HARMONi-A Trial

[Slide 1] Ivonescimab Mechanism of Action (MOA) Intended for HCPs/Investors/Media Designed to Optimize the Balance of Anti-tumor Activity and Safety1,2 Anti-VEGF Potential to drive SYNERGISTIC anti-tumor activity1,3.4 Simultaneous Blocking Cooperative of PD-1 & VEGF1,3,4 Binding5 Engineered Fc-null Region Linkers Anti-PD-1 T Cell Tumor Cell Anti-Angiogenesis VEGF Dimer Y PD-1 Receptor in T Cell --- [Slide 2] Acquired resistance to osimertinib in 1L Methods to detect AR mechanisms 15% ~30% ~15% Mechanisms of On-target mutations By-pass mechanisms Histologic transformation AR 4th Generation EGFR TKI EGFR TKI + TKI anti-bypass SCLC: Platinum-Etoposide 1st Generation EGFR TKI Sq.: Platinum-based CT Platinum-pemetrexed CT Potential treatment strategies to Amivantamab + CT +/- Lazertinib overcome AR Antibody drug conjugated Antibody drug conjugated + Osimertinib Inclusion in a clinical trial

[Slide 1] Immunotherapy at EGFR TKI PD in EGFRm NSCLC Study Treatment N PFS HR; 95%IC OS HR; 95%IC Nivolumab + PBC CheckMate 722 296 5.6 vs. 5.4 0.75; 0.56-1.00 19.4 vs. 15.9 vs. PBC 0.82; 0.61-1.10 Pembrolizumab + PBC KEYNOTE 789 480 5.6 vs. 5.5 VS PBC 0.80; 0.65-0.97 15.9 vs. 14.7 0.84; 0.69 -1.02 Sintilimab + PBC ORIENT 31 318 5.5 vs. 4.3 0.72; 0.55-0.94 20.5 vs. 19.2 vs. PBC 0.97; 0.71 -1.32 Atezolizumab + BVZ + PBC IMPOWER 150 58 10.3 vs. 6.1 0.41; 0.23-0.75 29.4 vs. 18.1 vs. BVZ + PBC 0.60; 0.31 -1.14 Atezolizumab + BVZ + PBC IMPOWER 151 163 8.5 vs. 8.3 vs. BVZ + PBC 0.86; 061-1.21 NR NR Atezolizumab + BVZ + PBC ATLAS 215 8.4 vs. 5.6 0.62; 0.45-0.86 20.6 vs. 20.3 vs. PBC 1.01; 0.69 -1.46 Sintilimab + IBI305 + PBC ORIENT 31 318 7.2 vs. 4.3 0.51; 0.39-0.67 21.1 vs. 19.2 VS. PBC 0.98; 0.72 -1.34 Atezolizumab + BVZ + PBC ABC-Lung 95 6.3 vs. 7.5 NR 15.4 vs. 15.5 NR vs. Atezolizumab + BVZ + Pem Ivonescimab + PBC HARMONi-A 322 7.1 vs. 4.8 0.46; 0.34-0.62 NR NR VS. PBC + Placebo Mok JCO 2024 Yang ASCO 2023 Lu LRM 2023 Reck LRM 2019 Nogami JTO 2021 Zho WCLC 2023 Ahn JCO 2024 Frueh ELCC 2024 Zhang ASCO 2024 (PBC: platinum-based chernotherapy. BVZ: Bevacizumab. Cross trial comparison should be undertaken with caution)

[Slide 1] Estimated sale 320 patients Analysis me A stratified 1 PFS was es regression All data (ex at which p 2024 ASCO #ASCO24 ANNUAL MEETING OF 2024ASCO --- [Slide 2] 2024 ASCO ANNUAL MEETING Ivonescimab combined with chemotherapy in patients with EGFR-mutant non-squamous non-small cell lung cancer who progressed on EGFR-TKIs treatment: a randomized, double-blind, multi-center, phase 3 trial (HARMONi-A study) Li Zhang¹, Wenfeng Fang¹, Yuanyuan Zhao¹, Yongzhong Luo², Runxiang Yang³, Yan Huang¹, Zhiyong He⁴, Hui Zhao⁵, Mingjun Li⁶, Kai Li⁷, Qibing Song⁸, Xiaobo Du⁹, Yulan Sun¹⁰, Wei Li¹¹, Fei Xu¹², Zhiyu Wang¹³, Kunning Yang¹⁴, Yun Fan¹⁵, Wenting Li¹⁶, Michelle Xia¹⁶ 'Sun Yat-sen University Cancer Center, Guangzhou, China: Hunan Cancer Hospital, Changsha, China; Yunnan Cancer Hospital, Kunming, China; Fujian Provincial Tumor Hospital, Fuzhou, China; SThe Second Hospital of Anhui Medical