
Corporate Presentation March 2026 Conditionally Active Biologics: Transforming Cancer Therapy

Important Notices & Disclaimers This presentation (the “Presentation”) by BioAtla, Inc. (“we”, “us”, “our”, “BioAtla”, or the “Company”) contains “forward-looking statements” within the meaning of the Private Securities Litigation Reform Act of 1995 relating to our business, operations and financial conditions, including but not limited to statements we make regarding BioAtla’s business plans and prospects; whether our clinical trials will support registration; achievement of milestones; results, progress and timing of our research and development programs and clinical trials; expectations with respect to enrollment and dosing in our clinical trials; the anticipated clinical benefits, safety, efficacy, and market potential of our product candidates; plans and expectations regarding future data updates, clinical trials, regulatory meetings and regulatory submissions; the timing of and the ability to establish collaborations or other strategic partnerships for selected assets; the potential regulatory approval path for our product candidates; and expectations about the sufficiency of our cash and cash equivalents to fund operations and expectations regarding R&D expenses and cash burn. Words such as, but not limited to, “anticipate”, “believe”, “could”, “estimate”, “expect”, “intend”, “may”, “plan”, “potential”, “predict”, “project”, “should”, “will”, “would” or the negative of those terms, and similar expressions that convey uncertainty of future events or outcomes, identify forward-looking statements. These forward-looking statements reflect management’s beliefs and views with respect to future events and are based on estimates and assumptions as of the date of this Presentation and are subject to risks and uncertainties, including those described in the Company's filings with the SEC, including but not limited to the Company's latest Annual Report on Form 10-K and any subsequently filed Quarterly Reports on Form 10-Q. Moreover, the Company operates in a very competitive and rapidly changing environment. New risks emerge from time to time. It is not possible for management to predict all risks, nor can the Company assess the impact of all factors on its business or the extent to which any factor, or combination of factors, may cause actual results to differ materially from those contained in any forward-looking statements we may make. Given these uncertainties, you should not place undue reliance on these forward-looking statements. The Company qualifies all the forward-looking statements in this Presentation by these cautionary statements. Except as required by law, the Company undertakes no obligation to publicly update any forward-looking statements, whether as a result of new information, future events or otherwise. Statements contained herein are made as of the date of this Presentation unless stated otherwise, and this Presentation shall not under any circumstances create an implication that the information contained herein is correct as of any time after such date or that the information will be updated or revisited to reflect information that subsequently becomes available or changes occurring after that date hereof. Certain information contained in this Presentation relates to or is based on statistical and other industry and market data obtained from independent industry publications and research, surveys and studies conducted by independent third parties as well as the Company’s own estimates of the prevalence of certain diseases and conditions. The market data used in this Presentation involves a number of assumptions and limitations, and you are cautioned not to give undue weight to such data. Industry publications and third-party research, surveys and studies generally indicate that their information has been obtained from sources believed to be reliable, although they do not guarantee the accuracy or completeness of such information. The Company’s estimates of the patient population with the potential to benefit from treatment with any product candidates the Company may develop include several key assumptions based on its industry knowledge, industry publications and third-party research, which may be based on a small sample size and may fail to accurately reflect the addressable patient population. While the Company believes that its internal assumptions are reasonable, no independent source has verified such assumptions. This Presentation may contain trademarks, trade names, or service marks belonging to other entities. The Company does not intend the use or display of other parties’ trade names, trademarks or service marks to imply a relationship with, or endorsement or sponsorship of, or by these other parties. None of the Company or any of its directors, officers, employees, contractors, agents, consultants, advisors or other representatives makes any representation or warranty, express or implied, as to the accuracy or completeness of the information contained in this Presentation.

Leadership Team Chris Vasquez Chief Financial Officer Sheri Lydick Chief Commercial Officer Jay Short, Ph.D. Chairman, CEO and Cofounder Eric Sievers, M.D. Chief Medical Officer Ji Hwan Lee VP Head of Portfolio Management Monica Sullivan Sr. VP, Intellectual Property & Contracts Susie Melody Sr. VP, Human Resources CapaIP

Board of Directors and Scientific Advisors Susan Moran, MD, MSCE Director Lawrence Steinman, MD Lead Director Mary Ann Gray, Ph.D. Director Sylvia McBrinn Director Scott Smith Director James Allison, Ph.D. MD Anderson Cancer Center Scientific Advisor Eddie Williams Director Lawrence Fong, MD Cancer Immunotherapy Program, UCSF Scientific Advisor Padmanee Sharma, MD, Ph.D. MD Anderson Cancer Center Scientific Advisor Jay Short, Ph.D. Chairman, Chief Executive Officer & Cofounder Director

BioAtla discovered that acidic pH at the cancer cell surface unveils binding sites that are shielded at normal pH of healthy cells BioAtla invented CAB technology, creating antibodies that bind only to these unveiled sites on cancer cells CAB binding region is not masked or caged and thus different from prodrugs that require irreversible enzymatic cleavage to become activated CAB antibodies have the potential for increased efficacy with improved safety relative to traditional antibodies H+ H+ H+ H+ H+ H+ H+ H+ H+ H+ Alkaline Healthy Cell Membrane Acidic Cancer Cell Membrane CAB Binding No CAB Binding Selective and Targeted CAB Technology Widens Therapeutic Window Thus has the potential to enhance clinical outcomes in multiple tumor types Chang, H.W., Frey, G., Liu, H., Xing, C., Steinman, L, Boyle, B.J., & Short, J.M. (2021) PNAS 118(9): 1-10, Suppl. 1-19.

All cancer cells are acidic (pH5.3-pH6.7) The most acidic regions are oxygenated, not anaerobic Acidity is a result of the need for precursor molecules from glycolysis for continuous cell replication Cancer cells use acidity to promote metastasis and defend against immune response CAB mechanism Leverages naturally occurring, negatively charged molecules (e.g. bicarbonate, hydrogen sulfide) to differentiate between targets on cancer cells versus normal cells These physiological molecules underpin the CAB mechanism and are referred to as Protein-associated Chemical Switches (PaCS)TM In normal tissues, PaCS shield epitopes, so CAB antibodies cannot bind. In contrast, cancer cells produce H+ ions that remove PaCS molecules from the epitopes, enabling cancer-specific binding. CAB Platform Technology Summary

Myths vs Facts of pH Therapies in Cancer BioAtla’s Solution: Conditionally Active Biologics (CAB) Myth Fact Not all tumors are acidic ALL tumors and cancer cells are acidic (i.e. pH5.3-6.7). Cancer cells are ATP limited Cancer cells are NOT generally ATP deficient but are limited in other precursor molecules whose synthesis depends upon glycolysis. pH technology will miss tumor cells because tumor pH is variable While tumor degree of pH acidity can vary, CABs are designed to bind any cancer cell at or below a predetermined pH. Tumor size influences the acidic environment so pH technology will not work Larger tumors with larger anaerobic regions are not necessarily more acidic since oxygenated regions have higher acidity due to the higher concentration of hydrogen ions from rapid glycolysis. Cancer cells – as opposed to the average pH of a tumor – are more acidic, especially at the membrane of the cancer cell.

