Advances in LAG3 cancer immunotherapeutics

Abstract

Cancer treatment has entered the age of immunotherapy. Immune checkpoint inhibitor (ICI) therapy has shown robust therapeutic potential in clinical practice, with significant improvements in progression-free survival (PFS) and overall survival (OS). Recently, checkpoint blockade of the lymphocyte activation gene 3 (LAG3) inhibitory receptor (IR) in combination with programmed death protein 1 (PD1) inhibition has been FDA approved in patients with advanced melanoma. This has encouraged the clinical evaluation of new LAG3-directed biologics in combination with other checkpoint inhibitors. Several of these studies are evaluating bispecific antibodies that target exhausted T (T_EX) cells expressing multiple IRs. This review discusses the current understanding of LAG3 in regulating antitumor immunity and the ongoing clinical testing of LAG3 inhibition in cancer.

Advent of LAG3 cancer immunotherapeutics

Immune checkpoint receptors serve an essential role in limiting T cell function and preventing the development of autoimmunity. However, in chronic infection and cancer, IRs hinder the immune system, enabling the persistence and progression of disease. Furthermore, tumors can hijack these immune checkpoint mechanisms to protect themselves from T cells. Thus, many cancer therapeutics target IRs to enhance T cell antitumor responses. The most commonly used immunotherapies are ICIs. These immunotherapies have obtained durable clinical responses, but efficacies vary, and only subsets of cancer patients benefit.

The primary goal of ICIs is to enhance antitumor function by blocking IRs on T_EX cells. T cell exhaustion was initially described in nonresponsive antigen-specific CD8⁺ T cells that exhibited reduced cell proliferation and effector cytokine production [1,2]. Later studies found that the majority of tumor-infiltrating lymphocytes (TILs) share similar characteristics of exhaustion [3]. A hallmark of T_EX cells is the chronic surface expression of multiple IRs, including LAG3, CTLA-4, PD1, TIM3, and TIGIT. For this reason, IRs are commonly referred to as markers of exhaustion. While all T_EX cells express IRs, they do not alone identify T_EX cells as their expression is not limited to these cells. Instead, the duration and combination of IRs distinguishes T_EX cells from other populations, along with other parameters [3-6]. Thus, the co-targeting of multiple IRs has become a leading strategy for new cancer treatments.

Immunotherapies have made significant advancements in the past decade. Initially, ICIs were used only to treat patients with advanced cancers that were no longer responding to standard treatments. Currently, anti-PD1/PDL1 and anti-CTLA4 immunotherapy is the preferred regimen for patients with metastatic melanoma, showing improvements in OS, response rate, and durability of response. However, patients receiving nivolumab (anti-PD1) and ipilimumab (anti-CTLA4) in combination have higher incidences of severe immune-related adverse events (irAEs), leading many patients to discontinue treatment or opt for monotherapy [7,8]. Thus, there is an urgent need for more tolerable therapeutics.

LAG3 was characterized in the 1990s by Frédéric Triebel in activated T and natural killer (NK) cells [9]. Later, LAG3 was shown to be co-expressed with PD1 on TILs. Antibody blockade of both LAG3 and PD1 promoted synergistic antitumor immunity [5]. Recently, the anti-LAG3 monoclonal antibody (mAb) relatlimab, in combination with nivolumab, became the third ICI to be approved by the FDA. In the RELATIVITY-047 Phase 3 trial, relatlimab and nivolumab combination therapy more than doubled the rate of PFS compared with nivolumab monotherapy in patients with metastatic melanoma [10]. Importantly, the safety profile with anti-PD1 + anti-LAG3 was substantially more favorable than the anti-PD1 + anti-CTLA4 (nivolumab + ipilimumab) combination. Despite these impressive clinical results, little is known of the function, critical ligands, and underlying synergistic mechanisms of LAG3 with PD1. Furthermore, the optimal treatment regimen, and whether LAG3 could synergize with other IRs such as CTLA4, TIM3, TIGIT and other immuno-oncology targets, are unclear. In this review we summarize the current understanding of the impact of LAG3 in cancer and examine the ongoing clinical strategies targeting this IR (Figure 1, Key figure).

Key figure

Lymphocyte activation gene 3 (LAG3) therapeutic strategies

Figure 1.

