AUTHORS
Yingyong Xu, MD, PhD, MBA, tianying chen2, jiajia wu3, Xu Feng4, Jiapan Hu5, lina zhou6, xiaohui ding7, pengying zhang8, jie zhou9, xiu zhao10, li he11, zhihong Lv12, xiaobin song13, yixuan zhou14, fuqiang peng15, zhijiang wang16, zhixiang yan17, nian chen18, yuqing lin19, lihong shen20, tony ho21, Yanan Yue22, Jianxiang Wang Sr, MD, PhD23, Luhan Yang, PhD24
AFFILIATIONS
1 Qihan Biotechnology, Hangzhou, China
2 Qihan Biotechnology, Hangzhou, China
3 Qihan Biotechnology, Hangzhou, China
4 Qihan Biotechnology, Hangzhou, China
5 Qihan Biotechnology, Hangzhou, China
6 Qihan Biotechnology, Hangzhou, China
7 Qihan Biotechnology, Hangzhou, China
8 Qihan Biotechnology, Hangzhou, China
9 Qihan Biotechnology, Hangzhou, China
10 Qihan Biotechnology, Hangzhou, China
11 Qihan Biotechnology, Hangzhou, China
12 Qihan Biotechnology, Hangzhou, China
13 Qihan Biotechnology, Hangzhou, China
14 Qihan Biotechnology, Hangzhou, China
15 Qihan Biotechnology, Hangzhou, China
16 Qihan Biotechnology, Hangzhou, China
17 Qihan Biotechnology, Hangzhou, China
18 Qihan Biotechnology, Hangzhou, China
19 Qihan Biotechnology, Hangzhou, China
20 Qihan Biotechnology, Hangzhou, China
21 Qihan Biotechnology, Hangzhou, China
22 Qihan Biotechnology, Hangzhou, China
23 Institute of Hematology & Blood Diseases Hospital, CAMS/PUMC, tianjing, China
24 Qihan Biotechnology, Hangzhou, China
ABSTRACT
Background: Allogeneic universal CAR-T (UCAR-T) therapy offers transformative potential for autoimmune diseases (AID), but host immune rejection remains the major barrier. While TCR knockout prevents graft-versus-host disease (GVHD), addressing host-versus-graft rejection (HVGR) driven by recipient alloreactive T cells and NK cells is far more challenging. Alloreactive CD8+ T cells recognize mismatched HLA class I molecules, while NK cells mediate "missing-self" killing when HLA class I is disrupted to evade T cell attack. This creates a fundamental trade-off in hypoimmune engineering. We compared eight dual CD19/BCMA-targeting UCAR-T prototypes across four hypoimmune strategies to identify principles for optimal product design.
Methods: Eight UCAR-T prototypes were engineered: (1) donor-recipient HLA-X matching (QT-019B, QT-219BG); (2) intensified lymphodepletion (QT-219CX); (3) counteract NK-killing (QT-119A); and (4) AI-guided engineering of HLA-X variants for simultaneous evasion of T cell and NK cell immunity (QT-019A, QT-019C, QT-219C1, QT-219C2). All incorporated TCR knockout. Prototypes with promising preclinical data were advanced to IIT in autoimmune indications. Cellular kinetics (Cmax, AUC), safety, and efficacy were assessed.
Results: A total of 44 patients were enrolled across all studies, with main indications including SLE, MS, AIHA, IgAN. Most patients received QT-019B (n=25) and QT-219C2 (n=5); while other products have completed enrollment with 1–3 patients each. Dose range was 3×105 to 6×106 cells/kg, with most patients receiving 1–3×106 cells/kg.
HLA-X matching (QT-019B, QT-219BG): QT-019B (n=25) demonstrated robust preclinical profiles; most patients achieved favorable Cmax and AUC with tolerable safety and promising clinical responses. QT-019B was advanced to IND with FDA FTD, RMAT, and BTD. QT-219BG (n=2) showed similar promise but was deprioritized in favor of QT-019B.
Intensified lymphodepletion (QT-219CX): Enrolled 1 patient. Suboptimal Cmax and AUC were observed despite strong preclinical data, reflecting host immune resilience. Further development was discontinued due to unfavorable risk-benefit.
Counteract NK-killing (QT-119A): Enrolled 4 patients. Highly heterogeneous PK—excellent expansion in some but poor in others, which is attributable to NK cell repertoire variability. Further development was discontinued due to PK inconsistency.
AI-Guided Engineering of HLA-X Variants for Simultaneous Evasion of T Cell and NK Cell Immunity (QT-019A, QT-019C, QT-219C1, QT-219C2): QT-019A (n=3), and QT-219C1 (n=2) showed pronounced rejection and were discontinued. QT-019C (n=2) and QT-219C2 (n=5) demonstrated favorable Cmax, AUC, and clinical results, supporting further IND-enabling development.
Regarding safety, overall tolerability was good for all 44 patients. The most common adverse events were fever or Grade 1 CRS, hypogammaglobulinemia, and Grade 1–2 infections. No severe adverse events, ICANS, or Grade 5 events were reported.
Conclusions: Systematic engineering can substantially mitigate host-versus-graft rejection, though complete immune evasion remains elusive. The optimal profile features moderate Cmax with broad AUC, suggesting that durable exposure, rather than maximal peak expansion, may be associated with a more favorable clinical profile. Drug design should prioritize generalizable benefit over isolated efficacy; intensified immunosuppression increases infection risk, particularly critical in autoimmune disease populations. These findings establish a framework for rational hypoimmune engineering in next-generation dual-target UCAR-T therapies.
DISCLOSURES
Luhan Yang, PhD: No relevant conflicts of interest to declare.