agree] REG [stacked bar: majority Strongly Disagree/Disagree/Undecided] p-value < 0.001 Insistence on proven PFS benefits delays access to innovative treatments for patients in MM HCP [stacked bar: majority Agree/Strongly agree] REG [stacked bar: majority Disagree/Undecided] p-value < 0.001 Axis: -100% to 100% Legend: Strongly Disagree | Disagree | Undecided | Agree | Strongly agree HCP = Healthcare professionals; REG = Regulators

International Myeloma Society IMS - SWOT STRENGTHS - Dedicated leadership - Clear mission - Our results (consensus building, advocacy work, research and education activities - Recognition form KOL, academic and cooperative groups, industry, patient advocacy groups, and regulatory agencies WEAKNESSES - Not yet enough activity in LMIC regions - Clinical trial designs - ASH, ASCO - Team is understaffed OPPORTUNITIES - Strategically growing and developing the society - LMIC - Creating new avenues of communicating our messaging/fundraising - Collaboration with IMWG : consensus meetings - « Young blood », GLO - Collaboration with EHA, AACR THREATS - Funding - Golden age for MM - Applicability of guidelines - Political situation 51

International Myeloma Society Attitudes towards the use of MRD as an early endpoint in MM MRD-negative status is the preferred endpoint in clinical trials for NDMM HCP REG p-value < 0.001 Insistence on proven PFS benefits delays access to innovative treatments for patients in MM HCP REG p-value < 0.001 [Diverging stacked bar chart; x-axis: -100% -50% 0% 50% 100%. For both statements the HCP bars skew toward Agree/Strongly agree and the REG bars skew toward Strongly Disagree/Disagree/Undecided; exact segment percentages not printed] HCP = Healthcare professionals REG = Regulators Legend: Strongly Disagree | Disagree | Undecided | Agree | Strongly agree

Attendee demographic details The 2026 IMS Annual Meeting is the most well-attended in IMS history! 77 Countries represented 3,840+ Total attendees 400+ Virtual attendees 146 Trainees 70 Media registrants from 42 outlets and 11 countries International Myeloma Society — 23rd Annual Meeting & Exposition