CAB Antibodies Bind Selectively and Reversibly Based on the Tumor Microenvironment (TME) Note: OD450nm = optical density measurements using a microplate reader with a 450nm filter; TME = Tumor Micro Environment; mABs = monoclonal antibodies; Data above based on non-human primate studies CABs Bind Selectively in the Lower pH TME Selective Binding Focused Tumor Killing Normal Cells Preserved ≥7.4 is pH of normal cell CABs Do Not Bind to Normal Tissues, Enhancing Safety Non-CABs Bind to Normal Tissues, Causing Toxicity Tumor pH Blood ELISA Data Non-CAB CABs Enhancing exposure and reducing toxicity

CAB Technology Eliminates On-target, Off-tumor Binding Thus, widening the therapeutic index by 12.6-fold over cetuximab CAB Anti-EGFR mAb Enhanced Tumor Selectivity (12.6-fold increase in TI) Cetuximab Control PET scan images

Conditional Binding Approaches: Prodrug vs BioAtla's CAB Platform Technology Feature Proteolytic cleavage of prodrug Conditionally Active Biologic (CAB) Mechanism of Action Mask obscuring antibody binding site is cleaved by tumor-associated proteases Biologic is engineered to bind only in TME at low pH* Trigger Type Enzymes overexpressed in tumors TME (i.e., low pH) Bioengineering Addition of foreign sequence No foreign sequences Activation Precision Requires activation; dependent on expression and enzymatic efficiency No activation required – maximizes potency Risk of Off-Target Effects Irreversible activation - can bind target in normal tissues Reversible binding - will not bind target in normal tissues TME = Tumor microenvironment *Warburg effect

Preclinical evidence of CAB selectivity Differential EGFR tumor vs. skin binding (12.6-fold improved TI) AXL-ADC reduced TMDD yielding Increased T1/2 and exposure in NHP (>2-fold increase in T1/2) Reduced liver enzymes (<10% of non-CAB ALT levels) EpCAM DualCAB TCE maintained efficacy with highly reduced toxicity MTD not reached in NHP >100-fold improvement in TI B7H3 Dual CAB TCE; target expression associated with high acidity via hyper-glycolysis MTD not reached in NHP Encouraging safety profile compared to other B7H3 TCEs in development CTLA4 reduced peripheral immune response while maintaining efficacy Maintains efficacy at same dose, while enabling higher and extended dosing Significant reduction in colitis in NHP compared to ipi MTD not reached at 30 mg/kg in NHP Selective reduction of activated T cells in the periphery or normal tissues Preclinical Evidence Summary CAB, Conditionally Active Biologics; TI, Therapeutic Index; T1/2, Half-life; TMDD, Target—Mediated Drug Disposition; NHP, Non-Human Primate; ALT, Alanine Aminotransferase; TCE, T-Cell Engager; MTD, Maximally Tolerated Dose; ipi, ipilimumab

Clinical evidence of CAB selectivity AXL-ADC good risk/benefit ratio Two non-CAB AXL-targeting ADCs terminated in P1 due to toxicity Potent and durable response with differentiated OS in mKRAS NSCLC and sarcoma patients ROR2-ADC good risk/benefit ratio Good tolerability with only 7% treatment-related discontinuation rate Potent and durable response in SCCHN patients EpCAM DualCAB TCE Non-CAB EpCAM TCE (BiTE) terminated in P1 Most advanced EpCAM TCE in the clinic showing tumor-reduction, ongoing in P1 MTD not yet reached CTLA4 I/O enables higher and prolonged dosing with reduced immune-meditated AEs Maintains PK and efficacy at similar dose, while enabling more intensive dosing MTD not reached at 14.3 mg/kg Extended dosing (>2x over ipi) and at higher doses Reduced grade 3 AEs such as colitis even at higher doses Clinical Evidence Summary CABs demonstrate universal clinical improvement in TI and enable therapeutic development "undruggable" targets

Key Advantages of the CAB Platform Widening The Therapeutic Index Conditional and reversible binding increases clinical activity and improves safety Not dependent on enzymatic activation for selective binding Enhances pharmacologic properties Broadly applicable to antibody formats including ADCs, Bispecific TCEs, CAR-Ts and other proteins Provides ability to create new therapies and combinations against targets that had previously been limited due to toxicity

CAB Program Target Indications IND Enabling Pre-Clinical Phase 1 Clinical Phase 2 Clinical Phase 3 Clinical CAB-Bispecific TCE BA3182 EpCAM x CD3 Adenocarcinomas CT-202 (Out licensed to Context Therapeutics) Nectin-4 x CD3 Solid Tumors BA3142 B7H3 x CD3 Solid Tumors BA3241 Trop2 x CD3 Solid Tumors BA3311 EGFR x CD3 Solid Tumors CAB-ADCs BA3021 Ozuriftamab Vedotin ROR2 2L+ OPSCC BA3011 Mecbotamab Vedotin AXL NSCLC Sarcoma BA3361 Nectin-4 Solid Tumors CAB-I/O BA3071 Evalstotug CTLA-4 1L and 2L Melanoma IND, investigational new drug; OPSCC, oropharyngeal squamous cell carcinoma; NSCLC, Non-small Cell Lung Cancer; TCE, T cell engager Clinical and Preclinical Pipeline of Differentiated CAB Assets Across Multiple Solid Tumors

BA3182 (Dual CAB EpCAM x CD3 Bispecific T-Cell Engager): Adenocarcinoma

Why EpCAM (Epithelial Cell Adhesion Molecule) As a Target? Targeting EpCAM has potential to serve over 1 Million patients (potential Pan-cancer drug) Estimated Number of New Cancer Cases in 2025 EpCAM Expression (TIS 1 to 12)2 Breast Cancer 319,750 81% Prostate Cancer 313,780 99% Lung Cancer 226,650 93% NSCLC 80% SCLC Colon Cancer 154,270 100% Pancreatic Cancer 67,440 99% Thyroid Cancer 44,020 97% Ovarian 20,890 92% Gallbladder & other biliary 12,610 97% 1Siegel RL, Kratzer TB, Giaquinto AN, Sung H, Jemal A. Cancer statistics, 2025. CA Cancer J Clin. 2025. 2G. Spizzo, et al. J Clin Pathol 2011;64:415e420. Normal Colon Colon Cancer Challenges of targeting EpCAM All normal epithelia express EpCAM which with traditional antibodies would lead to on-target, off-tumor toxicities CABs are essential for targeting EpCAM EpCAM IHC

Proposed Mechanism of Action of BA3182, a Dual-CAB EpCAM x CD3 bispecific T-cell Engager CAB-TCEs redirect all T cells in the TME to attack cancer cells, but not in healthy tissues Mabs. 2024 Mar 6;16(1):2322562

pH Range ELISA TME Physiologic pH BA3182 – First Dual-CAB T-cell Engager Targeting EpCAM Potent Lysis of EpCAM Positive Cancer Cells by BA3182 HCT116 CDX Model Highly reduced binding to both targets at normal physiological pH Similar efficacy in mouse xenograft models compared to Non-CAB parent molecule Tetravalent T-cell engager EpCAM* Mabs. 2024 Mar 6;16(1):2322562 *(2+1) functional format (only one anti-CD3scFv can engage a T cell based on steric hindrance)

CAB-EpCAM-TCE, BA3182, is Well Tolerated at High Doses in Non-human Primates Test Article Non-CAB Parent BA3182 Dose 0.05 mg/kg 5 mg/kg Clinical Outcome All Euthanized on Day 8 No Clinical Findings MCP-1 (pg/mL) IL-2 (pg/mL) IL-6 (pg/mL) Time (Hours) Non-CAB Parent 0.05 mg/kg BA3182 2.5 mg/kg Low cytokine levels with DualCAB vs Non-CAB Parent even at significantly higher doses Cytokine Levels Associated with Toxicity TI Non-CAB Parent BA3182 Increase in Safety and Tolerability Mabs. 2024 Mar 6;16(1):2322562

Phase 1 Dose Escalation Study of BA3182 in Advanced Adenocarcinomas Trial ongoing; evaluating various dosing and treatment schedules 0.003 to 0.1 mg 0.01/0.3 mg 0.01/0.6 mg 0.1, 0.3 or 0.6/1.2 mg 0.1/0.6/1.8 mg Continue Escalation No Priming Dose One Priming Dose Two Priming Doses Dose B (N~ 20-30) Dose A (N~ 20-30) R Part A: Dose Escalation (QW) Part B: Dose Expansion/Dose Optimization Treatment notes: Prophylactic acetaminophen and diphenhydramine delivered prior to all doses and prophylactic tocilizumab (no corticosteroids) given prior to Cycle 1 Day 1 treatment dose Post treatment ondansetron guided for nausea Ongoing weekly treatment dosing continued after DLT observation interval concluded Presented at ESMO Berlin, 2025.