Combination immune checkpoint inhibitor (ICI) therapy using monospecific antibodies: therapeutic strategy in which multiple inhibitory receptors (IRs) are targeted by combination treatment with two or more separate ICIs, such as anti-programmed death protein 1 (anti-PD1) monoclonal antibody (mAb) and anti-LAG3 mAb. Bispecific antibodies (BsAbs) targeting IRs on immune cells: BsAbs selectively binding T cells that express multiple IRs, such as LAG3 and PD1 or CTLA4. BsABs directed at cell-to-cell interactions: BsAbs designed to bind tumor-adjacent T cells by targeting tumor-expressing ligand (e.g., PD1 ligand, PDL1) and T cell IR (e.g., LAG3) or designed to bind and stabilize dendritic cell (DC) and T cell interactions to enhance cell activation by targeting DC-expressing surface receptor (e.g., FcγR) and T cell IR (e.g., LAG3). Abbreviation: TCR, T cell receptor.

LAG3 structure

LAG3 (CD223) is a type 1 cell surface molecule expressed on activated T, B, and NK cells. Importantly, LAG3 is not expressed on naïve T cells; instead, its expression is induced upon T cell receptor (TCR) stimulation of CD4⁺ and CD8⁺ T cells [11,12]. This includes regulatory T (Treg) cells, in which LAG3 has been shown to mediate Treg function and limit cell proliferation and effector function, depending on the environmental context [13,14]. LAG3 consists of four Ig-like extracellular domains (D1, D2, D3, and D4), a connecting peptide (CP), a transmembrane domain, and an intracellular domain [9,15-17]. The intracellular domain comprises three highly conserved motifs that mediate its inhibitory function: FxxL, KIEELE, and a glutamic acid–proline-rich region (EP) at the C terminus [18-24].

LAG3 homodimerization is required for ligand binding and downstream inhibitory signaling [25]. Recent crystal structures have revealed several hydrophobic residues within the D2 domain that contribute to LAG3 dimerization (mouse: Trp180 and Leu221) [25,26]. Mutations in the LAG3 D2 domain prevented tracking to the TCR/CD3 complex and co-localization to CD45 within the immune synapse in murine CD8⁺ T cells [27]. Furthermore, this abrogated LAG3 function, demonstrating enhanced antitumor immunity in a murine melanoma tumor model. It was recently shown that glycosylation at Asp184 within the D2 domain could also disrupt dimerization, albeit to a lesser extent than the hydrophobic residues [25]. Currently, anti-LAG3 therapeutics target the D1 domain to inhibit LAG3 from binding to its ligands. While the anti-D2 murine mAb C9B7W has been shown to disrupt LAG3 dimerization and enhance tumor clearance and survival in murine tumor models [5,27,28], the efficacy of D2-directed mAbs has yet to be assessed in trials.

LAG3 ligands

The canonical ligand of LAG3 is major histocompatibility complex II (MHCII) (Figure 2) [22,29-33]. Studies have shown that LAG3 binding to the peptide–MHCII (pMHCII) complex induces T cell inhibition. This interaction is attributed to the affinity of the peptide for MHCII, which stabilizes the pMHCII complex. Disrupting this interaction by targeting the D1 domain of LAG3 improved T cell-mediated antitumor activity in human patients and murine tumor models [10,34,35]. Recently, the crystal structure of single-chain mouse LAG3-MHCII demonstrated that CD4 and LAG3 bind to overlapping regions on the α2 and β2 domains of MHCII [36]. LAG3 binds to MHCII with ~1000-fold greater affinity than CD4. These findings suggest LAG3 may mediate T cell regulation by competing with CD4 for MHCII binding. Fibrinogen-like protein 1 (FLG1) is another ligand of LAG3 that has been correlated with poor prognosis in hepatocellular carcinoma (HCC) and is upregulated in a subtype of non-small-cell lung cancer (NSCLC) [37-39]. Like MHCII, FLG1 binds to the D1 domain of LAG3, but at an adjacent site. Blocking the FGL1-LAG3 axis promoted antitumor immunity in a murine colorectal cancer model [37]. The secretion of FGL1 by hepatocytes has been linked with immunotherapy resistance in metastatic cancers, and soluble FGL1 has been associated with reduced immunotherapy response [40].

Figure 2. Role of Lymphocyte activation gene 3 (LAG3) in the tumor microenvironment (TME).

Left: LAG3 homodimers colocalize to T cell receptor (TCR)/CD3 and inhibit downstream signaling. LAG3 function can be modulated by shedding via a disintegrin and metalloproteinase (ADAM) 10/17 to release soluble LAG3 (sLAG3). Middle: major histocompatibility complex II (MHCII) is the canonical ligand of LAG3. The glutamic acid–proline-rich (EP) motif of LAG3 causes disassociation of Lck from CD4 and CD8 coreceptors to inhibit T cell signaling. Right: FGL1 expressed on tumor cells or secreted (sFLG1) can bind LAG3 to attenuate T cell function. Abbreviation: APC, antigen-presenting cell.