What Are the Challenges With Rechallenge? Sequencing within established pathways: - IMiD: Len -> Pom - PI: Bort -> Car - Anti-CD38: Dara/Isa -> Dara/Isa - New complexity: CELMoDs The emerging BCMA challenge: - When BCMA therapies share a target but not a modality, what defines rechallenge? - How can we optimise BCMA rechallenge by switching modality? - Once a patient relapses, when is the best time to rechallenge with BCMA? - How might prior BCMA exposure and time since BCMA-targeted therapy change outcomes? BCMA sequencing considerations: - Prior BCMA therapy: exposure, response, BCMA refractoriness - BCMA modality: ADC, BsAb, CAR-T - Target strategy: stay with BCMA, switch target, return to BCMA - Time away from BCMA: immediate sequence, intervening therapy, interval - Patient and pathway: T-cell fitness, access, eligibility, logistics Additional challenge: retaining the same modality while changing target (e.g., BCMA -> GPRC5D) Rechallenge outcomes can vary widely depending on several factors, so how can we optimise BCMA treatments to maximise outcomes for patients? This slide presents a discussion framework based on faculty clinical experience. Relevant supporting evidence is cited on subsequent slides. ADC, antibody-drug conjugate; BCMA, B-cell maturation antigen; bort, bortezomib; BsAbs, bispecific antibodies; car, carfilzomib; CAR-T, chimeric antigen receptor T-cell; CD38, cluster of differentiation 38; CELMoD, cereblon E3 ligase modulatory drugs; dara, daratumumab; GPRC5D, G protein-coupled receptor class C group 5 member D; IMiD, immunomodulatory drug; isa, isatuximab; len, lenalidomide; PI, proteosome inhibitor; pom, pomalidomide. 21 [Panel name plates visible: Jesus Berdeja, Douglas Sborov, Joseph Mikhael] 1L Use of Anti-CD38 Is Increasing, Resulting in mPFS ~2-8 Months in RRMM With Anti-CD38 Retreatment RWE studies* have shown a mPFS of 1.5-8.4 months with anti-CD38 retreatment RWE Study | N | % Dara/anti-CD38 ref† (% multi-ref) | Anti-CD38-based retreatment regimen | Median PFS, months Kikuchi, 2023 | 39 | 72% | Isa | 5.6 Kostopoulos, 2023 | 35 | 100% | Dara, IMiD | 5.0 Hulin, 2022 (EMMY) | 173 | 73% | Anti-CD38-based retreatment NS | 4.7 Zhou, 2022 | 38 | 97% (55% penta ref) | Dara, Carf-based regimen | 4.1 Kastritis, 2022 | 42 | 100% | Anti-CD38-based retreatment NS | 4.0 Gandhi, 2019 (MAMMOTH) | 57 | (100% triple ref) | Anti-CD38-based retreatment NS | 3.9 Nooka, 2019 | 12 | 100% | Dara, Pom, Dex | 3.3 Yashar, 2021 | 49 / 23 / 15 / 8 | NR | Anti-CD38-based retreatment 2L / 3L / 4L / 5L | 3.2 / 2.6 / 1.9 / 1.5 Decaux, 2022 (IMAGE) | 56 | 28% | Isa, Pom, Dex (Dara ref) | 3.0 | 22 | | Isa, Pom, Dex (non-Dara ref) | 8.4 [Bar chart axis: 0 to 10 months] Adapted from Gay F, et al. Front Oncol. 2025;15:1550644. Clinical outcomes with anti-CD38-based retreatment are variable and may offer limited survival benefit *Median PFS data were reported in nine RWE studies; †In overall population. #L, # line of therapy; carf, carfilzomib; CD38, cluster of differentiation 38; Dara, daratumumab; Dex, dexamethasone; IMiD, immunomodulatory drug; Isa, isatuximab; mPFS, median progression-free survival; NR, not reported; NS, not specified; PFS, progression-free survival; pom, pomalidomide; ref, refractory; RRMM, relapsed/refractory multiple myeloma; RWE, real-world evidence. Gay F, et al. Front Oncol. 2025;15:1550644. 22

[2026.eventscribe.net] Treatment of Newly Diagnosed Myeloma Friday Poster Session PA-399: (PA-399) Real-World ISS-R-ISS Discordance, Extramedullary Disease Biology, and Treatment Patterns in Newly Diagnosed Multiple Myeloma: A North Indian Public Sector Cohort Friday, September 25, 2026 Abhishek Yadav, MBBS Junior Resident Vardhman Mahavir Medical College & Safdarjung Hospital Introduction: Multiple myeloma (MM) accounts for 10% of haematological malignancies globally; 188,000 new cases in 2022, with a projected 71% rise by 2045. Indian public sector data integrating staging concordance, serum free light chain (SFLC), fluorescence in situ hybridisation (FISH), and treatment patterns remain scarce. Methods: [cut off]

MRD and Biomarkers OA-55: Poor Outcomes Post Sequential BCMA and GPRC5D Exposed Myeloma Can Be Improved by Subsequent Targeted Agents Versus Conventional Therapy: An IMWG Immunotherapy Registry Analysis Wednesday, September 23, 2026 3:06pm - 3:18pm BST Location: Lomond Auditorium Abstract Presenter(s) Murali Janakiram, MD, MS Associate Professor, Department of Hematology & Hematopoietic Cell Transplantation City of Hope Duarte, California [browser bar] 026.eventscribe.net