BA3182 Phase 1 Patient Demographics Patients dosed SC per protocol as of September 10, 2025: N=35 Data cut as of 10Sep25 Presented at ESMO Berlin, 2025. Patient Characteristic N=35 Age, mean (SD), y 57 (10) Male (n, %) 19 (54) Female (n, %) 16 (46) ECOG performance status 0 (n, %) 24 (69) 1 (n, %) 11 (31) Presence of liver metastases (n, %) 22 (63) Carcinoma Type (n, %) N=35 Median # of prior treatments Adenoid Cystic Carcinoma 1 (3) 2 Appendiceal 1 (3) 2 Cholangiocarcinoma 2 (6) 1 Colorectal 22 (63) 4 Gallbladder 1 (3) 3 Ovarian 1 (3) 8 Pancreas 7 (20) 3

BA3182 Preliminary Prolonged Tumor Control With Increasing Doses Live Data cut as of 10Sep2025 *Pts with unconfirmed PD were treated beyond initial progression per protocol when pseudo-progression was suspected Presented at ESMO Berlin, 2025. Treatment Dose (mg) Average # of Doses All 8+ 1.2 3+ 0.6 11+ 0.3 9 0.1 8 0.032 6

BA3182 Achieved Objective Tumor Size Reductions Across Multiple Tumor Types Preliminary assessment of anti-tumor activity among pts receiving treatment doses of ≥0.6 mg (N=24) 14 CRC pts: among 7 pts who had one scan – 5 achieved SD at 1st scan; among the 7 pts without scans –2 withdrew consent before the 1st scan and 5 are pending 1st scan 6 pancreatic cancer pts: 2 pts had SD at 1st scan and 2 pts continued treatment beyond potential pseudo-progression 2 cholangiocarcinoma pts: 1 pt achieved a confirmed PR and 1 pt is pending first scan 1 ACC pt experienced SD Among pts treated with ≥0.6 mg , 9/10 pts achieved SD at a higher rate and remained on active treatment for prolonged intervals, generally longer than those who received lower doses Presented at ESMO Berlin, 2025. Live Data cut as of 10Sep2025 CRC, colorectal cancer; SD, stable disease; PR, partial response; ACC, adenoid cystic carcinoma

BA3182 Adverse Events Generally Transient and Readily Manageable Preliminary safety of subcutaneous dosing Data cut as of 10Sep25 Presented at ESMO Berlin, 2025. 1Early, transient elevation of hepatic analytes: AST, ALT, bilirubin, and/or alkaline phosphatase 2Non-febrile, transient neutropenia, possibly related to tocilizumab 3G2 pancreatitis, G2 atrial fibrillation, G3 diarrhea, G3 acute kidney injury, and G2 CRS 4G3 diarrhea; G3 diarrhea/lymphocyte count decrease; G3 acute kidney injury/white blood cell count decrease Characteristic N=35 (n, %) Any Adverse Events (AEs) 29 (83) Related AEs of Grade 3-4 14 (40) Related AEs of Grade 3 hepatic analytes1 11 (31) Related AEs of Grade 4 hepatic analytes1 1 (3) Related AEs of Grade 3-4 non-febrile neutropenia2 2 (6) Related AEs of Grade 3 excluding hepatic analytes1 and non-febrile neutropenia2,4 3 (9) Any related serious AEs3 5 (14) Related CRS of any grade (per ASTCT grading) 2 (6) Related AEs leading to death 0 Related AEs leading to treatment discontinuation 1 (3) TRAE >10% (N=35) TRAE Any (n, %) TRAE G3+ (n, %) Alanine aminotransferase increased* 15 (42.9) 7 (20.0) Aspartate aminotransferase increased* 15 (42.9) 9 (25.7) Nausea 14 (40.0) 0 Injection site reaction 12 (34.3) 0 Diarrhea 11 (31.4) 2 (5.7) Fatigue 9 (25.7) 0 Blood bilirubin increased* 8 (22.9) 0 Neutrophil count decreased 7 (20.0) 2 (5.7) Blood alkaline phosphatase increased* 6 (17.1) 1 (2.9) Decreased appetite 6 (17.1) 0 Dysgeusia 5 (14.3) 0 Vomiting 5 (14.3) 0 Abdominal pain 4 (11.4) 0 Constipation 4 (11.4) 0 *Transient laboratory changes; resolved

Early, Transient Increases in Hepatic Analytes Resolved with Continued Dosing Summary of hepatic analyte changes: Early, transient, and asymptomatic Enabling on-time weekly treatment dosing Consistent with cholestasis (not on target toxicity) Patients primed with 0.1 mg prior to treatment dose (N=9; treatment doses: 0.3, 0.6, and 1.2 mg) Grade 3 Grade 2 Grade 1 Normal Range 0.1 / 1.2 mg (n=4) 0.1 / 0.6 mg (n=3) 0.1 / 0.3 mg (n=2) Live Data cut as of 10Sep2025 Presented at ESMO Berlin, 2025.

71-year-old male with stage IV cholangiocarcinoma previously treated on clinical trial with gemcitabine, cisplatin, durvalumab, and investigational agent. Confirmed Partial Response (31% Tumor Reduction) BA3182 at 0.6 mg in Patient with Intrahepatic Cholangiocarcinoma without Progression for >6 months Presented at ESMO Berlin, 2025. Data cut as of 10Sep25

BA3182 Summary / Upcoming Milestones BA3182, a dual-conditionally binding CAB-EpCAM x CAB-CD3 T-cell engager continues to demonstrate a manageable safety profile with preliminary evidence of antitumor activity Early cytokine increases appear to cause brief, reversible cholestasis – consistent with tumor-selective targeting BA3182 treatment achieved an ongoing, confirmed partial response and multiple patients have experienced prolonged tumor control; currently testing dose levels up to 1.8mg and dose varying frequencies Dose escalation read out anticipated 1H’26

Ozuriftamab Vedotin (CAB-ROR2-ADC): Oropharyngeal Squamous Cell Carcinoma (OPSCC)

Significant Opportunity for Oz-V in OPSCC Rapidly Growing Patient Population OPSCC incidence is increasing, largely due to HPV-driven disease1,2 Up to 80% of OPSCC cases in the United States caused by HPV infection3 By 2030, OPSCC will be the most common subtype of head and neck cancer in the US1,2 Potential to Address Significant Unmet Need OPSCC poorly served by SOC, including EGFR inhibitors4-10 Oz-V has a compelling and differentiated profile in HPV+ OPSCC OPSCC, Oropharyngeal Squamous Cell Carcinoma; SOC, Standard of Care; WW, worldwide 1Med Sci (Basel). 2023 Jun 13;11(2):42.; 2Oral Oncol. 2021 Apr:115:105177.; 3Dela J Public Health. 2023 Apr 22;9(1):26-28. 4N Engl J Med 2016;375:1856-1867. 5Journal of Clinical Oncology 2018; 36(15): 1551-1558. 6Cohen E, et al. (2019) Lancet 393, 156–167. 7British Journal of Cancer (2018) 119:153–159; https://doi.org/10.1038/s41416-018-0131-9. 82008 Jun 15;112(12):2710-9. doi: 10.1002/cncr.23442, 9Erbitux USPI accessed 2024., 10INTERLINK-1: Phase 3 study of cetuximab ± monalizumab /Volume 34, Supplement 2S554-S555 October 2023., 11Internal BioAtla projections.; 12https://www.coherentmi.com/industry-reports/oropharyngeal-cancer-market; 13Cervical Cancer Therapeutics Market Size, Demand, Report to 2033; 23-cervix-uteri-fact-sheet.pdf Large Commercial Opportunity with Potential to Expand $800mm WW peak sales projections in 2L+ OPSCC11 Total WW OPSCC market value projected to be ~$3Bn by 203212 Total HPV+ solid tumors (ie, cervical) worldwide market valued at >$7Bn13