Recently, the TCR-CD3 complex was demonstrated to serve as a ligand in cis. LAG3 bound to the TCR-CD3 complex and accumulated within the immune synapse to inhibit proximal signaling in the presence or absence of MHCII [24]. The murine anti-LAG3 C9B7W mAb that is specific for the D2 domain disrupted LAG3 homodimers and reduced localization of LAG3 to the TCR [27]. Currently, it is unknown which chain in the TCR-CD3 complex LAG3 interacts with, or whether there is an intermediary. Other cancer-associated ligands for LAG3 may include LSECtin and galectin-3 (GAL-3); however, the clinical application of these ligands in cancer has been questioned, with recent data demonstrating that LSECtin and GAL-3 were unable to inhibit reporter-cell activation [41,42]. Further investigation is required to elucidate the role of these ligands and their order of importance in mediating LAG3 function.

LAG3 function

LAG3 serves to limit T cell activation by balancing co-stimulatory signals to regulate proliferation and effector functions [19-21]. LAG3 binding to its ligands inhibits proximal signaling, thereby attenuating cell expansion and effector function of T cells [19,21,43]. This is critical to limit autoimmunity by restraining pathogenic T cell clones that would contribute to chronic TCR stimulation [12,44-46]. The EP motif of LAG3 lowers the local pH near the CD4 and CD8 co-receptors, sequestering Zn²⁺, thus resulting in Lck dissociation from these co-receptors [24]. Deletion of the EP motif increased accumulation of phosphorylated ZAP-70, resulting in greater signaling distal to the TCR-CD3 complex. Interestingly, a recent study demonstrated that another LAG3 intracellular domain, RRFSALE (homologous to FxxL), is necessary and sufficient for LAG3 downstream signaling [42]. However, this was evaluated in reporter-cell lines and not in primary T cells where there are functional differences. A screen identified LAG3-associated protein (LAP) which appears to bind to the EP motif in the intracellular domain of LAG3; however, this interaction and the functional relevance of LAP has yet to be corroborated and remains unknown [47]. These findings highlight the gap in our knowledge of which intracellular domains are necessary for LAG3 inhibitory mechanism.

The extracellular domain of LAG3 can be shed from the surface of T cells, generating soluble LAG3 (sLAG3). A disintegrin and metalloproteinase domain-containing (ADAM) proteins, ADAM10 and ADAM17, cleave LAG3 near the membrane-proximal CP [16,48,49]. A high LAG3: ADAM10 ratio on CD4⁺Foxp3⁻ conventional T cells from the peripheral blood of patients with melanoma and head and neck squamous-cell carcinoma (HNSCC) correlated with poor prognosis [50]. In a noncleavable LAG3 (LAG3^NC) knockin mouse model, CD4⁺FoxP3⁻ T cells exhibited decreased responsiveness to anti-PD1 treatment, leading to reduced tumor clearance in a murine colorectal cancer model. While it appears that sLAG3 has no function and its shedding is simply a mechanism to rapidly turn off LAG3 function [16,49], more studies are needed to confirm the current premise.

Combination ICI therapy using monospecific antibodies in clinical trials

The therapeutic potential of LAG3 is a rapidly growing area of investigation in immuno-oncology, with 22 LAG3 monospecific and bispecific LAG3-targeted antibody drug candidates undergoing clinical testing. Co-expression of LAG3 and PD1 on TILs in clinical samples has been correlated with T cell exhaustion and dysfunction [51,52]. Furthermore, immunohistochemistry studies identified high LAG3 expression on TILs from a variety of tumor types – including ovarian cancer, melanoma, NSCLC, and HNSCC – as correlating with poor prognosis and disease-free survival [53-55]. A retrospective study of peripheral blood from pretreatment of patients with melanoma receiving ICIs identified CD8⁺ LAG3⁺ T cells as a negative prognostic biomarker [56]. In preclinical studies, single-agent anti-LAG3 treatment exhibited minimal antitumor responses. However, in combination with anti-PD1 treatment, significant tumor clearance and prolonged survival was observed in mice [5,57,58]. Together, these findings demonstrated the potential benefits of LAG3-directed therapeutic interventions, paving the way for LAG3 clinical trials.