International Myeloma Society 23rd Annual MEETING & EXPOSITION Sept. 23-26, 2026 - Glasgow, Scotland Myeloma Novel Drug Targets and Agents Wednesday Poster Session PA-304: (PA-304) Mezigdomide (CC-92480) Following Idecabtagene Vicleucel (Ide-Cel) in Relapsed/Refractory Multiple Myeloma Is Safe and Leads to Deepening of Responses Wednesday, September 23, 2026 Murali Janakiram, MD, MS Associate Professor, Department of Hematology & Hematopoietic Cell Transplantation City of Hope Introduction: Idecabtagene vicleucel (ide-cel), a BCMA-directed CAR T-cell therapy, achieves high response rates in relapsed/refractory multiple myeloma (RRMM), yet most patients relapse. Immune rec[obscured] infusion represents a window to de[obscured] responses. Mezigdomide (CC-92480) is an oral CELMoD with potent anti-myeloma activity that can stimulate anti-tumor immunity and synergize with residual CAR T-cells. Translational work from this study reported separately shows that mezigdomide use after [illegible] is associated with [illegible] immune remodeling, with a trend toward decrease in the [cut off] [browser bar] imsannual2026.eventscribe.net

METHODS/ STUDY DESIGN Key Eligibility Criteria: Age ≥18 years ECOG PS 0-2 NDMM Transplant-eligible ≤1 cycle of prior induction therapy Phase 2 Treatment Dosing: CFZ 20 mg/m2 IV on C1D1, then 56 mg/m2 on C1D8 and D15, then 56 mg/m2 C2-4 D1, D8, and D15 Iberdomide 1 mg orally on D1-21 DEX 40 mg if ≤75 years or 20 mg is >75 years on D1, 8, and 15 KID induction regimen given in 28-day cycles for 2-3 cycles followed by ASCT Phase 2 Primary Endpoint: Rate of CR + sCR Secondary Endpoints: ORR, MRD-negativity, PFS, OS, TEAEs Figure 1. Updated Treatment Schema Cycle 1: 20/56 mg/m2 D1, D8, D15 Cycle 2*: 56 mg/m2 D1, D8, D15, D22 Cycle 3*: 56 mg/m2 D1, D8, D15, D22 Cycle 4*: 56 mg/m2 D1, D8, D15, D22 Iberdomide dosing shown across cycles Dexamethasone 40 mg or 20 mg for patients >75 years old Iberdomide: 1mg** -> Autologous stem cell transplantation -> Follow-up every 3 months for 2 years, or until PD or start of a new line of therapy *Stem cell mobilization can occur after the 2nd, 3rd, or 4th cycle determined per physician discretion and institutional guidelines **Iber 1.6 mg was initially selected as the phase 2 dose (P2D). Following emerging data from BMS trials on 3/17/2025, the P2D was amended to 1 mg. Table 2. Response Criteria | At end of induction N=42, n (%) | 3-months post-ASCT N=36, n (%) | Best response N=58, n (%) | High-risk best response N=28, n (%) | sCR: - | 2 (6) | 2 (3) | 1 (4) CR: 5 (12) | 9 (25) | 16 (28) | 8 (29) VGPR: 21 (50) | 20 (56) | 12 (21) | 7 (25) PR: 15 (36) | 4 (11) | 27 (47) | 11 (39) SD: 1 (2) | 1 (3) | 1 (2) | 1 (2) ORR: 41 (98) | 35 (97) | 57 (98) | 27 (96) Table 3. Treatment-Emergent Adverse Events (Grade III-V) Adverse Event | # of Patients | Grade Acute Kidney Injury | 1 | 3 Anemia | 3 | 3 Atrial Fibrillation | 2 | 3 Blood per Rectum | 1 | 3 Constipation | 1 | 3 Death | 1 | 5 Dizziness | 1 | 3 Electrolyte abnormalities | 2 | 3 Fatigue | 2 | 3 Hypovolemia | 1 | 4 Liver enzyme abnormalities | 2 | 3 Maculopapular rash | 2 | 3 Mood, sleep/wake cycle | 2 | 3 Myalgias | 2 | 3 Neutropenia | 11 | 3 (n=11), 4 (n=12) Syncope | 1 | 3 Thrombocytopenia | 3 | 3 CONCLUSIONS Induction therapy with KID leads to deep and durable responses regardless of high-risk status, despite the short treatment duration. Notably, iber-based induction also demonstrated successful stem cell mobilization. Further studies will evaluate KID with post-ASCT iber + elranatamab consolidation followed by iber maintenance.