OPSCC Incidence is Increasing Proportion of new HPV+ OPSCC diagnoses now approach 80% in the US Years Annual number of OPC Cases Oropharyngeal Cancer Burden by Infection Status Damgacioglu H, Sonawane K, Chhatwal J, et al. Long-term impact of HPV vaccination and COVID-19 pandemic on oropharyngeal cancer incidence and burden among men in the USA: A modeling Study. The Lancet Regional Health – Americas. 2022;8:100143. *CDC report. Oropharyngeal cancer has now surpassed cervical cancer as the most common HPV-related malignancy More than 21,000 U.S. cases yearly compared to nearly 12,000 cervical cancer cases*

Ozuriftamab Vedotin (Oz-V): CAB-ROR2-ADC ROR2 is expressed in a variety of tumor types, with overexpression associated with metastasis, tumor resistance to chemotherapy, and poor prognosis CAB-tumor cell target Cytotoxic payload and linker BA3021 pH binding inflection point adjusted for tumor microenvironment selectivity 7.4 is pH of normal cell Blood MMAE-containing ADC (DAR4) with cleavable linker Humanized anti-ROR2 (N-terminal) IgG1 ~2nM affinity (pH 6) MMAE-containing ADC (DAR4) with cleavable linker Epitope in Ig loop region

ROR2 Overexpression is Driven by HPV E6 / E7 Oncoproteins 80% of OPSCC cases in the United States caused by HPV infection HPV driven cancers Highest ROR2 expression among SCCHN HPV E6 and E7 oncoproteins drive ROR2 overexpression ROR2 overexpression results in increased proliferation and invasiveness Oz-V conditionally and selectively eliminates ROR2-expressing cells HPV infection drives ROR2 overexpression ROR2 expression Carrero, et al. Oncogene (2019) 38:3551–3568; https://doi.org/10.1038/s41388-018-0659-4 Avincsal, et al. Oncol Rep. 2021 Jul;46(1):148. doi: 10.3892/or.2021.8099. Epub 2021 Jun 3. PMID: 34080643. Figure adapted from Z. Lu, et al. Cancers. 2024, 16, 3474.

Demographics and Baseline Clinical Characteristics HPV p16 positive and all patients Ozuriftamab vedotin1.8 mg/kg OPSCC p16+a Full Analysis 2Q3W (n=10) Q2W (n=13) 2Q3W and Q2W (N=40) Age, mean (SD), y 65 (5) 62 (7) 65 (8) Sex, n (%) Male 10 (100) 12 (92) 36 (90) Female 0 1 (8) 4 (10) ECOG performance, n (%) 0 3 (30) 6 (46) 15 (38) 1 7 (70) 7 (54) 25 (63) Number of prior lines of therapy, median 3 3 3 Prior anti–PD-1 exposure, n (%) 10 (100) 13 (100) 40 (100) Prior platinum-based chemotherapy exposure, n (%) 10 (100) 11 (85) 34 (85) Prior taxane exposure, n(%) 7 (70) 7 (54) 26 (65) aHPV status was determined using p16 immunohistochemistry. 2Q3W = Days 1 and 8 of 21-day cycle Data Cut Date: 31Oct2025

Oz-V in SCCHN Continues to Demonstrate Clinical Responses in a Heavily Pretreated Population 1.8 mg/kg Q2W and 2Q3W; Median of 3 prior lines of therapy CR CR Response evaluable patients defined as pts that had at least 1 scan after treatment with study drug Not evaluable prior to first scan: 3 pts had clinical progression 2 pts withdrew consent Data Cut Date: 31Oct2025

Oz-V in SCCHN p16+^ OPSCC 1.8 mg/kg Q2W; Median of 3 prior lines of therapy * Response evaluable patients defined as patients that had at least 1 scan after treatment with study drug Not evaluable prior to first scan: 1 patients had clinical progression Q2W Responders (confirmed & unconfirmed) 42% (5/12*) Responders (confirmed) 25% (3/12*) DCR 92% (11/12*) Median DOR (months) 9.9 ongoing Median PFS (months) 4.7 ongoing Median OS (months) 11.6 ongoing Data Cut Date: 31Oct2025 ^ p16 is strongly associated with HPV; HPV testing in progress for unknown patients Previous lines of therapy 3 5 1 2 3 4 3 3 5 2 3 2

2L+ HPV+ OPSCC: Cross Trial Comparisons of ORR and OS Considerably improved response rate and survival among a heavily pretreated trial population Cross trial comparisons** Median prior lines of therapy ORR (%) OS (months)*** Ozuriftamab vedotin monotherapy (1.8 mg/kg Q2W) Study Ongoing* 3 42% ORR 25% cORR 11.6 ongoing SOC (methotrexate, docetaxel, or cetuximab)1,2 2 3.4% 4.4 Cetuximab monotherapy3,4,5 1 0% NA Not approved Petosemtamab (1500 mg Q2W) 2 13% NA 1N Engl J Med 2016;375:1856-1867. 2 Journal of Clinical Oncology 2018; 36(15): 1551-1558. 32008 Jun 15;112(12):2710-9. doi: 10.1002/cncr.23442, 4Erbitux USPI accessed 2024., 5INTERLINK-1: Phase 3 study of cetuximab ± monalizumab /Volume 34, Supplement 2S554-S555 October 2023 SOC, Standard of Care (Cetuximab, Methotrexate or Docetaxel); NA, not available **The comparisons above are not based on data resulting from a head-to-head trial and are not direct comparisons. Different protocol designs, trial designs, patient selection and populations, number of patients, trial endpoints, trial objectives and other parameters that are not the same between the relevant trials may lead to bias in the results causing comparisons from different trials to be unreliable. *Response evaluable patients defined as patients that had at least 1 scan after treatment with study drug; Not evaluable prior to first scan: 1 patients had clinical progression Anti-EGFR therapies have inferior therapeutic outcomes among HPV+ patients Accelerated Approval Opportunity (ORR) ***From Dr. Alan Ho at MSKCC: “
extrapolating the OS of CheckMate 141 patients after progression on Nivo, we are figuring the OS of these patients to be about 7 months (OS of 9 -PFS of 2 months) post PD-1. Some of the retrospective papers (https://pubmed.ncbi.nlm.nih.gov/31574417/ & https://pubmed.ncbi.nlm.nih.gov/31864957/ we reviewed post-PD1 saw that upper limit of OS was of about 7-8.5 months post PD-1. Figuring what the upper and lower estimates are, we think 6 months OS is a fair estimate for the null.” Data Cut Date: 31Oct2025

Confirmed Complete Response Oz-V in SCCHN (1.8 mg/kg Q2W) – HPV Positive 76-year-old male, stage IV – recurred after surgery and radiation therapy; prior treatments: pembrolizumab; clinical trial bispecific anti-PD1/CD47; patient remains in complete response >16 months after Oz-V treatment initiation Baseline - July 14, 2023 On Treatment – December 8, 2023

Partial Response (-80%) Oz-V in SCCHN (1.8 mg/kg Q2W) – HPV Positive 63-year-old male, stage IV – recurred after surgery and radiation therapy; prior treatments: platinum, investigational agents including pembrolizumab Baseline – December 4, 2025 On Treatment – February 15, 2025

Most frequent AEs of any grade (>15% of patients) (N=40) 2Q3W (n=20) Treatment-Related Adverse Events Related Grade 3 6 (30%) Related Grade 4 2 (10%) Related Serious 2 (10%) Treatment-Related Adverse Events Related Grade 3 3 (15%) Related Grade 4 0 (0%) Related Serious 0 (0%) Q2W (n=20) Related AEs Related AEs Data Cut Date: 31Oct2025