Relatlimab from Bristol Myers Squibb (BMS) was the first anti-LAG3 mAb checkpoint therapy to be approved by the FDA. This approval resulted from impressive therapeutic benefits observed in the RELATIVITY-047 clinical trial, in which anti-LAG3 (relatlimab) and anti-PD1 (nivolumab) were formulated as a fixed-dose combination (FDC) therapy. The recommended dose for Opdualag is 160 mg of relatlimab plus 480 mg of nivolumab administered intravenously every 4 weeks (Q4W). In patients with previously untreated metastatic or unresectable melanoma, Opdualag more than doubled the median PFS compared with nivolumab monotherapy, 10.1 months [95% confidence interval (CI) 6.4–15.7] versus 4.6 months (95% CI: 3.4–5.6) [hazard ratio (HR) 0.75; 95% CI: 0.62–0.92, P = 0.0055] [10]. Importantly, Opdualag exhibited a comparable safety profile when compared with nivolumab monotherapy, and no new safety events were identified. Recently, the results from the Phase 1/2a RELATIVITY-020 clinical trial reinforce these findings, showing that Opdualag is safe and offers durable clinical activity in patients with heavily pretreated advanced or nonresectable melanoma patients in the first line [59]. By contrast, combination treatment with nivolumab plus the CTLA-4 inhibitor ipilimumab has been associated with substantial toxicity, leading some oncologists to predict that Opdualag will supplant it as the new standard of care for advanced melanoma [60,61].

Despite the FDA approval of Opdualag, the mechanisms by which LAG3 and PD1 synergize have remained elusive. Recently, LAG3 and PD1 synergism was found to abrogate autocrine interferon γ (IFNγ)-dependent antitumor activity of CD8⁺ T cells in mice [28]. In melanoma patients treated with Opdualag, CD8⁺ T cells exhibited higher co-expression of cytotoxic and exhaustion gene modules, compared with patients treated with monotherapy [62]. Consistent with this, LAG3 sustained expression of TOX, a transcription factor that is critical for the development of T cell exhaustion [63-65]. Further, LAG3 expression represses the expression of the activating receptor NKG2C on CD8⁺ T cells, hindering engagement with its ligand Qa-1b on target cells [66]. Disruption of LAG3 improved CD8⁺ T_EX cell functionality and Qa-1b-dependent killing in a murine chronic viral infection model. Together, these studies provide mechanistic insight into the observed efficacy of LAG3/PD1 combinatorial blockade.

While Opdualag therapy showed efficacy in advanced or unresectable melanoma patients in the first line, questions remain about the optimal dosing of anti-LAG3. Some investigators speculate that the efficacy of relatlimab is dampened by the presence of sLAG3 in patients’ serum, suggesting that a higher dosage of anti-LAG3 could enhance efficacy. Fianlimab is an anti-LAG3 mAb developed by Regeneron Pharmaceuticals and Sanofi. It is being administered in patients at a tenfold higher dose (1600 mg), compared with relatlimab (160 mg), in combination with the FDA-approved anti-PD1 mAb cemiplimab (350 mg) Q3W for 12 months. In Phase 1 trials (NCT03005782), fianlimab plus cemiplimab FDC therapy showed an objective response rate (ORR) of 61.2% – 12 complete responders (CRs), 48 partial responders (PRs) – from a total of 98 patients, and a Kaplan–Meier estimation of median PFS was 15.3 (95% CI: 9.4 to not estimated) months [67]. This finding was consistent with preclinical results, in which fianlimab and cemiplimab showed early efficacy and antitumor activity in a human Pdcd1/Lag3-knockin murine model [68]. The safety profile of fianlimab plus cemiplimab FDC therapy was comparable with the safety profile of cemiplimab monotherapy. However, higher incidences of adrenal insufficiency were observed (12% of patients), which were grade ≥2 in 64% of cases. Furthermore, adverse events occurred in 94% of patients, with 44% of patients experiencing severity grade ≥3.

Currently, there are multiple ongoing clinical trials assessing the therapeutic potential of fianlimab plus cemiplimab, including an ongoing Phase 3 trial (NCT05352672) in treatment-naïve advanced melanoma patients. Importantly, a Phase 3 study (NCT06246916) is beginning enrollment to directly compare the ORRs from the FDC of fianlimab plus cemiplimab with the FDC of Opdualag in patients with unresectable or metastatic melanoma. Results from this head-to-head study are likely to have a large impact on the treatment options for cancer patients.

