LAG‐3 : recent developments in combinational therapies in cancer

Abstract

The study of anticancer immune responses and in particular the action of immune checkpoint inhibitors that overcome T cell inhibition has revolutionized metastatic patients' care. Unfortunately, many patients are resistant to these innovative immunotherapies. Over the last decade, several immune checkpoint inhibitors, currently available in the clinic, have been developed, such as anti‐PD‐1/PD‐L1 or anti‐CTLA‐4. More recently, other immune checkpoints have been characterized, among them lymphocyte activation gene 3 (LAG‐3). LAG‐3 has been the subject of numerous therapeutic studies and may be involved in cancer‐associated immune resistance phenomena. This review summarizes the latest knowledge on LAG‐3 as an immunotherapeutic target, particularly in combination with standard or innovative therapies. Indeed, many studies are looking at combining LAG‐3 inhibitors with chemotherapeutic, immunotherapeutic, radiotherapeutic treatments, or adoptive cell therapies to potentiate their antitumor effects and/or to overcome patients' resistance. We will particularly focus on the association therapies that are currently in phase III clinical trials and innovative combinations in preclinical phase. These new discoveries highlight the possibility of developing other types of therapeutic combinations currently unavailable in the clinic, which could broaden the therapeutic spectrum of personalized medicine.

This review summarizes the latest knowledge on LAG‐3 as an immunotherapeutic target, particularly in combination with standard or innovative therapies.

Abbreviations

BCMA

B cell maturation antigen

CTLA‐4

cytotoxic T lymphocyte antigen 4

FGL‐1

fibrinogen‐like protein 1

galectin‐3

galactose lectin‐3

Gsk‐3

glycogen synthase kinase 3

HDAC

histone deacetylase

ICI

immune checkpoint inhibitor

LAG‐3

lymphocyte activation gene 3

LSECTin

liver and lymph node sinusoidal endothelial cell C‐type lectin

MDSC

myeloid‐derived suppressor cells

NK

natural killer

PBMC

peripheral blood mononuclear cell

PD‐1

programmed death 1

PI3Kδ

phosphoinositide‐3‐kinase δ

TIGIT

T cell immunoglobulin and ITIM domain

TIL

tumor‐infiltrating lymphocyte

Tim‐3

T cell immunoglobulin 3

VDC

virus‐like drug conjugate

1. INTRODUCTION

Cancer immunotherapy has shaped the therapeutic landscape of many metastatic patients over the past decade, notably with the discovery of the immune checkpoints, such as programmed death 1 (PD‐1) ¹ and cytotoxic T lymphocyte antigen 4 (CTLA‐4). ² Immune checkpoints are activating or inhibitory receptors on the surface of immune cells and contribute to the regulation of immune cell activation. ³ , ⁴ , ⁵ PD‐1 acts in the later phases of lymphocyte activation mainly in peripheral tissues and the tumor microenvironment, while CTLA‐4 acts in the early phases of T cell activation in lymphoid organs. They are expressed on T cells and interact with their ligands, PD‐L1/PD‐L2 (present notably on the surface of cancer cells) and B7.1 (present on the surface of antigen‐presenting cells), respectively. ⁶ , ⁷ In the tumor microenvironment, T cells are subjected to a chronic stimulation with cancer cells antigens, inducing expression of immune checkpoints. The interaction of these receptors with their ligands expressed by tumor and antigen‐presenting cells, inhibits T cell function and therefore limits the antitumor immune response. Although the use of immune checkpoint inhibitors (ICIs) targeting CTLA‐4 and PD‐1 in various types of cancer has been beneficial in the clinic, unfortunately, many patients do not respond to these promising immunotherapies. ⁸ , ⁹ , ¹⁰ Resistance to these ICIs may be explained by the expression of other immune checkpoints in T cells such as lymphocyte activation gene 3 (LAG‐3, also known as CD223). ¹¹ , ¹² LAG‐3 is highly expressed in cancer patients, ¹³ , ¹⁴ , ¹⁵ often coexpressed with PD‐1 ¹⁶ , ¹⁷ and associated with poor prognosis. ¹⁴ , ¹⁸ , ¹⁹ LAG‐3 is found within the cell, but also an extracellular form exists, following cleavage of the extracellular part by the metalloprotease ADAM 10/17 ²⁰ (Figure 1). Both forms have been reported to be increased in cancer. ²¹ , ²² LAG‐3 is also expressed in immune cells, including CD4⁺ and CD8⁺ tumor‐infiltrating lymphocytes (TILs), ¹⁸ , ²³ natural killers (NKs), invariant natural killer T (NKT) cells, activated B cells, and plasmacytoid dendritic cells. ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ Of note, its expression on T CD8⁺ cytotoxic immune cells prevents them from proliferating, ¹⁸ which impedes a proper antitumor immune response and thus may limit the efficacy of treatments. Structurally, LAG3 is composed of an extracellular, a transmembrane, and an intracellular region. The extracellular part, where the ligands bind, comprises four domains belonging to the immunoglobulin superfamily (D1–D4) whose D1 domain has a loop domain rich in proline, and glycosylation sites. The transmembrane part consists of a connecting peptide, and the intracellular part comprises a serine phosphorylation site and two motifs named KIEELE and EP ²⁹ , ³⁰ involved in LAG‐3 function (Figure 1). LAG‐3 binds to several ligands, the best characterized being MHC class II present on the surface of antigen‐presenting cells. LAG‐3 is unique in having amino acid sequence homology with the CD4 receptor, ²⁴ , ²⁸ and it binds to MHC class II with higher affinity than CD4. ²⁵ Interaction of MHC class II with LAG‐3 suppresses T cell activation. ³¹ LAG‐3 can bind to other ligands such as liver and lymph node sinusoidal endothelial cell C‐type lectin (LSECTin), ³² fibrinogen‐like protein 1 (FGL1), ³³ and galactose lectin‐3 (galectin‐3), ³⁴ which can be present on different cell types such as dendritic cells, monocytes, stromal cells, liver cells, and cancer cells ³² , ³³ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ (Figure 1). The interaction between LAG‐3 and these ligands also induces the inhibition of T cells. ³² , ³³ , ³⁴ Concerning its way of action, LAG‐3 inhibits T cells’ anticancer response by reducing T cells’ cytokine and granzyme production and proliferation while encouraging differentiation into T regulatory cells, thereby providing a tumor microenvironment favorable to tumor escape. In certain cancers such as renal cancer, LAG‐3, compared with the immune checkpoints Tim‐3 or TIGIT that have also recently come to light, might be a better target, as its expression in tumors is associated with the worst overall survival and recurrence‐free survival rates. ⁴¹ Although targeting LAG‐3 alone could be a possible therapeutic strategy, it is in combinational therapies that it has proven its clinical interest to better enhance T cell activation, thereby improving the efficacy of various treatments. In this review, we will give an overview of combination therapies involving targeting LAG‐3, highlighting those in phase III studies.