Ph2 Oz-V: Overall Safety Summary of SCCHN patients Investigator Choice3 related AEs G3 or 4 = 35% - Considerably higher than Oz-V 1.8 mg/kg Q2W 1.8 mg/kg Q2W (N=20) 1.8 mg/kg 2Q3W (N=20)4 Total (N=40)4 Any Adverse Events (AEs) 19 (95%) 20 (100%) 38 (95%) Related AEs with CTCAE1 Grade 3 or 42 3 (15%) 7 (35%) 10 (25%) Any related serious AEs2 0 2 (10%) 2 (5%) Possibly Related AEs leading to death2 0 0 0 Related AEs leading to treatment discontinuation2 1 (5%) 1 (5%) 2 (5%) 1CTCAE: Common Terminology Criteria for Adverse Events. The NCI Common Terminology Criteria for Adverse Events is a descriptive terminology which is utilized for Adverse Event (AE) reporting. A grading (severity) scale is provided for each AE term. 2As assessed by the investigator. Missing responses are counted as related. 3N Engl J Med 2016;375:1856-1867. 2 Journal of Clinical Oncology 2018; 36(15): 1551-1558 4One patient from Phase 1 not included Data Cut Date: 31Oct2025

Phase 2 Oz-V Safety Data Most frequent treatment-emergent related adverse events (>15%) 1.8 mg/kg Q2W (N=20) 1.8 mg/kg 2Q3W (N=20)^ Total (N=40)^ Preferred Term All Grades n (%) Grade 3-4 n (%) All Grades n (%) Grade 3-4 n (%) All Grades n (%) Grade 3-4 n (%) Number of subjects with at least one TRAE 19 (95) 12 (60) 20 (100) 16 (80) 39 (98) 28 (70) Fatigue 11 (55) 0 12 (60) 1 (5) 23 (58) 1 (3) Nausea 5 (25) 1 (5) 7 (35) 0 12 (30) 1 (3) Anaemia 8 (40) 2 (10) 3 (15) 1 (5) 11 (28) 3 (8) Constipation 5 (25) 0 6 (30) 0 11 (28) 0 Lymphocyte count decreased 4 (20) 2 (10) 7 (35) 5 (25) 11 (28) 7 (18) Decreased appetite 3 (15) 0 7 (35) 0 10 (25) 1 (3) Peripheral neuropathy¥ 7 (35) 1 (5) 3 (15) 1 (5) 10 (25) 2 (5) Diarrhoea 1 (5) 0 8 (40) 2 (10) 9 (23) 2 (5) Hyponatraemia 3 (15) 1 (5) 6 (30) 4 (20) 9 (23) 5 (13) Arthralgia 5 (25) 0 3 (15) 0 8 (20) 0 Weight decreased 3 (15) 0 5 (25) 1 (5) 8 (20) 1 (3) Cough 4 (20) 0 3 (15) 0 7 (18) 0 Blood lactate dehydrogenase increased 4 (20) 0 3 (15) 0 7 (18) 0 Hypercalcaemia 3 (15) 1 (5) 3 (15) 1 (5) 6 (15) 2 (5) Neutropenia* 1 (5) 0 5 (25) 2 (10) 6 (15) 2 (5) White blood cell count decreased 4 (20) 0 2 (10) 0 6 (15) 0 ^ One patient from Phase 1 not included * Derived from neutropenia, and neutrophil count decreased ¥ Derived from neuropathy peripheral, peripheral motor neuropathy, and peripheral sensory neuropathy Data Cut Date: 31Oct2025

FDA End of Phase 2 Meeting: Key Outcomes Dual primary endpoints of ORR and OS to support potential accelerated approval followed by full approval Pivotal Trial Design: For full approval, approximately 300 patients prospectively randomized and stratified, one to one between two open label treatment arms Oz-V Dose and Regimen: Patients randomized to the investigational arm will receive 1.8 mg/kg every other week Investigator’s Choice (IC) control arm: Patients randomized to the control arm will receive either cetuximab, docetaxel, or methotrexate monotherapy Accelerated Approval Endpoint: Based on interim analysis of enrolled patients, statistically significant improvement of confirmed Overall Response Rate (ORR) by Blinded Independent Central Review (BICR) supported by an adequately characterized Duration of Response (DOR) without detriment in OS Full Approval Endpoint: Statistically significant improvement of OS

Mecbotamab Vedotin (CAB-AXL-ADC): mKRAS Non-Small Cell Lung Cancer (NSCLC)

Mecbotamab Vedotin (Mec-V): CAB-AXL-ADC AXL is expressed in a variety of tumor types, with overexpression associated with metastasis, tumor resistance to chemotherapy, and poor prognosis 1. Gay CM, Balaji K, Byers LA. Br J Cancer. 2017;116(4):415-423. 2. Zhang G, Wang M, Zhao H, Cui W. Oncol Lett. 2018;15(3):2726-2734. Abbreviations: ADC, antibody-drug conjugate; AM, affinity matched; CAB, conditionally active biologic; DAR, drug antibody ratio; ELISA, enzyme linked immunosorbent assay; Ig, immunoglobulin; OD, optical density; VC-MMAE, valine-citrulline monomethylauristatin E. CAB-tumor cell target Cytotoxic payload and linker BA3011 pH binding inflection point adjusted for tumor microenvironment selectivity ≥7.4 is pH of normal cell pH value Blood Non-CAB AM CAB AM 6.00 6.25 6.50 6.75 7.00 7.25 7.50 0 25 50 75 100 Non-CAB BA3011 Normalized ELISA Data (OD 450 nm) Humanized anti-AXL IgG1 ~100 pM affinity (pH 6.5) VC-MMAE (DAR 4) linker and payload Epitope in Ig loop region

AXL Plays a Crucial Role in the Survival of mutated KRAS (mKRAS) NSCLC Cells Adapted from Morimoto et al. Cancer Letters 587 (April 2024) mKRAS leads to upregulation and activation of AXL expression Tumor resistance mechanism mKRAS represents 30% of all NSCLC patients 70% to 85% of mKRAS NSCLC express AXL by IHC and higher by mRNA analysis AXL over-expression drives aggressive tumor characteristics, resistance to therapies, and poor patient outcomes Significant opportunity for Mec-V (CAB-AXL-ADC) * Morimoto et al., Cancer Letters 587 (2024) 216692 and BioAtla study BA3011-002 IHC- Immunohistochemical assay

Mec-V 1.8 mg/kg Q2W Overall Survival mKRAS vs wtKRAS NSCLC Median of 3 prior lines of tx Events n (%) Median (95% CI) mKRAS 7 (41) NR (4.0 – NE) wtKRAS 20 (74) 8.3 (3.5-10.2) ^Response evaluable patients defined as patients that had at least 1 scan after treatment with study drug Prior to first scan: one patients withdrew consent median of 2 prior lines of tx mKRAS (Q2W only) Responders (confirmed & unconfirmed) 31% (5/16^) Responders (confirmed) 19% (3/16^) DCR 81% (13/16^) PFS 4.6 One-year landmark OS 67% Data Cut Date: 19Aug2025

Mec-V 1.8 mg/kg Q2W Associated with Exceptional Overall Survival in mKRAS NSCLC Overall survival cross trial comparison1 mecbotamab vedotin (1.8 mg/kg Q2W) Docetaxel3 Population mKRAS mKRAS Number of patients 17 256 Prior Lines of tx 2 1 Prior Taxane 53% 0% ORR 31%2 14% PFS (months) 4.5 2.8 OS (months) Not reached 7.9 Survival at 12 months 67% <40% Survival at 24 months 59% <20% . mecbotamab vedotin (KRAS) 12 months 24 months Docetaxel + Placebo (SOC in KRAS) Selumetinib + Docetaxel (mKRAS) FDA guidance: 2L+ NSCLC Phase 3 trial will be randomized mecbotamab vedotin versus docetaxel (full-approval = OS) 1 The comparisons above are not based on data resulting from a head-to-head trial and are not direct comparisons. Different protocol designs, trial designs, patient selection and populations, number of patients, trial endpoints, trial objectives and other parameters that are not the same between the relevant trials may lead to bias in the results causing comparisons from different trials to be unreliable. 2 Confirmed and unconfirmed responders. 3 JAMA. 2017 May 9;317(18):1844–1853 Data Cut Date: 19Aug2025

