AR166, a first-in-class PD-1xLAG-3xIL-2v tri-specific immunocytokine delivering Cis -acting IL-2v to overcome immune checkpoint inhibitor resistance

Immunocytokines fusing anti-PD-1 antibody with interleukin-2 (IL-2) have demonstrated therapeutic potential in clinical settings in overcoming resistance to PD-(L)1 inhibitors. However, clinical response rates were varied, highlighting the need for improved efficacy, particularly in tumors with primary resistance to PD-1 blockade expressing high levels of LAG-3. Emerging evidence indicates LAG-3, a key immune checkpoint, acts synergistically with PD-1 to promote T cell exhaustion, while dual blockade of LAG-3 and PD-1 can reprogram CD8 + T/Tregs to enhance antitumor immunity. Here, we present AR166, a novel tri-specific immunocytokine engineered with novel Fc-silencing, delivering dual PD-1 and LAG-3 blockade while inducing PD-1-targeted CD8 + T cell activation via an optimized IL-2 variant (IL-2v). Patient-derived samples were profiled by flow cytometry and IHC. PD-1×LAG-3 bispecific activity and IL-2v cis -acting were assessed using PD-1/PD-L1 and LAG-3/MHC II blockade assays, and PD-1 and/or LAG-3-dependent CD8 + T cell pSTAT5 signaling. Fc-silencing was assessed using ADCC, ADCP, and CDC assays. Repeatedly stimulated PBMCs were co-cultured with tumor cells to determine T cell proliferation and activation in the presence of AR166. Patient-derived samples were treated with AR166 followed by immune cell profiling. Antitumor efficacy was evaluated in tumor-bearing (≥300 mm³) humanized syngeneic mouse models and tumor infiltrating lymphocytes were profiled. Complete responders were rechallenged with tumor cells to assess memory response. Pharmacokinetic and safety profiles were analyzed in hIL-2Ralphabetagamma knock-in mice and cynomolgus monkeys. Expression of PD-1 and LAG-3 was confirmed in patient-derived tumor samples. AR166 effectively inhibited PD-1/PD-L1 and LAG-3/MHC II pathways while IL-2v showed cis -activity only when anchored to PD-1 and/or LAG-3. Assessment of Fc-silencing showed minimal Fc-dependent effector functions. AR166 elicited robust T cell proliferation and activation compared to a PD-1xIL2v competitor in an in vitro T cell exhaustion model, with similar results observed in patient-derived CD8 + T cells. AR166 showed remarkable tumor growth inhibition compared to anti-PD-1 and PD-1xIL-2v competitors in humanized mouse models that recapitulate anti-PD1 resistance with high LAG-3 expression, while immune profiling showed rapid expansion of PD-1 + TCF1 + CD8 + stem-like T cells. AR166 elicited durable tumor regression in a tumor rechallenge model, suggesting tumor-specific immune memory. AR166 achieved favorable safety and pharmacokinetic profiles in both mice and cynomolgus monkeys, supporting its broad therapeutic window. Our findings show that AR166, a first-in-class PD-1xLAG-3xIL-2v tri-specific, can exceed the efficacy of both PD-1 inhibitors and PD-1xIL-2v with good safety.