FIGURE 1.

Structure of lymphocyte activation gene 3 (LAG‐3) and its interaction with ligands. LAG‐3 is present on the surface of T cells. It is composed of an extracellular, transmembrane, and intracellular region. The extracellular part comprises four domains belonging to the immunoglobulin superfamily (D1–D4), whose D1 domain has a loop domain rich in proline, and glycosylation sites. The transmembrane part consists of a connecting peptide. The intracellular part comprises a serine phosphorylation site, a KIEELE motif, and an EP motif. ²⁹ , ³⁰ A soluble form of LAG‐3 exists, which appears as a result of cleavage of the extracellular part by ADAM 10/17. LAG‐3 can interact with several ligands such as HMC II, FGL‐1, LSECTin, or galectin‐3, which are present in several cell types including cancer cells, liver cells, monocytes, stromal cells, and dendritic cells. Binding of LAG‐3 to these ligands induces inhibition of T cells.

2. LAG‐3 COMBINATIONAL THERAPIES IN ADVANCED CLINICAL EVALUATION

2.1. Molecules targeting LAG‐3 in clinical trials

Several neutralizing antibodies targeting LAG‐3 have been developed as potential immunotherapeutic drugs to be used in various types of cancer. ⁴² They bind to human LAG‐3 with high affinity and specificity to block the interaction of LAG‐3 with its ligands, notably MHC II, and reverse LAG‐3‐mediated inhibition of T cell function in vitro. ⁴³ Among them, three monoclonal antibodies, in combinational therapies, are currently in phase III clinical trials, which is the most advanced phase at present for molecules targeting LAG‐3. Those are: relatlimab (also called BMS‐986016), ⁴⁴ fianlimab (also called REGN3767), ⁴⁵ and favezelimab (also called MK‐4280) ⁴⁶ (Table 1). A bispecific antibody targeting LAG‐3 and PD‐1 is also in phase III clinical trials: tebotelimab (also called MGD013) ⁴⁷ (Table 1). These phase III studies are described in Section 2.2.

TABLE 1.

Molecules targeting lymphocyte activation gene 3 (LAG‐3) undergoing clinical evaluation.