Competitive Response Rate in 2L+ mKRAS NSCLC Median Overall Survival ranges from 6 to 11 months in mKRAS NSCLC when treated with docetaxel in the 2L+ Cross trial comparisons** Pts Median prior lines of therapy ORR Median OS (months) Mecbotamab Vedotin (1.8 mg/kg Q2W regimen) Study Ongoing* 17 3 31%*** Not Reached (1 year landmark at 67%) SELECT-11 (all mKRAS variants) 256 1 13.7% 7.9 Real-life ESME cohort2 (all mKRAS variants) 1000+ 1 NA 6.1 to 10.6* Codebreak 2003 (mKRAS G12C) 174 2 13.2% 11.3 KRYSTAL-124 (mKRAS G12C) 152 2 9.2% NA Janne, P et al. JAMA. 2017 May 9;317(18):1844–1853. (2) Thomas QD, et al. ESMO Open, Volume 9, Issue 6, 103473. (3) de Langen AJ, et al. Lancet 2023; 401:733-746; (4) Mok TS, Journal Clinical Oncology 2024; 42(17_suppl):LBA8509. * PD-1/PD-L1 monotherapy, PT-based CT without a PD-1/PD-L1, or Docetaxel monotherapy or combination **The comparisons above are not based on data resulting from a head-to-head trial and are not direct comparisons. Different protocol designs, trial designs, patient selection and populations, number of patients, trial endpoints, trial objectives and other parameters that are not the same between the relevant trials may lead to bias in the results causing comparisons from different trials to be unreliable. ***Confirmed and Unconfirmed Docetaxel Studies

Ph2 Mec-V: Overall Safety Summary of NSCLC patients 1.8 mg/kg Q2W with or without nivolumab generally well-tolerated; only 7% discontinuation due to related AEs Data Cut Date: 19Aug2025 N=45 (%) Any Adverse Events (AEs) 45 (100) Related AEs with CTCAE1 Grade 3 or 42 14 (31) Any Related Serious AEs2 6 (13) Possibly Related AEs leading to death2 0 Related AEs leading to treatment discontinuation2 3 (7) 1CTCAE: Common Terminology Criteria for Adverse Events. The NCI Common Terminology Criteria for Adverse Events is a descriptive terminology which is utilized for Adverse Event (AE) reporting. A grading (severity) scale is provided for each AE term. 2As assessed by the investigator. Missing responses are counted as related. Most frequent AEs any grade occurring at a rate >15%

Potential for Mec-V to Address All mKRAS NSCLC 1.8 mg/kg Q2W associated with improved overall survival and favorable benefit / risk profile Promising anti-tumor activity among patients whose tumors express KRAS mutations mKRAS represents 30% of all NSCLC patients and is associated with increased AXL expression 1.8 mg/kg Q2W associated with exceptional overall survival even in heavily pretreated patients 67% alive at a landmark of one year 59% alive at a landmark of two years; standard of care agents result in less than 20% alive at the two year* Anti-tumor activity across nine different KRAS mutation variants Partial response observed in a patient who had experienced prior failure of sotorasib Patient treated with mecbotamab vedotin + anti-PD-1 antibody remains in complete response for >2 years Potential for a pan mKRAS strategy in NSCLC; currently positioning for a future pivotal trial *The comparisons above are not based on data resulting from a head-to-head trial and are not direct comparisons. Different protocol designs, trial designs, patient selection and populations, number of patients, trial endpoints, trial objectives and other parameters that are not the same between the relevant trials may lead to bias in the results causing comparisons from different trials to be unreliable.

Significant Opportunity for Mec-V to Expand Beyond mKRAS NSCLC KRAS mutations most commonly found in CRC, NSCLC and PDAC Estimated US incidence of select mKRAS cancers and distribution of selected mKRAS variants Lee J., Sivakumar S., Schrock A., et al. NPJ Precision Oncology, 2022. PMID: 36494601. CRC: colorectal cancer; NSCLC: non-small cell lung cancer; PDAC: pancreatic ductal adenocarcinoma 75K patients/year 50K patients/year 50K patients/year

Improved Overall Survival also Observed in Soft Tissue Sarcoma Mirrors Overall Survival Observed in mKRAS NSCLC LMS, UPS, and Liposarcoma: 1.8 mg/kg Q2W Mono and anti-PD-1 Combo; cross-trial comparison* Data Cut Date: 24Mar2025 placebo eribulin dacarbazine trabectedin mecbotamab vedotin 21.5 (14.2 – 29.9) trabectedin 13.6 (7.1 – 17.3) pazopanib 12.5 (10.6 – 12.8) eribulin 13.5 (10.9 – 15.6) dacarbazine 11.5 (9.6 – 13.0) best supportive care 10.8 (6.0 – 18.8) anti-PD1 12.5 (7.7-17.3) placebo 10.7 (8.7 – 12.8) mecbotamab vedotin Best supportive care pazopanib Graaf, Winette van der et al. The Lancet 379 (2012): 1879-1886. Le Cesne, A. et al Annals of Oncology, Volume 32, Issue 8, 1034 - 1044 Schöffski, Patrick et al. The Lancet, Volume 387, Issue 10028, 1629 – 1637 Miao, R. Cancer Immunol Immunother 72, 2521–2527 (2023) Responders, n (%) 2 (5) DCR, n (%) 23 (52) Mec-V in Sarcoma: Median of 2 prior lines of Tx Median OS of 21.5 months Landmark OS at 1 year 73% (54-85) Anti-PD1 monotherapy *The comparisons above are not based on data resulting from a head-to-head trial and are not direct comparisons. Different protocol designs, trial designs, patient selection and populations, number of patients, trial endpoints, trial objectives and other parameters that are not the same between the relevant trials may lead to bias in the results causing comparisons from different trials unreliable.

Evalstotug (CAB-CTLA-4)

Evalustotug is a Next Generation Adaptation of Ipilimumab CAB-CTLA4 selectively active in tumor microenvironment, thereby reducing immune mediated adverse events (imAEs) Ipilimumab Evalstotug CAB CDR Engineering Inability to distinguish between normal and tumor tissue Similar efficacy and PK enhanced with selective pH-dependent tumor binding

Evalstotug is a “CABified” Ipilimumab: A Next Generation CTLA-4 Inhibitor Preserved efficacy with reduced toxicity in combination with PD-1 Ipilimumab (ipi) CDRs modified to bind at tumor cell acidic pH, but not at normal pH leading to evalstotug: Preserved affinity and epitope Equivalent EMax and EC50 in preclinical models However, observed substantially reduced G.I. toxicity in primates CAB CTLA-4, evalstotug, enables targeted exposure in TME enabling lower imAE relative to ipi Abbreviations: ADCC, antibody-dependent cellular cytotoxicity; CAB, Conditionally Active Biologic; CD, cluster of differentiation; CDR, complementarity-determining region; CTLA-4, cytotoxic T-lymphocyte associated protein 4; EC50, concentration producing 50% Emax; Emax, maximum effect; imAE, immune mediated adverse event; PD-1, programmed cell death protein 1; t1/2, half-life; Treg, regulatory T cells. 1. Chang HW, et al. Proc Natl Acad Sci USA. 2021;118(9):e2020606118.