Several other mAbs that target other IRs – such as CTLA-4, PD1, TIM3, and TIGIT, in addition to anti-vascular endothelial growth factor (anti-VEGF) and anti-ICOS mAbs – are being tested in combination with anti-LAG3 (Table 1). Furthermore, several investigators are assessing the benefits of adjuvant and neoadjuvant anti-LAG3/PD1 in combination with chemotherapy and radiation [69]. However, the therapeutic benefit of these various combination strategies is unknown. Contrary to the success in melanoma, BMS recently terminated its Phase 3 RELATIVITY-123 trial of Opdualag in previously treated, microsatellite-stable (MSS) metastatic colorectal cancer (mCRC) patients due to unlikelihood of meeting primary endpoints of OS. It is unclear how these results will impact future testing of LAG3-targeted therapies in mCRC; however, these results highlight our lack of knowledge surrounding LAG3 treatment applications.

Table 1.

Monoclonal antibodies targeting LAG3 in clinical trials for use in treatment of cancers (as of August 2024)

Anti-LAG3 mAbs	Manufacturer	Combination arms	Total number of clinical trials
Relatlimab (BMS-986016)	BMS	Nivolumab (anti-PD1); Ipilimumab (anti-CTLA4); Urelumab (anti-4-1BB); Bevacizumab (anti-VEGF); BMS-986205 (IDO1 inhibitor)	72 (including Opdualag)
Fianlimab (REGN3767)	Regeneron	Cemiplimab (anti-PD1); Degarelix + pTVG-AR	13
Favezelimab (MK-4280)	Merck	Pembrolizumab (anti-PD1)	10
LBL-007	BeiGene	Tislelizumab (anti-PD1); Surzebiclimab (anti-TIM3); chemotherapy	10
Tuparstobart (INCAGN02385)	Incyte	INCAGN02390 (anti-TIM3); Retifanlimab (anti-PD1)	6
HLX26	Shanghai Henlius Biotech	HLX10 (anti-PD1); chemotherapy	4
Ieramilimab (LAG525)	Novartis	PDR001 (anti-PD1); carboplatin	4
Miptenalimab (BI-754111)	Boehringer Ingelheim	BI907828 (MDM2-p53 inhibitor); Ezabenlimab (anti-PD1)	4
Negalstobart (IBI110)	Innovent Biologics	Sintilimab (anti-PD1); etoposide, platinum	4
Sym022	Symphogen	Sym021 (anti-PD1); Sym023 (anti-TIM3)	3
Encelimab (TSR-033)	Tesaro/GSK	Dostalimab (anti-PD1); TSR-022 (anti-TIM3); Bevacizumab (anti-VEGF); FOLFIRI; mFOLFOX6	2
GLS-012	Gloria Pharma	GLS-010 (anti-PD1); paclitaxel + carboplatin; pemetrexed + carboplatin	2
TQB2223	Chia Tai Tianqing Pharmaceutical Group	Penpulimab (anti-PD1)	2

Bispecific antibodies targeting multiple IRs on immune cells

Due to the success of the Opdualag, interest in LAG3 therapeutics has increased and the clinical development of multiple therapeutics accelerated. The utility of bispecific antibodies (BsAbs) has increased due to the possibility that they might selectively target T_EX cells with eight BsAbs directed at LAG3 currently in trials (Table 2). BsAbs can target double-positive cells (e.g., PD1⁺LAG3⁺ T cells) and thus may exhibit increased potency and reduced toxicity due to the dampening of off-target effects. Also, it has been suggested that BsAbs sterically exclude IRs from the immune synapse to prevent inhibitory function [70]. The majority of LAG3-targeting BsAbs are assessing the benefits of co-targeting other IRs, such as PD1 or CTLA4, expressed on TILs.

Table 2.

Bispecific antibodies targeting LAG3 in clinical trials for use in treatment of cancers (as of August 2024)

Bispecific antibody	Manufacturer	Target	Combination arms	Total number of clinical trials
Tobemstomig (RO7247669)	Roche	LAG3 x PD1	Tiragolumab (anti-TIGIT); axitinib (tyrosine kinase inhibitor); bevacizumab; paclitaxel; pemetrexed; carboplatin	8
IBI323	Innovent Biologics	LAG3 x PDL1	Bevacizumab; platinum chemotherapy	2
Pavunalimab/Bavunalimab (XmAB22841)	Xencor	LAG3 x CTLA4	Pembrolizumab; XmAb23104 (anti-PD1xICOS)	2
Tebotelimab (MGD013)	Macrogenics	LAG3 x PD1	Margetuximab (anti-HER2)	2
ABL501	Abl bio	LAG3 x PD1	Monotherapy	1
AK129	Akeso	LAG3 x PD1	Monotherapy	1
FS118	invoX	LAG3 x PDL1	Paclitaxel	1
INCA32459	Incyte	LAG3 x PD1	Monotherapy	1