Name of the drug	Type	Isotype	Laboratory	References
Relatlimab	Monoclonal antibody	IgG4κ	Bristol‐Myers Squibb	44
Fianlimab	Monoclonal antibody	IgG4	Regeneron pharmaceuticals	45
Favezelimab	Monoclonal antibody	IgG4	Merck	46
Ieramilimab	Monoclonal antibody	IgG4	Novartis	49
INCAGN02385	Monoclonal antibody	IgG1κ	Incyte corporation	50
Sym022	Monoclonal antibody	IgG	Symphogen	51
GSK2831781	Monoclonal antibody	IgG1	GlaxoSmith	52
TSR‐033	Monoclonal antibody	IgG4	Tesaro	53
Tebotelimab	Bispecific antibody (targeting LAG‐3 and PD‐1)	IgG4κ	MacroGenicx	47
FS118	Bispecific antibody (targeting LAG‐3 and PD‐L1)	IgG1	F‐star therapeutics	48
RO7247669	Bispecific antibody (targeting LAG‐3 and PD‐1)	Undisclosed	Hoffmann‐La Roche	54
EMB‐02	Bispecific antibody (targeting LAG‐3 and PD‐1)	Undisclosed	EpimAb Biotherapeutics	55
Pavunalimab	Bispecific antibody (targeting LAG‐3 and CTLA‐4)	Undisclosed	Xencor	56

Note: The antibody type, isotype, and supplier are listed.

Three other antibodies will soon enter phase III clinical trials (Table 1). Ab FS118 is a tetravalent bispecific antibody targeting LAG‐3 and PD‐1. A recent phase I first‐in‐human study (NCT03440437) has demonstrated that it was well tolerated in patients with advanced solid cancers resistant to anti‐PD(L)1 therapy ⁴⁸ (Table 1). The second one is Ieramilimab (LAG525), a humanized anti‐LAG‐3 monoclonal antibody, well tolerated in patients alone or in combination with the anti‐PD‐1 antibody spartalizumab. A recent phase II multicenter study (NCT02460224) including patients with different locally advanced or metastatic solid malignancies demonstrated that response to the combined immunotherapy was durable (>24 months) in some patients across all enrolled indications. ⁴⁹ The third is INCAGN02385, another anti‐LAG‐3 monoclonal antibody, soon to enter a phase III clinical trial, which proved to be well tolerated in a phase I study on different types of cancer such as lung, gastric, breast, ovarian, endometrial cancer, or even melanoma (NCT03538028). ⁵⁰ A phase II clinical trial evaluating the combination of INCAGN02385 with retifanlimab (anti‐PD‐1) and INCAGN02390 (anti‐Tim‐3) in squamous cell carcinoma of the head and neck will end in 2024 (NCT05287113). Similarly, a phase II clinical trial testing INCAGN02385 in urothelial cancer will come to an end in 2024 (NCT04586244) (Table 1).

Finally, although still in early clinical trials, it is worth mentioning several other mono‐ or bispecific antibodies targeting LAG‐3, such as the monoclonal antibodies Sym022 (LAG‐3‐inhibiting antibody), ⁵¹ GSK2831781 (LAG‐3‐depleting antibody), ⁵² and TSR‐033 (LAG‐3‐inhibiting antibody) ⁵³ or bispecific antibodies RO7247669 (PD‐1/LAG‐3‐targeting antibody), ⁵⁴ EMB‐02 (PD‐1/LAG‐3‐targeting antibody) ⁵⁵ and pavunalimab (XmAb22841, LAG‐3/CTLA‐4‐targeting antibody) ⁵⁶ (Table 1).

2.2. Combination treatments involving LAG‐3 inhibitors in phase III clinical trials

Relatlimab was the first anti‐LAG‐3 antibody approved by the FDA in a phase III clinical trial (RELATIVITY‐047) evaluating a dual anti‐PD‐1 (nivolumab) and anti‐LAG‐3 (relatlimab) therapy in patients with advanced melanoma (NCT03470922) (Table 2). Previous phase I/II studies demonstrated good efficacy and safety for the relatlimab–nivolumab association in melanoma (NCT01968109), with a disease control rate of 45%. ⁴⁴ The phase III clinical trial showed a higher progression‐‐free survival for melanoma patients treated with the double therapy (10.1 months) compared with patients treated with nivolumab alone (4.6 months), ⁵⁷ which implied that blocking LAG‐3 improved the efficacy of anti‐PD‐1 therapy. This promising clinical result highlights the important role of LAG‐3 as a therapeutic target that potentiates the effects of other immunotherapies. In this phase III trial, the percentage of the patients presenting grade 3 or 4 side effects was 18.9% in the population treated with the dual therapy and 9.7% for the patients treated with nivolumab alone. ⁵⁷ Four other phase III clinical trials evaluating the combination of relatlimab with nivolumab are underway in colon cancer, melanoma, and urothelial bladder cancer (NCT05328908, NCT05625399, NCT05002569, NCT05987241) (Table 2).

TABLE 2.

Phase III clinical trials involving lymphocyte activation gene 3 (LAG‐3) with other molecules.