Reversible Binding in the TME Comparison of evalstotug binding to CTLA-4 in different pH conditions Abbreviations: CTLA-4, cytotoxic T-lymphocyte associated protein 4; ICI, immune checkpoint inhibitor; IgG, immunoglobulin G; OD450, optical density at 450 nm; TME, tumor microenvironment. Note: Figures modified from Chang HW, et al. Proc Natl Acad Sci USA. 2021;118(9):e2020606118. CAB-CTLA4 Shows Potent Anti-Tumor Activity in MC38/hCTLA4 KI Model BA3071 induces complete tumor regression in mouse tumor model. Mice were dosed with IgG control and anti-CTLA antibodies at 10mg/kg (equivalent to 1mg/kg anti-CTLA-4 human dose), IP, BIW, N=12 mice/group. pH range ELISA

Study Date 10 20 30 0 0 5 10 15 20 KI67+ CD4+ T cells Ratio to Baseline Vehicle Ipilimumab + nivolumab Evalstotug + nivolumab Evalstotug Reduced GI Toxicity in Primates Abbreviations: CD, cluster of differentiation; Cyno, cynomolgus macaque; GI, gastrointestinal; QW, once weekly. Note: Ipilimumab and evalstotug had the same half-life and exposure in this model. Figure modified from Chang HW, et al. Proc Natl Acad Sci USA. 2021;118(9):e2020606118. aIpilimumab analog or evalstotug 15 mg/kg (≈11 mg/kg human dose) + nivolumab 20 mg/kg (≈14.6 mg/kg human dose) both administered QW for 4 weeks. Liquid feces Non-formed feces Other GI symptoms GI Symptoms Vehicle Control Ipilimumaba + nivolumab 15 mg/kg + 20 mg/kg Evalstotuga + nivolumab 15 mg/kg + 20 mg/kg

Evalstotug Across Multiple Tumor Types

Evalstotug at 350mg Q3W (5mg/kg based on 70 kg pt) in Combination With PD1: Demographics – Across Multiple Tumor Types; n=17 82% patients experienced failure of prior PD-1 treatment Total (N=17) Total (N=17) Prior # of treatments Age, y, mean (SD) 60 (14) Tumor type, n (%) Sex, n (%) Melanoma 11 0 - 1 Female 8 (47) Gastric 2 3 - 5 Male 9 (53) Renal cell 1 3 White race, n (%) 14 (82) Cervical 1 3 ECOG, n (%) NSCLC 1 3 0 12 (71) aHCC 1 5 1 5 (29) Prior Anti-PD-1 Therapy, n (%) 14 (82) Data Cut Date: 31Oct25

Patients Treated with Evalstotug Received More Than Twice as Many Doses Compared with Reported Ipi Dosing Mean number of evalstotug 350 mg doses vs ipilimumab when in combination with PD1 Note: Mean number of doses for ipilimumab is based on a retrospective observational study. Evalstotug 350 mg and 700 mg are dose-equivalent to ipilimumab 5 mg/kg and 10 mg/kg, respectively. *Mohr P, et al. J Eur Acad Dermatol Venereol. 2018;32(6):962-971. Evalstotug 350 mg (5 mg/kg) N=17 Ipilimumab 3 mg/kg* Number of doses 3.4 7.8 0 2 4 6 8 Data Cut Date: 31Oct25

Evalstotug at 350mg Q3W (5mg/kg based on 70 kg pt) in Combination with PD1 Across Multiple Tumor Types 3 Complete Responders with 2 in melanoma and 1 in cervical cancer Total Responders (confirmed & unconfirmed) 44% (7/171) Responders (confirmed) 41% (7/171) DCR 76% (13/171) OS ongoing 1Response evaluable patients defined as patients that had at least 1 scan after treatment with study drug Data Cut Date: 31Oct25

Evalstotug in Combination with PD1: 7 of 17 Achieved Response Durable antitumor activity across multiple solid tumor types 70 mg Q3W + PD1 210 mg Q3W + PD1 350 mg Q3W + PD1 1 year Data Cut Date: 31Oct25

Evalstotug Melanoma (with or without prior ICI treatment)

Evalstotug In Combination with PD1 in Melanoma (with or without prior ICI treatment; n=16): Demographics Evalstotug 350 mg or 700 mg Q3W (5 mg/kg or 10 mg/kg based on 70 kg patient) Total (N=16) Age, y, mean (SD) 61 (14) Sex, n (%) Female 10 (62) Male 6 (38) White race, n (%) 16 (100) ECOG, n (%) 0 12 (75) 1 4 (25) Prior ICI Therapy Status, n (%) No prior ICI treatment 3 (19) Adjuvant 10 (62) Metastatic setting 3 (19) Data Cut Date: 31Oct25 ICI = PD1 or PD1 + LAG3

Evalstotug in Combination with PD1 in Melanoma (with or without prior ICI treatment) 8 responders among 16 patients; 81% of patients had prior ICI Total Responders (confirmed & unconfirmed) 50% (8/161) Responders (confirmed) 50% (8/161) DCR 94% (15/161) PFS Not Reached OS Not Reached 1Response evaluable patients defined as patients that had at least 1 scan after treatment with study drug Data Cut Date: 31Oct25 Prior ICI Y Y Y N Y Y Y N Y Y Y Y N Y Y Y

50% ORR and 94% DCR in Melanoma (with or without prior ICI treatment) 81% patients had received prior PD1 Data Cut Date: 31Oct25 Evalstotug

Evalstotug 1L Melanoma – prior adjuvant therapy or no prior ICI treatment

Total (N=16) Age, y, mean (SD) 61 (14) Sex, n (%) Female 10 (62) Male 6 (38) White race, n (%) 16 (100) ECOG, n (%) 0 12 (75) 1 4 (25) Prior ICI Therapy Status, n (%) No prior ICI treatment 3 (19) Adjuvant 10 (62) Metastatic setting 3 (19) Evalstotug In Combination with PD1 in Melanoma (no prior treatment or prior adjuvant therapy; n=13): Demographics and Waterfall Evalstotug 350 mg or 700 mg Q3W (5 mg/kg or 10 mg/kg based on 70 kg patient) Data Cut Date: 31Oct25 ICI = PD1 or PD1 + LAG3 1L Melanoma Patients 62% ORR 92% DCR

62% ORR and 92% DCR in Melanoma (with or without prior ICI treatment; n=13) Evalstotug in Combination with PD1 81% patients had received prior PD1 adjuvant treatment Data Cut Date: 31Oct25 All 13 melanoma pts living as of data cut date Evalstotug Evalstotug

Evalstotug 1L and 2L+ Melanoma who received prior ICI treatment

Total (N=16) Age, y, mean (SD) 61 (14) Sex, n (%) Female 10 (62) Male 6 (38) White race, n (%) 16 (100) ECOG, n (%) 0 12 (75) 1 4 (25) Prior ICI Therapy Status, n (%) No prior ICI treatment 3 (19) Adjuvant 10 (62) Metastatic setting 3 (19) Evalstotug In Combination with PD1 in Melanoma (received prior ICI treatment; n=13): Demographics and Waterfall Evalstotug 350 mg or 700 mg Q3W (5 mg/kg to 10 mg/kg based on 70 kg patient) Data Cut Date: 31Oct25 ICI = PD1 or PD1 + LAG3 Melanoma progressed on prior ICI 54% ORR 85% DCR

54% ORR and 85% DCR in 1L and 2L+ Melanoma who received Prior ICI All 13 patients received prior ICI; evalstotug at higher doses achieves greater efficacy ipilimumuab (3mg/kg) ipilimumab (3mg/kg) + nivolumab Evalstotug (5mg/kg or 10mg/kg) + pembrolizumab Evalstotug + pembrolizumab lifileucel ipilimumab + nivolumab ipilimumuab lifileucel opdualag opdualag Data Cut Date: 31Oct25 Evalstotug Evalstotug Evalstotug (5mg/kg or 10mg/kg) + pembrolizumab opdualag Evalstotug lifileucel ipilimumab (3mg/kg) + nivolumab