Developed by MacroGenics, tebotelimab (MGD013) is a tetravalent anti-LAG3/PD1 BsAb. Tebotelimab blocks the interaction of PD1 to PDL1 and LAG3 to its respective ligands, similarly to relatlimab plus nivolumab [71]. The Phase 1 trial (NCT03219268) of tebotelimab assessed its safety profile as a monotherapy in advanced solid tumors. Tebotelimab demonstrated irAEs in 68.4% of patients (184/269), with 22.3% (60/269) grade ≥3. Additionally, tebotelimab in combination with margetuximab, an anti-HER2 mAb, was assessed in this trial. HER2 is a growth factor expressed on tumor cells. Interestingly, 16.7% of patients (14/84) exhibited irAEs of grade ≥3 with an ORR of 19.4% and a median duration of response of 16.7 months (95% CI: 11.04–NE). Of 72 patients, there were one CR (1.4%), 13 PR (18.1%), 25 stable disease (SD) (34.7%), and 33 (45.8%) progressive disease (PD). This success resulted in the Phase 2/3 Mahogany trial (NCT04082364) with an additional arm assessing tebotelimab plus margetuximab with chemotherapy in patients with HER2-positive gastric cancer/gastroesophageal junction cancer.

Another anti-LAG3/PD1 BsAb is tobemstomig (RO7247669/RG6139, developed by Roche), which is designed to have higher affinity for PD1 than for LAG3, to preferentially bind TILs. Tobemstomig is currently being evaluated in eight clinical trials. Several of these are assessing tobemstomig as a monotherapy. In a Phase 1 trial (NCT04140500) in patients with advanced or metastatic solid tumors an ORR of 17.1% was observed, with 6/35 patients experiencing grade 3 irAEs. Additionally, tobemstomig is being tested in combination with platinum-based chemotherapy (NCT05775289), anti-TIGIT mAb (tiragolumab), and/or the tyrosine kinase inhibitor axitinib (NCT05805501).

CTLA4 is an IR expressed constitutively on Tregs and upregulated upon activation on T cells. It competitively binds to CD80 and CD86 with greater affinity than CD28 to attenuate T cell activation. Anti-CTLA4 mAbs have improved patient survival in combination with anti-PD1 mAbs; however, severe irAEs have been reported, which frequently led to treatment discontinuation, thereby hindering its efficacy. This has prompted the need to target other IRs and evaluate different dosages. Xmab22841 is a BsAb targeting CTLA4/LAG3 from Xencor. NCT05695898 is the first trial to target four receptors: LAG3, CTLA4, PD1, and ICOS using two BsAbs, XmAb22841 (CTLA4 x LAG3) and Xmab23104 (PD1 x ICOS), in patients with metastatic melanoma. Importantly, these two BsAbs exhibited low occurrence of irAEs (20–24%) in their separate Phase 1 trials (NCT03752398 and NCT03849469). Together, the early clinical results from LAG3-targeting BsAbs highlight their enhanced safety profiles, compared with combination therapies with monospecific mAbs, potentially due to their increased specificity; however, their impact on antitumor immunity is unclear.

Bispecific antibodies that promote cell-to-cell interactions

The capability of BsAbs to bridge interactions of T cells with antigen-presenting cells (APCs) or tumor cells is being evaluated in clinical trials. One strategy aims to utilize the dual-targeting nature of BsAbs to selectively stimulate TILs by targeting a tumor-expressing ligand in addition to an immune cell IR.

FS118, a tetravalent BsAb targeting LAG3/PDL1 developed by F-star Therapeutics, was one of the earliest BsAbs targeting LAG3 to enter trials, having demonstrated efficacy in preclinical cancer models [72]. FS118 recently completed its Phase 1 trial in patients with advanced cancer and PDL1 resistance, with no dose-limiting toxicities observed, and a disease control rate of 54.8% in patients receiving ≥1 mg/kg of FS118. Currently, FS118 is in Phase 2/3 trials assessing safety, tolerability, and efficacy as a monotherapy or in combination with paclitaxel chemotherapy in advanced malignancies.

ABL501 (developed by ABL Bio) is a tetravalent IgG4 BsAb targeting LAG3 and PDL1. In preclinical studies, ABL501 promoted dendritic cell (DC) maturation, indicated by upregulation of CD40 and HLA-ABC, and increased the accumulation of CD8⁺ TILs in a humanized murine cancer model [73]. ABL501 had minimal toxicity in preclinical assessments, with 200 mg/kg/dose twice weekly for eight doses tolerated in cynomolgus monkeys. It is currently being tested in a Phase 1 study as a monotherapy in patients with progressive or metastatic solid tumors (NCT05101109).