University, Hefei, China; The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China; Tianjin Medical University Cancer Institute&Hospital. Tianjin, China; Renmin Hospital of Wuhan University, Wuhan, China; Mianyang Central Hospital, Mianyang, China; "Shandong Cancer Prevention and Treatment Institute, Jinan, China; "The First Affiliated Hospital of Bengbu Medical University, Bengbu, China; 12The First Affiliated Hospital of Nanchang University, Nanchang, China; "The Fourth Hospital of Hebei Medical University, Shijiazhuang, China; "Weifang No.2 People's Hospital, Weifang, China; 15Zhejiang Cancer Hospital, Hangzhou, China; 16Akeso Biopharma, Inc., Zhongshan, China 2024 ASCO #ASCO24 PRESENTED BY: Li Zhang, MD ASCO AMERICAN SOCIETY Of CLINICAL ONCOLOGY ANNUAL MEETING Presentation is property of the author and ASCO Permission required for reuse, contact permissions@asco.org KNOWLEDGE CONQUERS CANCER --- [Slide 3] Background For patients with EGFR-mutant NSCLC, upfront treatment with tyrosine kinase inhibitors is standard. However, drug resistance Anti-VEGF remains a challenge, and an effective therapy after progression is needed. Ivonescimab (AK112/SMT112) is an anti-PD-1/VEGF bispecific antibody displaying cooperative binding characteristics. Engineered Fc-Null Region Phase II clinical studies have shown potential efficacy of Ivonescimab plus chemotherapy in NSCLC patients with EGFR mutations who progressed on prior EGFR-TKIs therapies¹⁻². This phase 3 study aimed to evaluate and confirm the efficacy Anti-PD-1 and safety of ivonescimab combined with chemotherapy compared to chemotherapy alone in this population (NCT05184712). 1.L Zhang et al: ASCO 2023; 2. YY Zhao et al: eClinicalMedicine 2023;62: 102106. 2024 ASCO #ASCO24 PRESENTED BY: Li Zhang, MD ASCO AMERICAN SOCIETY OF CLINICAL ONCOLOGY ANNUAL MEETING Presentation is property of the author and ASCO Permission required for reuse, contact permissions@asco.org KNOWLEDGE CONQUERS CANCER --- [Slide 4] HARMONi-A Study Design Ivonescimab Ivonescimab Key Eligibility Criteria 20mg/kg 20mg/kg Q3W + nsq-NSCLC (Stage IIIB/C ineligible for surgery or + Treatment until: Pemetrexed 500 mg/m2 Pemetrexed 500 local therapy and IV) Carboplatin AUC5 mg/m2 Q3W Intolerable toxicity EGFR sensitive mutation positive Q3W for 4 cycles No clinical benefit ECOG PS 0 or1 R (1:1) Initiation of new Regardless PD-L1 expression anti-tumor therapy Placebo Placebo Stratification Factors: + Q3W Up to 24 months Exposure to 3rd EGFR-TKI before (yes vs. no) Pemetrexed 500 mg/m2 + Carboplatin AUC5 Pemetrexed 500 Brain metastases (yes vs. no) Q3W for 4 cycles mg/m2 Q3W Endpoints Primary: Progression-free survival by independent radiologic review committee (IRRC) Secondary: Overall survival, Response rate, Duration of response, Time to response and Safety ClinicalTrials.gov, NCT05184712; NSCLC, non-small cell lung carcinoma; EGFR, epidermal growth factor receptor; ECOG, eastern copperative oncology group; TKI, tyrosine-kinase inhibitor; Q3W, every 3 weeks. 2024 ASCO #ASCO24 PRESENTED BY: Li Zhang, MD ASCO AMERICAN SOCIETY OF CLINICAL ONCOLOGY ANNUAL MEETING Presentation is property of the author and ASCO Permission required for reuse contact permissions@asco.org KNOWLEDGE CONQUERS CANCER