Combinational therapy	Type of cancer	NCT number	Trial starting date
Relatlimab/nivolumab	Untreated metastatic or unresectable melanoma	NCT03470922	2018‐04
Relatlimab/nivolumab	Metastatic colorectal cancer	NCT05328908	2022‐04
Relatlimab/nivolumab	Metastatic or unresectable melanoma	NCT05625399	2023‐03
Relatlimab/nivolumab	Melanoma	NCT05002569	2021‐10
Relatlimab/nivolumab	Bladder urothelial carcinoma	NCT05987241	2024‐01
Fianlimab/cemiplimab/pemetrexed + cisplatin or paclitaxel + carboplatin	Advanced non‐small cell lung cancer (NSCLC)	NCT05800015	2023‐08
Fianlimab/cemiplimab	Melanoma	NCT05608291	2023‐01
Fianlimab/cemiplimab	Locally advanced or metastatic melanoma	NCT05352672	2022‐07
Fianlimab/cemiplimab	NSCLC	NCT05785767	2023‐06
Favezelimab/pembrolizumab	Colorectal cancer	NCT05600309	2022‐06
Favezelimab/pembrolizumab	Colorectal cancer	NCT05064059	2021‐11
Favezelimab/pembrolizumab	Hodgkin lymphoma	NCT05508867	2022‐10
Tebotelimab/enoblituzumab	Squamous cell carcinoma of the head and neck	NCT04129320	2019‐10
Tebotelimab/margetuximab/capecitabin + oxaliplatin or 5‐Fluorouracil + leucovorin + oxaliplatin	Metastatic or locally advanced HER2‐positive gastric or gastroesophageal junction cancer	NCT04082364	2019‐09
IMP321/paclitaxel	Metastatic breast cancer	NCT05747794	2023‐05

Note: Indicated are the immunotherapies and chemotherapies used. Relatlimab, fianlimab, favezelimab, and tebotelimab are anti‐LAG‐3; nivolumab, cemiplimab, and pembrolizumab are anti‐PD‐1; enoblituzumab is anti‐B7‐H3; IMP321 (eftilagimod alpha) is a dimeric soluble form of LAG‐3.

There are also three other phase II/III trials assessing the dual LAG‐3 + PD‐1 therapy, this time using fianlimab as a LAG‐3 inhibitor and cemiplimab (REGN2810) as a PD‐1 inhibitor, in two types of cancer: melanoma and non‐small cell lung cancer (NSCLC) (NCT05608291, NCT05352672, NCT05785767) (Table 2). Former phase I clinical trials on this dual therapy demonstrated an interesting safety profile (NCT03005782, NCT03233139), ⁵⁸ , ⁵⁹ , ⁶⁰ , ⁶¹ which further supported the results obtained with nivolumab and relatlimab, confirming that dual targeting of PD‐1 and LAG‐3 appears to be a very promising avenue in cancer immunotherapy.

The dual immunotherapy targeting PD‐1/LAG‐3 is also being tested associated with chemotherapies in two different phase III clinical trials. The first one is evaluating fianlimab + cemiplimab combined with pemetrexed and cisplatin or paclitaxel and carboplatin in advanced NSCLC (NCT05800015) (Table 2). The second ongoing phase III trial aims to evaluate the bispecific PD‐1 and LAG‐3 antibody (tebotelimab) in combination with capecitabin + oxaliplatin or 5‐fluorouracil + leucovorin + oxaliplatin (with margetuximab, anti‐HER2) in metastatic or locally advanced HER2‐positive gastric or gastroesophageal junction cancer (NCT04082364) ⁶² (Table 2). The bispecific tebotelimab is also being tested in association with another immunotherapeutic antibody called enoblituzumab (an anti‐B7 homolog 3 protein, B7‐H3), compared with the combination of MGA012 (an anti‐PD‐1, also called retifanlimab) with enoblituzumab (Table 2). B7‐H3 (also known as CD276) has been described as a costimulatory molecule for T lymphocytes, promoting epithelial–mesenchymal transition and associated with poor prognosis and tumor progression. ⁶³ , ⁶⁴ , ⁶⁵ A recent phase I clinical trial (NCT03219268) has demonstrated a tolerable safety and an interesting objective response rate of tebotelimab in patients with different cancer types including nonresponders to PD‐1 blockade (i.e., ovarian cancer and diffuse large B cell lymphoma), ⁴⁷ which is encouraging for ongoing and future phase III clinical trials.

Finally, three phase III clinical trials are currently underway, testing the combination of favezelimab and the anti‐PD‐1 pembrolizumab in colon cancer and Hodgkin lymphoma (NCT05600309, NCT05064059, NCT05508867) (Table 2).