Safety

Evalstotug vs Ipilimumab: Cross Trial Comparison Lower imAE rate vs Ipilimumab despite more patients previously treated with ICI Treatment Evalstotug (5 mg/kg) + PD1 Q3W N=17 Evalstotug (5 mg/kg or 10 mg/kg) + PD1 Q3W N=25 Ipilimumab (3 mg/kg) + nivolumab Q3W N=178-3141,2,3 Doses 1 – ≤18 weeks exposure (≤6 doses) 1 – ≤12 weeks exposure (≤4 doses) Tumor Types Multiple tumor types Melanoma % Patients w/ Prior Tx 90% 15%4 imAE (G3-4) 12% (no G4) 16% (no G4) 40% 1Wolchok, J; Five-Year Survival with Cmbined Nivolumab and Ipilimumab in Advanced Melanoma; N Engl J Med 2019;381:1535-1546; 2Lebbe, C; Evaluation of Two Dosing Regimens for Nivolumab in Combination With Ipilimumab in Patients With Advanced Melanoma: Results From the Phase IIIb/IV CheckMate 511 Trial; J Clin Oncol. 2019 Feb 27;37(11):867–875; 3Larkin, J. Combined Nivolumab and Ipilimumab or Monotherapy in Untreated Melanoma; N Engl J Med 2015;373:23-34; 4Allouchery, M; Safety of immune checkpoint inhibitor rechallenge after discontinuation for grade ≥2 immune-related adverse events in patients with cancer; J Immunother Cancer. 2020 Dec;8(2):e001622. Data Cut Date: 31Oct25

Evalstotug in Combination with PD1: Overall Safety Summary Generally well-tolerated; All G3 and G4 resolved Related AE Summary Total (N=17) Total (N=8) Total (N=25) Any Related Adverse Events (AEs) 16 (94%) 8 (100%) 24 (96%) Related AEs with CTCAE1 Grade 32 6 (36%) 4 (50%) 12 (48%) Grade 42 transient hypercalcemia 1 (6%) 1 (13%) 8 (32%) imAE Grade 3 or 4 5 (29%) 2 (25%) 7 (28%) Any related serious AEs2 5 (29%) 5 (63%) 10 (40%) Possibly Related AEs leading to death2 0 0 0 Related AEs leading to treatment discontinuation2 5 (29%) 3 (38%) 7 (28%) 1CTCAE: Common Terminology Criteria for Adverse Events. The NCI Common Terminology Criteria for Adverse Events is a descriptive terminology which is utilized for Adverse Event (AE) reporting. A grading (severity) scale is provided for each AE term. 2As assessed by the investigator. Missing responses are counted as related. Data Cut Date: 31Oct25

Evalstotug in Combination with PD1 Safety Data Treatment-emergent related adverse events >15% Data Cut Date: 31Oct25 Evalstotug 350 mg + PD1 (N=17) Evalstotug 700 mg + PD1 (N=8) Total (N=25) AE All Gr3+ All Gr3+ All Gr3+ Chills 11 (64.7) 0 4 (50.0) 0 15 (60.0) 0 Fatigue 11 (64.7) 0 3 (37.5) 1 (12.5) 14 (56.0) 1 (4.0) Nausea 8 (47.1) 0 4 (50.0) 0 12 (48.0) 0 Cytokine release syndrome 6 (35.3) 0 5 (62.5) 0 11 (44.0) 0 Vomiting 9 (52.9) 0 2 (25.0) 0 11 (44.0) 0 Arthralgia 9 (52.9) 0 1 (12.5) 1 (12.5) 10 (40.0) 1 (4.0) Diarrhoea 6 (35.3) 1 (5.9) 3 (37.5) 0 9 (36.0) 1 (4.0) Abdominal pain 6 (35.3) 1 (5.9) 2 (25.0) 0 8 (32.0) 1 (4.0) Pruritus 4 (23.5) 0 4 (50.0) 0 8 (32.0) 0 Rash 6 (35.3) 0 2 (25.0) 0 8 (32.0) 0 Anaemia 5 (29.4) 3 (17.6) 2 (25.0) 0 7 (28.0) 3 (12.0) Headache 6 (35.3) 0 1 (12.5) 0 7 (28.0) 0 Oedema peripheral 3 (17.6) 0 3 (37.5) 0 6 (24.0) 0 Hypokalaemia 4 (23.5) 1 (5.9) 1 (12.5) 0 5 (20.0) 1 (4.0) Decreased appetite 4 (23.5) 0 1 (12.5) 0 5 (20.0) 0 Infusion related reaction 2 (11.8) 0 3 (37.5) 0 5 (20.0) 0 Lipase increased 3 (17.6) 2 (11.8) 1 (12.5) 0 4 (16.0) 2 (8.0) Pneumonia 1 (5.9) 0 3 (37.5) 2 (25.0) 4 (16.0) 2 (8.0) Dehydration 3 (17.6) 1 (5.9) 1 (12.5) 0 4 (16.0) 1 (4.0) Back pain 3 (17.6) 0 1 (12.5) 0 4 (16.0) 0 Cough 4 (23.5) 0 0 0 4 (16.0) 0 Dyspnoea 3 (17.6) 0 1 (12.5) 0 4 (16.0) 0 Hypomagnesaemia 3 (17.6) 0 1 (12.5) 0 4 (16.0) 0 Influenza like illness 4 (23.5) 0 0 0 4 (16.0) 0 Pyrexia 3 (17.6) 0 1 (12.5) 0 4 (16.0) 0

Confirmed PR - Gastro-esophageal Cancer 63-year-old male, stage IV gastro-esophageal cancer HER2 negative, post-FOLFOX, taxane, TKI, anti-PD1 and anti-VEGFl Baseline - July 31,2023 On Treatment - October 23, 2023

Confirmed CR - Cervical Cancer 43-year-old female, stage IV cervical cancer HPV+16 positive, post-platinum, taxane, anti-PD1 and anti-VEGF “No enlarged mediastinal, hilar or axillary lymph nodes are present. There is persistent resolution of previously noted enlarged mediastinal and paraesophageal lymph nodes.” “Multiple enlarged mediastinal, paraesophageal, and right hilar lymph nodes
” Baseline – March 23, 2023 On Treatment – August 9, 2023

Evalstotug Enables the Combination with ADC and PD1 Antibodies PR (-50%) – Well Tolerated Triplet; Oz-V + Evalstotug + PD1 69 yo M with oral cavity, floor of mouth/ mandibular, mucoepidermoid carcinoma; neo-adjuvant patient Disease had recurred after prior surgery and chemoradiation (platinum / taxane / pembrolizumab as well as cetuximab). Prior to Oz-V triplet patient had tumor filling the maxillary sinus measuring 5.3cm in longest diameter. After triplet, the tumor nearly resolved and is difficult to measure. A transient, clinically asymptomatic elevation of hepatic transaminases was documented that didn't recur with subsequent dosing.

FDA Guidance Regarding Evalstotug Pivotal Trial in 1L Unresectable and / or Metastatic Melanoma Centrally reviewed PFS acceptable as primary endpoint General agreement with proposed study population and sample size Additional guidance received on ongoing dose optimization and control arm: IO-based combination regimen should be included in the control arm Project Optimus should guide determination of Phase 3 evalstotug dose IO – immuno-oncology

Overall Summary: Evalstotug + PD1 Greater efficacy and lower imAE rate vs ipilimumab + PD1 Preclinical & clinical data demonstrate that ipi & evalstotug are similar, i.e. epitope, affinity, T1/2 and tumor exposure, and efficacy ipi & evalstotug are NOT similar with respect to normal tissue environment and safety; e.g., reduced imAEs and extended treatment 54% ORR and 85% DCR in 1L and 2L+ melanoma patients who received prior ICI treatment Higher dosing yields encouraging efficacy with low incidence and severity of imAEs, consistent with CAB-driven tumor selectivity for potential best-in-class CTLA4 Significant opportunity for an effective and better tolerated CTLA-4 in combination regimens

www.bioatla.com