EXHIBIT 99.2

Topic: 26. Gene therapy, cellular immunotherapy and vaccination - Clinical

EHA-7249

First-in-human trial of LB2501, an in vivo CD19/CD20 dual targeting CAR-T therapy, in relapsed/refractory B-Cell NHL

Type: None selected

Xiaoyan Qu¹ Yajing Zhang² Haisheng Liu³ Heng Mei⁴ Kaiyang Ding⁵ Jujuan Wang¹ Sanmei Wang¹ Haorui Shen¹ Zhengxu Sun¹ Shuangshuang Xing¹ Guoai Su² Zheng Li³ Lin Liu⁴ Ran Li¹ Hailing Liu¹ Ling Zhou⁶ Yinrui Jiang⁶ Yongxin Luo⁶ Baoming Ni⁷ Dong Geng⁷ Guowei Fang⁷ Yanjie Xu⁶ Cong Feng⁶ Wenjie Wang⁶ Da Xu⁶ Zhongyuan Tu⁶ Hongchen Zheng⁶ Bing Gao⁶ Jin Liu⁶ Lei Fan¹

¹ The First Affiliated Hospital of Nanjing Medical University, Department of Hematology, Nanjing, China, ² Beijing GoBroad Boren Hospital, Gobroad Medical Institute of Hematology, Department of Myeloma and Lymphoma, Beijing, China, ³ The Fourth Hospital of Hebei Medical University, Shijiazhuang, China, ⁴ Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Institute of Hematology, Wuhan, China, ⁵ The First Affiliated Hospital of USTC Anhui Provincial Hospital, Department of Hematology, Hefei, China, ⁶ Legend Biotech Co, Nanjing, China, ⁷ Legend Biotech USA Inc, Somerset, NJ, United States of America

Background

Ex vivo CAR-T therapies have advanced treatment for relapsed/refractory non-Hodgkin lymphoma (R/R NHL) but are limited by complex manufacturing and lymphodepletion. In vivo CAR-T offers an off-the-shelf alternative that generates CAR-T cells directly in patients without requiring lymphodepletion. While feasibility has been shown in multiple myeloma, clinical data for B-cell malignancies remain unreported. LB2501, a third-generation, replication-incompetent lentiviral vector (LVV) pseudotyped with a modified fusion glycoprotein and a CD3 binder for T-cell-specific transduction, encodes a CD19/CD20 dual targeting CAR to generate CAR-T in vivo.

Aims

To evaluate the safety, pharmacokinetics, and preliminary efficacy of LB2501 in R/R B-NHL.

Methods

This ongoing phase 1 trial (NCT07002112) enrolled patients with measurable R/R NHL who had primary refractory disease or progression after ≥2 prior lines of therapy. A 3+3 dose-escalation design with backfilling was used across dose levels (DLs). LB2501 was administered as a single intravenous infusion without lymphodepletion.

Results

As of April 1, 2026, 12 patients had been dosed with LB2501 at DL1 (n=6) or DL2 (n=6). The median age was 58.5 years, median prior therapy lines was 3, and 58.3% were refractory to last line. Patients included 7 (58.3%) LBCL, 3 (25%) FL and 2 (16.7%) MCL. No DLT, SAE or fatal cases were reported. Infusion-related reactions occurred in 9 patients (75.0%); all were grade 1 or 2, resolved within a median of 2.0 days, and required no tocilizumab or glucocorticoids. Cytokine release syndrome (CRS) occurred in 66.7% patients (grade 1 in 58.3%, grade 2 in 8.3%); No neurotoxicity (ICANS) was observed. Grade≥3 LVV-related and CAR-T-related adverse events were limited to decreased lymphocyte count (33.3% each) and decreased neutrophil count (25.0% and 50.0%, respectively). The objective response rate (ORR) was 50% (6/12), with a complete response rate (CR) of 41.7% (5/12). At DL2, the ORR was 100% (6/6) and the CR was 83.3% (5/6). All responses were ongoing at cut-off date (Figure 1). The median follow-up for DL2 was 2.2 months (range, 2.0 to 3.8). PK analysis via qPCR confirmed dose-dependent in vivo CAR-T cell expansion in 83% (5/6) of patients at DL1 and 100% (6/6) at DL2. At DL2, the median Cmax was 109,117.5 copies/μg DNA (vs. 1,068.0 at DL1), with a median Tmax of 15.0 days. At the time of data cutoff, patients exhibited persistent PK, with CAR-T cells detectable in peripheral blood for up to 116 days. Insertion site analyses indicate in vivo transduction is highly T cell specific, polyclonal, and diverse with mostly single copy insertion per T cell. The median VCN for in vivo CAR-T was 1.05, lower than ex vivo manufactured CAR-T.

Summary/Conclusion

This study represents the initial clinical cohort of in vivo CAR-T therapy in relapsed/refractory non-Hodgkin lymphoma. LB2501, a first-in-class CD19/CD20 dual-targeting in vivo vector, introduces a novel paradigm by eliminating the need for ex vivo manufacturing and lymphodepletion. LB2501 has demonstrated a favorable safety profile and encouraging efficacy with a 100% ORR and 83.3% CR at DL2. While a longer-follow up is warranted, these data highlight the potential of LB2501 as a scalable, readily accessible “off-the-shelf” immunotherapy for B-cell malignancies.

Figure 1: Patient Responses to DL2