IBI323 (from Innovent Biologics) targets LAG3/PDL1 as single-domain tetravalent BsAb [74]. In preclinical testing, mice receiving IBI323 exhibited significantly reduced tumor growth in a human PD-L1 knock-in MC38 colon adenocarcinoma and an A375 human melanoma model. In an allogeneic mixed lymphocyte reaction (MLR) of CD4⁺ T cells with DCs, IBI323 increased IFNγ and interleukin-2 (IL-2) secretion. This effect may be mediated by the crosslinking of immune cells with either tumor cells or APCs. Increased cell-to-cell interactions were observed between LAG3⁺Jurkat cells and PDL1-expressing MDA-MB-231 human adenocarcinoma cells. Currently, there are two clinical trials evaluating IB323: a Phase 1 monotherapy in patients with advanced malignancies (NCT04916119), and a Phase 2 investigating a combination of bevacizumab (anti-VEGF-A mAb) plus platinum-based chemotherapy in NSCLC (NCT05296278).

A second strategy of utilizing BsAbs to bridge cell-to-cell interactions is to target DC receptors to enhance cross-priming [75]. Immunoglobulin Fc domains bind to Fcγ receptors (FcγRs), inducing antibody-dependent cell-mediated cytotoxicity (ADCC) of the antibody-coated cell. Therapeutic antibodies can be mutated in their Fc region – such as L234A/L235A (LALA) – to reduce this interaction [76]. Currently, an Fc-competent trispecific antibody GB266T (anti-PDL1/TIGIT/LAG3) is in preclinical development [77]. This design creates bridges of cell-to-cell interactions between APCs, cancer cells, and T cells. In preclinical testing GB266T treatment increased human CD4⁺ and CD8⁺ T cell viability and promoted antitumor activity in humanized NOD scid gamma (NSG) mice implanted with an RKO human colon cell line, compared with the LALA-mutated antibody. These findings suggest retaining Fc function with trispecific antibodies as novel engagers of immune cells to eliminate tumor cells. Together, these results highlight the potential therapeutic benefits of BsAb strategies targeting APCs or tumor-adjacent T cells to enhance CD8⁺-mediated antitumor immunity.

Biomarkers for LAG3-based therapy

With the emergence of LAG3 ICIs, there is a critical need for biomarkers to identify patient populations that would benefit from these therapies. Several studies have correlated high LAG3 expression on TILs from a variety of tumor types with poor prognosis. This observation demonstrates that LAG3 could be a mechanism of resistance to PD1 blockade and indicates that LAG3 could be used as valuable biomarker for anti-PD1 therapies. Furthermore, LAG3 could be combinatorically targeted in patients with high levels of LAG3⁺ TILs to overcome such resistance. Analysis of the RELATIVITY-047 trial revealed a trend toward extended PFS in patients with ≥1% LAG3 expression in the tumor microenvironment (TME); however, this trend was additionally observed in the nivolumab monotherapy arm, indicating that LAG3 expression alone could not predict which patients would benefit from the addition of anti-LAG3 therapy [10]. A recent study evaluating baseline LAG3 expression levels and clinical outcomes following relatlimab and nivolumab combination treatment in patients with metastatic melanoma corroborated these findings, showing that patients with≥1% LAG3⁺ cells in the TME had significantly longer PFS and higher rates of response [78].

The LAG3 ligand, FGL1, could potentially be a biomarker, with high levels of FGL1 in plasma being correlated with anti-PD1 resistance in patients with melanoma and NSCLC [37]. Additionally, while the function of sLAG3 is unclear, sLAG3 in patient sera is elevated in various malignancies and can be prognostic of outcome. Varying by cancer type, high levels of sLAG3 in the serum have been associated with a better or worse prognosis [79,80]. However, there is currently no information on how sLAG3 or FGL1 levels correlate with response in patients treated with anti-LAG3 therapies. LAG3’s ligand MHCII is also expressed on several cancers. HLA-DR expression on melanoma is correlated with a transcriptional signature that is inflammatory and increased CD4⁺ T cell infiltration. Melanomas expressing MHCII also correlated with better PFS in patients treated with anti-PD1/PDL1 therapies [81,82]. Further investigations are required to discern which patients will benefit from anti-LAG3-based combination therapy.

Concluding remarks and future perspectives

The FDA approval of Opdualag has spiked interest in LAG3 as a target for cancer therapy. Currently, investigators are testing different strategies in a variety of cancer types to find new applications of LAG3-directed therapy. These include assessments of the impact of anti-LAG3 mAbs in combination with other ICIs, BsAbs selective for T_EX, and BsAbs directed at promoting T cell interactions with APCs and tumor cells.