Concerning undesirable effects, LAG‐3 inhibitors alone or in combination with other ICIs are generally well tolerated with side effects such as pruritus, fatigue, and nausea in various cancers. The immune‐related side effects described are mainly hypothyroidism, thyroiditis, rash, diarrhea or colitis, hyperthyroidism, hepatitis, adrenal insufficiency, pneumonitis, hypophysitis, nephritis, and renal dysfunction and/or hypersensitivity. ⁴⁶ , ⁴⁷ , ⁵⁷ , ⁶⁶ However, they did not warrant the arrest of the clinical assays. This highlights the need of biomarkers of toxicity or of response to LAG‐3‐based therapies to select the patients that would benefit from these innovating immunotherapies. Although no toxicity biomarkers have yet been validated, LAG‐3 and PD‐L1 levels have been studied as potential biomarkers of answer to the therapy. Indeed, LAG‐3 and PD‐L1 baseline expression in tumors has been demonstrated to be associated with a better response in various cancers (melanoma, epithelial ovarian cancer, triple‐negative breast cancer, NSCLC), as well as with an improved progression‐free survival in metastatic melanoma. ⁴⁷ , ⁶⁷

It should be noted that LAG‐3 is used and studied not only as a therapeutic target but also as a molecule of interest for the activation of immune cells presenting MHC II since MHC II is a major ligand for LAG‐3. Accordingly, eftilagimod alpha (IMP321), a recombinant dimeric soluble form of LAG‐3 that has been developed to activate dendritic cells via MHC II, has been approved in a phase I study in advanced renal cell carcinoma for its activity and safety. ⁶⁸ A phase III clinical trial (NCT05747794) is currently evaluating IMP321 in combination with the chemotherapeutic drug taxol (paclitaxel) in metastatic breast cancer (Table 2).

Altogether, these clinical studies allow us to conclude that LAG‐3 holds promises in cancer therapy. Although clinical trials are providing major answers on the tolerance/efficacy of treatment combinations involving LAG‐3 targeting, numerous preclinical studies have also been carried out, providing further insight into how they work at the immune level.

3. PRECLINICAL AND TRANSLATIONAL STUDIES INVOLVING LAG‐3 IN COMBINATIONAL THERAPIES

3.1. Immunological response to anti‐LAG‐3 and anti‐PD‐1/PD‐L1 association

Bridging the gap between preclinical and clinical studies, translational research experiments using melanoma patients' samples have demonstrated that anti‐PD‐1/LAG‐3 therapy increases the production of IL‐12 and IL‐18, which promotes IFNγ production and T cell costimulatory molecules such as CD27 and OX40, thereby promoting T cell activation. ⁶⁹ It has also been reported that this dual therapy increases leukocyte chemoattraction via molecules such as CXCL9/10/11/12 and CCL3/20, ⁶⁹ which could facilitate immune responses. It should be noted that this dual therapy also increases IL‐10 production, which is an immunosuppressive cytokine whose role here remains elusive. ⁶⁹ Dual inhibition of LAG‐3 and PD‐1 has an impact on regulatory T lymphocytes as well, by promoting the expansion of these cells but reducing their suppressive effect. ⁶⁹ In multiple myeloma patients, dual inhibition of LAG‐3 and PD‐1 also has an impact on regulatory T lymphocytes, by reducing their proliferation in bone marrow mononuclear cells. ⁷⁰ Analysis of tumor biopsies from patients with different solid cancers has shown that response to a bispecific anti‐PD‐1/LAG‐3 antibody can be associated with an increase in the activation of a CD8⁺ T subtype called HLA‐DR CD25⁺ granzyme B⁺ and in the proliferation of NK cells. ⁷¹ The dual anti‐PD‐1/LAG‐3 therapy also promotes NK cells degranulation and production of IFNγ and TNFα. ⁷¹ Knowing that NK cells are immune cells described as having an antitumor activity, ⁷² , ⁷³ in particular by releasing cytotoxic granules, ⁷³  we can therefore assume that the efficacy of the dual therapy also is mediated by NK cells.

In preclinical models, blocking both LAG‐3 and PD‐1 was also reported to enhance the inhibition of tumor growth in MC38 colorectal cancer and Sa1N fibrosarcoma, ⁷⁴ , ⁷⁵ , ⁷⁶ and the activation of the immune system, ⁷⁵ , ⁷⁶ compared with the monotherapies. The immune response induced by blocking LAG‐3 and PD‐1 acts on the production of IFNγ, IL‐2, ⁷⁷ and TNFα by peripheral blood mononuclear cells (PBMCs). ⁷⁶ IL‐2 is a cytokine more associated with an antitumor response through its ability to stimulate T cell expansion, ⁷⁸ , ⁷⁹ while the pro‐ or antitumoral role of IFNγ and TNFα is still ambiguous today. ⁸⁰ , ⁸¹ , ⁸² , ⁸³ Further studies are needed to determine whether they are involved in tumor cytotoxicity or immune suppression. A bispecific antibody targeting LAG‐3 and PD‐1 (YG‐003D3) has also demonstrated its effectiveness in increasing IL‐6 production by PBMCs. ⁷⁶ This result may seem difficult to conciliate with the anticancer immune potentiation effect of the combinational therapy, IL‐6 being a cytokine known for its effects promoting cancer cell invasion and immunosuppression. ⁸⁴ , ⁸⁵ However, IL‐6 has also been described as a molecule that can promote extravasation of T CD8⁺ lymphocytes into tumors. ⁸⁶ It would therefore be interesting to identify whether IL‐6, in the therapeutic context of LAG‐3 and PD‐1 inhibition, promotes T CD8⁺ recruitment within the tumor.