[Slide 1] Final analysis of overall survival (OS) Data cut-off date: April 2025 100 (median follow-up of 32.5 months) Ivonescimab+Chemotherapy 90 Placebo+Chemotherapy Ivonescimab Placebo + + 80 12-month rate 24-month rate 30-month rate chemotherapy chemotherapy 65.0% 35.3% 29.1% 70 60.0% 28.8% 18.4% ORR¹, % 50.6 35.4 (95% CI) (42.6, 58.6) (28.0, 43.3) DCR¹, % 83.2 Overall Survival Rate (%) 60 93.1 (95% CI) (88.0, 96.5) (76.5, 88.6) 50 mPFS1, months 7.1 4.8 40 (95% CI) (5.9, 8.7) (4.2, 5.6) PFS HR1 0.46 (0.34, 0.62) 30 (95% CI) p<0.001 mOS², months 16.8 14.1 20 (95% CI) (14.5, 20.0) (12.8, 16.3) 10 os HR² 0.74 (0.58, 0.95) (95% CI) p=0.019 (two-side) 0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 Interim analysis Number of Risk Final analysis Ivonescimab+Chemotherapy 161 159 155 143 132 117 104 95 85 74 68 61 55 47 43 33 27 10 5 0 Placebo+Chemotherapy 161 159 152 138 124 109 96 81 69 63 55 50 46 40 33 22 13 6 0 HR and P-value were stratified by previous 3rd Gen EGFR-TKI ues (yes VS. no) and presence of brain metastases (yes VS. no), TH 4 Society for Immunotherapy of Cancer and were calculated with stratified Cox model and log rank test. The two-sided P-value boundary is 0.05 as calculated using sitc 2025 Lan-Demets spending function with O'Brien-Fleming approximation; HR, hazard ratio; CI, confidence interval.

[Slide 1] Achieved comprehensive coverage across core lung cancer indications, establishing a complete multi-line therapy portfolio Ivonescimab has achieved comprehensive coverage across core lung cancer indications and established a complete multi-line therapy portfolio, demonstrating its potential to transform the global treatment paradigm for advanced lung cancer. The first approved indication of ivonescimab is for the treatment of EGFR-mutated, locally advanced or metastatic non-squamous NSCLC progressed after EGFR-TKI treatment. This indication was approved in Mav 2024 and successfully included in the NRDL in 2025. In August 2025, the Company announced that the final OS analysis of this clinical trial showed that ivonescimab met the OS clinical endpoint, demonstrating a statistically significant and clinically meaningful OS benefit. Detailed results of this study will be presented at an upcoming medical conference. During the Reporting Period, this treatment received 4 authoritative clinical guidelines as recommendation, including a Class I recommendation in the CSCO Guidelines for the Diagnosis and Treatment of Non-Small Cell Lung Cancer (2025). In April 2025, the sNDA for ivonescimab as monotherapy for the first-line treatment of PD-L1-positive, locally advanced or metastatic NSCLC received approval from the NMPA. This marks the second approved indication for ivonescimab, providing a novel, efficacious and safe "chemo-free" regimen for first-line NSCLC. This treatment was granted a prominent recommendation in the CSCO Guidelines for the Diagnosis and Treatment of Non-Small Cell Lung Cancer (2025).