While all LAG3-directed clinical trials are assessing the impact of LAG3 blockade in a combinatorial setting, questions still remain about whether anti-LAG3 monotherapy alone could be efficacious, especially if given at higher doses. Regeneron’s Phase 1 trial testing of anti-LAG3 (fianlimab) monotherapy showed no dose-limiting toxicities, reporting grade ≥3 irAEs in 7.4% patients (NCT03005782). Currently, the Phase 3 clinical trial is testing the combination therapy of anti-PD1 (cemiplimab, 350 mg) with a tenfold higher dose of anti-LAG3 (fianlimab; 1600 mg) than Opdualag, which administers 160 mg of relatlimab. While the efficacy of this higher dose of anti-LAG3 is not being assessed as a monotherapy, this increased dose of anti-LAG3 showed a 63% ORR in advanced PDL1-naïve metastatic melanoma patients with an acceptable risk–benefit profile. In this cohort, 19 patients (47%) had M1c metastases and 17 (42.5%) had lactate dehydrogenase (LDH) levels greater than the upper limit of normal (ULN). Together, these results highlight that anti-LAG3 therapy is a highly tolerable treatment option and suggests that a higher dose of anti-LAG3 could be an effective monotherapy with sufficient dosing, contrary to the current understanding of LAG3-directed therapies in the immuno-oncology field.

Epitope mapping of LAG3 revealed the binding domains of several human anti-LAG3 therapeutic mAbs [83]. These mAbs were found to target the D1 domain of LAG3, which prevents LAG3–MHCII interactions [83,84]. Recent findings on the structure of LAG3 demonstrated mAb targeting the murine LAG3 at D1 (M8-4-6), D2 (C9B7W), and D3 (410C9) domains can disrupt LAG3 homodimerization, thereby preventing ligand binding and its inhibitory function [25,27]. While relatlimab (BMS), miptenalimab (Boehringer Ingelheim), favezelimab (Merck), and likely other mAbs bind to or near the LAG3 D1 loop to inhibit MHCII binding, it remains unclear whether other epitopes would be more effective.

Lastly, despite the known clinical benefits of targeting PD1 and LAG3, there is still surprisingly little known about how PD1 and LAG3 synergize to limit T cell function and antitumor immunity, which can be unleashed with combinatorial blockade. The contribution of CD4⁺ T cells to the observed efficacy of anti-PD1 and anti-LAG3 combination therapy is unknown. Recently, PD1 and LAG3 were shown to synergize to alter the exhaustion profiles of CD8⁺ T cells, and that the blockade of PD1 and LAG3 enhances their antitumor immunity driven by an autocrine IFNγ-dependent mechanism [28]. Furthermore, PD1- and LAG3-deficient CD8⁺ T cells have significantly lowered expression of TOX accompanied with enhanced secretion of IFNγ and granzyme B, and increased cell proliferation and TCR diversity [62,66]. Further studies could reveal additional underappreciated aspects of the combinatorial activity of PD1 and LAG3, which could be utilized in future clinical development (see Outstanding questions).

Outstanding questions.

Are the current LAG3 therapeutics optimal? Anti-LAG3 combination therapy has shown impressive results; it is unclear whether better responses can be achieved with new therapeutic strategies. This is, in part, because the mechanism of anti-LAG3/PD1 synergy is not fully understood.

Which LAG3 immunotherapeutic strategies will be the most efficacious? LAG3 blockade is being combined with many other IRs and surface receptors; however, it is unknown which of these targets will result in durable responses in patients.

Will bispecific antibodies prove to be more potent and continue to result in low toxicity, compared with combinatorial monotherapeutic strategies? Which bispecific platform should be used in developing dual-targeting therapies?

Can radiation and/or chemotherapy improve anti-LAG3/PD1 therapy? These have been shown to induce neoantigen expression on tumor cells, which are highly immunogenic. Several clinical trials are assessing anti-LAG3 therapies with platinum doublet chemotherapy and other standard chemotherapies.

Highlights.

Lymphocyte activation gene 3 (LAG3) D2-domain-directed monoclonal antibodies exhibit enhanced antitumor immunity in preclinical models.

Higher dosing of anti-LAG3 in combination with anti-programmed death protein 1 (anti-PD1) shows clinical efficacy with a favorable safety profile in patients with advanced melanoma.

Bispecific antibody strategies targeting T cells and tumor or antigen-presenting cells are in preclinical development.

Bispecific antibodies targeting LAG3 and PD1 or CTLA4 have demonstrated safety in early-phase clinical trials.