Concerning the blocking of PD‐L1 instead of PD‐1, we can note similar results. Indeed, blocking PD‐L1 and LAG‐3 inhibited tumor growth in colon cancer ⁸⁷ , ⁸⁸ and promoted survival in colon cancer. ⁸⁷ Moreover, this dual therapy induced higher secretion of IL‐2 and IFNγ by PBMCs especially when bispecific antibodies or tribodies targeting PD‐L1 and LAG‐3 were used ⁷⁷ , ⁸⁹ , ⁹⁰ and limited suppression of effector T cells by Treg in melanoma essentially with the bispecific antibody ABL501. ⁸⁹ This dual therapy also increased the proportion of CD8⁺ and CD4⁺ T cells in colorectal cancer or melanoma essentially with the bispecific antibodies IBI323 and ABL501, respectively, ⁸⁸ , ⁸⁹ increased cytotoxicity by CD8⁺ T cells in melanoma, ⁸⁹ and increased TNFα expression by CD4⁺ TILs in MC38 colorectal cancer. ⁸⁸ It also increased the proportion of CD4⁺ and CD8⁺ T cells within the tumor and in the blood in the MC38 colorectal cancer model. ⁸⁸

All in all, the high anti‐tumor effect induced by targeting both LAG‐3 and PD(L)‐1 could be explained by reduced immunosuppression and potentiation of immune responses by CD8⁺ and CD4⁺ T cells.

3.2. LAG‐3 inhibition with other immunotherapies

Targeting other immune checkpoints, such as T cell immunoglobulin and ITIM domain (TIGIT) and T cell immunoglobulin 3 (Tim‐3), has also been tested in combination with the inhibition of LAG‐3 in a model of colon cancer. Indeed, targeting LAG‐3 + TIGIT or LAG‐3 + PD‐1 + Tim‐3 lead to a decrease in tumor growth. ⁹¹ , ⁹² Of note, the targeting of LAG‐3 + TIGIT induced complete tumor regression in a large number of mice as well as an improved survival rate. ⁹¹ As for the targeting of LAG‐3 + PD‐1 + Tim‐3, it increased the production of B‐granzyme by CD8⁺ TILs ⁹² and promoted cytotoxic effects of T lymphocytes against cancer cells. ⁹³ These results once again demonstrate the major role played by CD8⁺ T lymphocytes in the antitumor response induced by anti‐LAG‐3.

3.3. LAG‐3 inhibition and radiation

Radiotherapy is known to destroy cancer cells and is an effective cancer treatment. Nevertheless, its effectiveness can be limited in some patients because the dose required to destroy the tumor tissue can damage the surrounding healthy cells. ⁹⁴ , ⁹⁵ New technologies have been developed to overcome these drawbacks and target cancer cells more specifically. These include microwave ablation and nanoparticles. Indeed, the aim of microwave ablation is to specifically destroy the tumor area by inserting a needle into the tumor and generating heat. The heat emitted comes from microwaves. On the other hand, metal nanoparticle radio‐enhancer activated by ionizing radiation can deliver a highly concentrated dose of radiation directly into the tumor. In a mouse model of MC38 colorectal cancer, the combination of microwave ablation with anti‐LAG‐3 induced a decrease in tumor growth as well as an increase in mice survival, ⁹⁶ which was also observed in a mouse model of anti‐PD‐1‐resistant lung cancer (344SQR), treated with metal nanoparticle radio‐enhancer (NBTXR3), radiation, anti‐PD‐1, anti‐TIGIT, and anti‐LAG‐3 compared with the NBTXR3 + radiation + anti‐PD‐1 + anti‐TIGIT group. ⁹⁷ This indicates that blocking LAG‐3 can improve the effectiveness of radiation therapies. Moreover, the addition of anti‐LAG‐3 to microwave ablation increased the number of CD8⁺ IFNγ⁺ and CD8⁺ TNFα⁺ TILs, ⁹⁶ indicating that the effect of anti‐LAG‐3 was probably mediated by the activation of CD8⁺ T cells.