[Slide 1] Final analysis of overall survival (OS) Data cut-off date: April 2025 (median follow-up of 32.5 months) - I . vonescimab Placebo + + % 12-month rate 24-month rate chemotherapy 10-munth rate chemotherapy 45.0% 15.7% 29.1% ORR', % 50.6 35.4 (28.0,43.3) DCR', % 93.1 83.2 Overall Survival Rate (%) 70 60 0% % 28.8% 18.0% (95% CI) (42.6,58.6) 0 (95% CI) (88.0,96.5) (76.5,88.6) % mPFS', months 7.1 4.8 (95% CI) 45 (5.9,8.7) (4.2,5.6) PFS HR1 0.46 (0.34, 0.62) * (95% CI) p<0.001 mOS2, months 16.8 14.1 * (95% CI) (14.5,20.0) (12.8,16.3) 15 os HR2 0.74 (0.58,0.95) (95% CI) p=0.019 (two-side) 1 I 1. . - - Interim analysis $ 11 14 it 11 . 22 24 a a a 12 M a . Number If Rok Final analysis - E de in to If 11 - is " . . " " e 6 . D 0 - . - - " 14 OF 8 M " " 0 = " R n a : - 1 - HR and P-value were stratified by previous " Gen EGFR-T usa (yes is. no) and presence of brain metastases (yes n. no), TH and were calculated with stratified Cox model and log rank test, The two-sided P-value boundary is 0.05 as calculated using 4 - sitc 2025 Lan-Demets spending function with 0 Brien Pleming approximation: HR, hazerd ratio, CL confidence interval.

[Slide 1] ASCO Sign in Ivonescimab combined with chemotherapy in patients with EGFR-mutant non-squamous non-small cell lung cancer who progressed on EGFR tyrosine- kinase inhibitor treatment (HARMONi-A): A randomized, double-blind, multi-center, phase 3 trial. Abstract Authors Li Zhang i Sun Yat-sen University Cancer Center, Guangzhou, China Li Zhang, Wenfeng Fang, Yuanyuan Zhao, Yongzhong Luo, Runxiang Yang, Yan Huang, Zhiyong He, Hui Zhao, Mingjun Li, Kai Li, Qibin Song, Xiaobo Du, Yulan Sun, Wei Li, Fei Xu, Zhiyu Wang, Kunning Yang, Yun Fan, Wenting Li, Yu Xia --- [Slide 2] Background: Ivonescimab (AK112/SMT112) is a anti-PD- 1/VEGF bispecific antibody. Previous phase I/II clinical studies have shown potential efficacy of ivonescimab in NSCLC patients with EGFR mutations who had failed prior EGFR-TKIs therapies. This phase 3 study aimed to evaluate and confirm the efficacy and safety of ivonescimab combined with chemotherapy versus chemotherapy alone in this population. Methods: Patients were randomized 1:1 to receive ivonescimab (20 mg/kg) plus pemetrexed (500 mg/m²) and carboplatin (AUC 5) or placebo plus chemotherapy once every 3 weeks for four cycles, with stratification according to the third- generation EGFR-TKI (received vs not received) and brain metastases (presence vs absence), followed by maintenance therapy of ivonescimab and pemetrexed or placebo and pemetrexed. The primary endpoint was progression-free survival (PFS) in intention-to-treat (ITT) population assessed by independent radiographic review committee (IRRC) per RECIST v1.1. Here we report the results of the first planned interim analysis. --- [Slide 3] Results: Total 322 patients were randomized (161 to the ivonescimab plus chemotherapy arm, 161 to the placebo plus chemotherapy arm). 