I-SPY 2.2 Trial

[Slide 1] ASCO 2024 Conclusions Dato + Durva for 4 cycles was effective 33% of patients were able to go to surgery with this combination and skip traditional chemotherapy Dato + Durva was particularly effective in the Immune + response predictive subtype 20/47 (43%) of patients in the immune subtype achieved a pCR after Block A (most conservative estimate) The modeled pCR rate for the Immune+ subtype is 65% Reported toxicity profile of this combination was consistent with prior studies Data about the performance for the whole sequence of therapy will be reported at a future meeting I-SPY I The right drug. The right patient. The right time. Now.

[Slide 1] ASCO 2024 Key Takeaways ISPY 2.2 is a novel neoadjuvant trial for Mammaprint high-risk, Stage 2/3 BC that offers the opportunity to personalize treatment to maximize pCR rate for each individual patient. Dato + Durva for 4 cycles (Block A) was effective and allowed 33% of patients to go straight to surgery and skip traditional chemotherapy (i.e. skip taxane - >AC) Dato + Durva was particularly effective in the Immune+ response predictive subtype 43% of patients in the immune subtype achieved a pCR after Block A (most conservative estimate) The modeled pCR rate, based on all data after Block A, is 65% Reported toxicity profile of this combination was consistent with prior studies EARLY ESCALATION EARLY ESCALATION AFTV AFTV 000 o.o-o DATO-DxD + Durva Taxol + Carbo + PD1i 0-0-0 AC + PD1i N=106 N=66 N=26 N=20 Screen Randomize preRCB preRCB Surgery Block A N=35 N=35 Surgery Surgery

[Slide 1] ASCO 2024 New I-SPY 2.2 Design Features: Multiple Sequential Regimens Called Blocks Block A Block B Block C Surgery Investigational agents Optimal regimens Adriamycin/Cytoxan pCR and RCB without standard based on Response Adriamycin/Cytoxan + endpoints chemo across RPS Predictive Subtypes IO per SOC (RPS) and SOC Investigational agents to improve response EARLY ESCALATION EARLY ESCALATION AFTV AFTV o.o-o DATO-DxD + Durva Taxol + Carbo + PD1i AC + PD1i o N=106 N=66 N=26 N=20 Screen Randomize preRCB preRCE Surgery Block A Block B Block C All patients must screen Mammaprint High Risk N=35 N=35 to be eligible Surgery Surgery I-SPY The right drug. The right patient. The right time. Now.

[Slide 1] SAN ANTONIO BREAST CANCER 12:30 - 13:30 Rapid Fire 5 SYMPOSIUM UT Hold AACR I - - MarCime Classe - CHAIR ERICA MAYER MD, MPH, FASCO SAN ANTONIO BREAST CANCER SYMPOSIUM Methods I-SPY2 trial: ARX788 arm 5 THeat AAGR I A B C Block B Block c THIP [N-43] AC (N=2) n = 100 Block A Paula Pohlmann MD, MSc, ARX788 (N=100) N Exit Block 8 After THP . 41 (Sep 2022 to Dec 2024) Screen Randomize PhD Block 8 TCHP (N=17) Pathologic complete response rates (pCR) aner N Exit Block A *40 the novel HER2 ADC ARX788: Results from the :- SPY22 - N Ext Block D after TCMP - 17 Efficacy of ARX788 (Block A followed by surgery): Bayesian covariate adjusted model estimating pCR and Compared to fixed subtype-specific thresholds. Efficacy of ARX788 +/- subsequent chemotherapy (pCR rate Blocks A/B/C): Bayesian model considering timing of pCR Rates compared subtype specific Dynamic Control modeled from I-SPY2 data (N=1,818). an Antonio ecember 0 - o 12. 2025 --- [Slide 2] UT Health AACR - - I Hast - Career CHAIR: ERICA MAYER MD, MPH, FASCO RCB in HER2+ RPS: SAN ANTONIO BREAST CANCER SYMPOSIUM ARX788 vs. I-SPY2 control Health AACR - - ARX788 I-SPY2 CONTROL THP-AC : 4(6%) 2(7%) 1.0 MCBOYCM - 3(4%) RCB4 19(13%) 6(14,6%) MC8-36CR 12 RCB-1 0.8 (17%) 9 to 12(9%) PCB: non-pORt o (31%) RCB-1 - Paula Pohlmann MD, MSc, 0.6 : 0 PhD 7 25(61%) (24%) Pathologic complete response rates (pCR) after - 0.4 52 108(77%) the novel HER2 ADC ARX788: Results from the - 0 (73%) SPY2.2 trial RCB-0/I rate: RCB-0/I rate: 0.2 11 HER2+non-Luminal: 90% 0.2 4(9.8%) (38%) HER2+non-Luminal: 85% HER2+Luminal: 62% 6 (14 6%) - HER2+Luminal: 24% 0.0 - HER2+ HER2+ HER2-nonLuminal HER2+Luminal non-Luminal Luminal (N=141) (N=41) (N-71) (N=29) In HER2+/Luminal subtype, ARX788 improved pCR by 23.4% (38% vs 14.6%) and RCB 0/1 by 37.6% (62% vs 24.4%) compared to I-SPY2 control The probability that the pCR rate of ARX788 treatment strategy is higher than dynamic control in HER2+/Luminal is 0.99 --- [Slide 3] BREAST CANCER SYMPOSIUM UT Health AACR - - Mays - I CHAIR: ERICA MAYER MD, MPH, FASCO SAN ANTONIO BREAST CANCER Results: Safety SYMPOSIUM UT Health AACR - ADC Corneal.pseudomicrocysts ¥ ¥ Macropinocytosis inhibitor eye drops Lysosome ¥ Paula Pohlmann MD, MSc, Macropinosome PhD K if Pathologic complete response rates (pCR) after the novel HER2 ADC ARX788: Results from the 1- Payload release Neel Pasricha, MD SPY2.2 trial Lindgren ES et al. Curr Ophthalmol Rep. 2024 Ocular toxicity: 95% of patients (9% grade 3) Pneumonitis in 6% (2% grade 3) 10 (10%) Dues APOLTER 21 Revelved 67% pts completed all 4 doses of ARX-788 (21%) 2(2%) 8 patients discontinued ARX788 due to adverse events. 67 (67%) No treatment related deaths San Antonio December 9 - 12. 2025 --- [Slide 4] BREAST CANCER SYMPOSIUM UT Health AACR - - Caram CHAIR: ERICA MAYER MD, MPH, FASCO SAN ANTONIO BREAST CANCER Conclusions SYMPOSIUM UT Health AAGR - I Sequential ARX788 followed by standard anti-HER2 therapy demonstrated high efficacy despite ocular toxicity pCR 63% and RCB 0/1 82% is consistent with 2nd generation anti-HER2 ADC Paula Pohlmann MD, MS performance PhD In HER2+/Luminal subtype, +23% pCR improvement VS I-SPY2 controls Pathologic complete response rates (pCR) at the novel HER2 ADC ARX788: Results from SPY2.2 trial Many patients avoided standard chemotherapy (28%) and most completely avoided AC Incidence of ILD was low and there were no fatalities Ongoing studies aim to mitigate ocular toxicity from ADCs I-SPY2.2 continues to demonstrate how response-adaptive neoadjuvant therapy can be tailored to each individual patient [NCT01042379]. San Antonio December 9 - 12, 2025

OCR text not available for this slide. View the original post on X for context.