3.4. LAG‐3 inhibition with adoptive cell therapies

T cell‐based adoptive cell treatments are attracting attention as emerging cancer therapies. This new form of therapy involves using T lymphocytes and genetically modifying them to express a CAR receptor capable of specifically recognizing tumor cells. In a mouse model of B16 melanoma, anti‐LAG‐3 combined with melanoma‐specific CD4⁺ T cells and irradiation increased mouse survival and resulted in significant tumor regression. ⁹⁸ It is noteworthy that the addition of anti‐PD‐L1 instead of anti‐LAG‐3 was less efficient, resulting in fewer mice with a regressed tumor. ⁹⁸ Moreover, the number of melanoma‐specific TCD4⁺ cells increased in the presence of anti‐LAG‐3 compared with anti‐PD‐L1, supporting the notion that anti‐LAG‐3 could also act on CD4⁺ T cells. ⁹⁸

XBP1/CD138/CS1 is a multipeptide that induces multiple myeloma‐specific cytotoxic T lymphocytes, ⁹⁹ , ¹⁰⁰ , ¹⁰¹ , ¹⁰² while B cell maturation antigen (BCMA) is a transmembrane receptor which belongs to the family of tumor necrosis factor receptors and contributes to the growth and survival of myeloma cells. The combination in multiple myeloma of anti‐LAG‐3 with cytotoxic T lymphocytes specific for XBP1/CD138/CS1 or specific for BCMA (YLMFLLRKI) induced an increase in the degranulation of CD8⁺ T cells and of CD45RO⁺ memory cells, respectively. ⁷⁰ , ¹⁰³ Furthermore, anti‐LAG‐3 combined with XBP1/CD138/CS1‐specific cytotoxic T lymphocytes directed against multiple myeloma resulted in an increased proliferation of CD8⁺, central memory, effector memory, CD28⁺, and CD38⁺ CTLs and an increase in IFNγ production. ⁷⁰ This again indicates that the addition of anti‐LAG‐3 promotes an antitumor response mediated by cytotoxic TCD8⁺.

Another adoptive cell therapy based on the use of CD3⁺CD4⁻CD8⁻ double negative T cells showed in breast cancer an upregulation in the mRNA expression of Tbet (a transcription factor involved in the regulation of T cell activation), IFNγ, and perforin (involved in the cell death process) when combined with anti‐LAG‐3. ¹⁵ This demonstrates that anti‐LAG‐3 promoted the cytotoxicity of double‐negative T cells. ¹⁵ This highlights that T cells not differentiated into CD4⁺ or CD8⁺ can also play a role in the antitumor response and that this can be favored by an anti‐LAG‐3 therapy.

3.5. LAG‐3 inhibition with targeted therapies

Various targeted therapies such as those targeting cell membrane markers (e.g., CD20), kinases, or histone deacetylases (HDACs) have been tested in combination with LAG‐3 inhibition. Indeed, molecules targeting CD20, glycogen synthase kinase 3 (Gsk‐3), phosphoinositide‐3‐kinase δ (PI3Kδ), or domatinostat (a HDAC inhibitor) have been tested in combination with LAG‐3 inhibition in B16 melanoma, 4 T1 breast cancer, CT26 colon cancer, or C38 colon carcinoma. ¹⁰⁴ , ¹⁰⁵ , ¹⁰⁶ , ¹⁰⁷ Overall, these dual therapies demonstrated a reduction in tumor growth ¹⁰⁴ , ¹⁰⁵ , ¹⁰⁶ , ¹⁰⁷ or even complete tumor regression in 60% or 80% of mice for Gsk‐3 inhibition + anti‐LAG‐3 combination or domatinostat + anti‐PD‐1 + anti‐LAG‐3 combination, respectively. ¹⁰⁴ , ¹⁰⁶ Gsk‐3, PI3Kδ, and HDAC are present in cancer cells and are involved in proliferation processes. We can therefore hypothesize that inhibition of these molecules prevents cancer cells from proliferating, thus facilitating their elimination by the immune system. The presence of anti‐LAG‐3 limits the inhibition of T cells by cancer cells resistant to Gsk‐3, PI3Kδ, and HDAC inhibitors, which could explain the effects observed above. For these experiments, SB415286 was the molecule used for Gsk‐3 inhibition and PI‐3065 for PI3Kδ inhibition. ¹⁰⁴ , ¹⁰⁶ It should be noted that SB415286 + anti‐LAG‐3 combination treatments led to increasing proportion of CD8⁺ TIL cells expressing granzyme B or IFNγ and TNFα. Therefore, inhibition of LAG‐3 promoted a cytotoxic immune response mediated by CD8⁺ T cells. ¹⁰⁴

In addition, the combination of anti‐LAG‐3 with 41BB agonist antibodies (41BB, also called CD137/TNFSF9, a costimulatory receptor on T cells) and with a CXCR1/2 inhibitor (SX‐682) already used in several phase II clinical trials (NCT04599140, NCT05570825, NCT05604560, NCT04574583) was tested in a mouse model of iKRAS pancreatic cancer. CXCR1/2 are receptors for CXCL8 chemokines and are present on several cell types including myeloid‐derived suppressor cells (MDSCs), which are involved in immunosuppression. ¹⁰⁸ , ¹⁰⁹ SX‐682 has been shown to decrease granulocytic MDSC levels and increase CD8⁺ T tumor infiltration. Moreover, the addition of the anti‐LAG‐3 + anti‐41BB combination prolonged mouse survival, increased CD4⁺ and CD8⁺ T levels in tumors, as well as their effector and memory phenotypes. In addition, IFNγ and TNFα levels in tumor‐infiltrating T cells were increased in the presence of the triple therapy. ¹¹⁰ Although this work did not allow to assert whether anti‐LAG‐3 potentiates the effect of SX‐682 because it was not tested alone without the anti‐41BB agonist, it clearly demonstrated that the dual anti‐LAG‐3/anti‐41BB combination played an important role in enhancing SX‐682 efficacy and antitumor immunity.