86.3% versus 85.1% of patients had received the third generation EGFR-TKIs treatment, 21.7% versus 23.0% of patients had brain metastases. As of March 10, 2023, median follow up time was 7.89 months. PFS was significantly improved in the ivonescimab plus chemotherapy arm (HR 0.46 [0.34, 0.62], P < 0.0001). Median PFS (95%CI) by IRRC were 7.06m (5.85, 8.74) in the ivonescimab arm versus 4.80m (4.21, 5.55) in chemotherapy arm. The prespecified subgroup analysis showed PFS benefit favoring patients receiving ivonescimab over placebo across almost all subgroups, including in patients who progressed on the third-generation EGFR-TKIs therapy (HR 0.48, 95% CI 0.35-0.66), those with brain metastases (HR 0.40, 0.22-0.73), those with EGFR mutation of deletion 19 (HR 0.48, 0.32- 0.73), and individuals with T790M mutation positive (HR 0.22, 0.09-0.54). The ORR were 50.6% and 35.4%, respectively. Grade ≥3 TEAEs occurred in 99 (61.5%) patients versus 79 (49.1%) patients, the most common grade ≥3 TEAEs were chemotherapy related adverse events. Grade ≥3 immune-related adverse events occurred in 10 (6.2%) patients versus 4 (2.5%). Grade ≥3 VEGF blocking related adverse events occurred in 5 patients (3.1%) versus 4 patients (2.5%). --- [Slide 4] Conclusions: Ivonescimab plus chemotherapy significantly improved PFS while maintaining a manageable safety profile in patients who had failed EGFR- TKIs treatments. Clinical trial information: NCT05184712.

[Slide 1] Safety Summary Ivonescimab + Placebo + TRAE, n(%) Chemotherapy Chemotherapy (N=161) (N=161) Any grade 158 (98.1) 153 (95.0) Grade>3 87 (54.0) 69 (42.9) Serious* 46 (28.6) 26 (16.1) Led to discontinuation of ivonescimab/placebo 9 (5.6) 4 (2.5) Led to death 0 (0.0) 0 (0.0) Grade>3 immune-related 10 (6.2) 4 (2.5) Grade>3 VEGF-related 5 (3.1) 4 (2.5) . For any PT (excluding PO) in SAE the PT with more than 2 cases in the experimental group compared to the control group were plates count decreased (7.5% vs. 4.3%) and hepatic function abnormal (2.5% vs. 0%) TRAE, treatment-colated adverse event (related to any drug) PT. preferred term PD. disease progression: BAE serious adverse event 2024 ASCO #ASCO24 RESORTS or u Zhang, MD ASCO MEDICAN SOCIETY OF COMICAL ONCOLOGY ANNUAL MEETING - spreats w'the satur and ANDO KNOWLEDGE CONQUERS CANCER --- [Slide 2] ORR, DCR and DoR per IRRC 100% Ivonescimab Placebo 90% 93.1 Chemo Chemo 95% 80% 188.0 56.5) 83.2 95% CI 70% (76.5, 88.6) ORR, % 50.6 35.4 60% (95% CI) (42.6, 58.6) (28.0, 43.3) 50% 50.6 40% 95% CI (42.0, 58.6) DCR, % 93,1 83.2 30% 35.4 (95% CI) (88.0, 96.5) (76.5, 88.6) 95% CI 20% (28.8, 43.3) 10% Median DoR, 6.6 4.2 0% month (4.3,7.6) Ivonescimab (3.0, 4.7) Chemotherapy Placebo . Chemotherapy (95% CI) RD, rate difference: a confidence interval RD and a were calculated using Mattinen-Nummer method stratified by exposure to you generation EGFR-TK before (yes v3.no) and brain metastases (yes vs.no) 2024 ASCO #ASCO24 PRESENTED UT U Zhang MD ASCO LOCITY OF CONICAL CACOLOGY ANNUAL MEETING . - of E - ASCO KNOWLEDGE CONQUERS CANCER