3.6. Other reported therapies associated with LAG‐3

Oxysophocarpine, a bioactive alkaloid, is a traditional Chinese medicine that possesses various pharmacological functions. In a hepatocellular carcinoma model, combining it with anti‐LAG‐3 decreased tumor growth and the viability of cancer cells, which was associated with an increase in the cytotoxicity of CD8⁺ T cells. ¹¹¹

Finally, combination of LAG‐3 blockade with a therapeutic vaccine could be beneficial; more precisely, a vaccine based on mesenchymal stromal cells expressing the immunoproteasome complex which gives these cells an antigen cross‐presentation capacity whose epitopes are very different from those of dendritic cells. In a mouse EL4 lymphoma model, adding anti‐LAG‐3 to the vaccine improved the vaccine's efficacy on mouse survival, at a 30%–50% survival rate. ¹¹² Another study showed that AU‐011, a virus‐like drug conjugate (VDC) which induces antitumor immune responses, when combined with anti‐PD‐L1 and anti‐LAG‐3 improved mouse survival in an MC38 colorectal cancer model. ¹¹³

All combinational strategies involving targeting LAG‐3 tested in preclinical studies discussed in this review are summarized in Figure 2.

FIGURE 2.

New treatment strategies that involve targeting lymphocyte activation gene 3 (LAG‐3). Numerous new therapeutic combinations based on LAG‐3 inhibition have been tested. These include (1) bispecific antibodies targeting LAG‐3 and PD‐(L)1, such as ABL501, IBI323, and YG‐003D3; (2) targeting LAG‐3 and other immune checkpoints such as targeting TIGIT + anti‐LAG‐3 or anti‐PD‐1 + anti‐Tim‐3 + anti‐LAG‐3; (3) radiation and immunotherapies such as NBTXR3 (metal nanoparticle radio‐enhancer) + radiation + anti‐PD‐1 + anti‐TIGIT + anti‐LAG‐3; (4) adoptive cell therapies based on cancer‐specific T lymphocytes associated with anti‐LAG‐3; (5) vaccines, such as AU‐011, a virus‐like drug conjugate (VDC) associated with anti‐LAG‐3 and anti‐PD‐L1; (6) targeted therapies in combination with anti‐LAG‐3 +/− other therapy such as CD20 targeting + anti‐LAG‐3, SB415286 (Gsk‐3 inhibitor) + anti‐LAG‐3, domatinostat (histone deacetylase [HDAC] inhibitor) + anti‐PD‐1 + anti‐LAG‐3, PI‐3065 (PI3Kδ inhibitor) + anti‐LAG‐3, and SX‐682 (CXCR1/2 inhibitor) + anti‐41BB + anti‐LAG‐3.

4. CONCLUDING REMARKS AND CHALLENGES AHEAD

Still many cancer patients either gain minimal benefit from anticancer therapies or eventually progress, leaving an unmet need for the development of novel therapeutic agents and strategies. LAG‐3 acts by inhibiting T cells’ anticancer response and providing a tumor microenvironment favorable for tumor escape different from other checkpoint receptors. Molecules targeting LAG‐3 prevent the interaction of LAG‐3 with its ligands (mainly MHC II, LSECTin, FGL‐1, and galectin‐3), thus preventing T lymphocytes from being inhibited. It is in combinational therapies, notably with inhibitors of other immune checkpoints, that anti‐LAG‐3 has become a promising option for oncologists in precision medicine. The association of anti‐LAG‐3 with anti‐PD‐1 is particularly exciting since it shows a remarkable synergy and a decrease in the number of resistant patents. The clinical studies highlighted in this review show that dual therapies combining anti‐PD‐1 with anti‐LAG‐3 are booming and are proving to be highly effective in improving patient longevity, as demonstrated for instance in the RELATIVITY‐047 phase III clinical trial. To date, at least two LAG‐3‐directed agents (eftilagimod alpha, relatlimab) have been approved by the FDA in combination with PD‐1 inhibitors in the setting of advanced solid tumors. Dual inhibition of PD‐1 and LAG‐3 appears to facilitate sustained responses and disease control. It is worth mentioning that, overall, the molecules targeting LAG‐3 currently in clinical trials appear to be well tolerated, whether used alone or in combination.

Preclinical studies focused on revealing the detailed cellular and molecular mechanisms by which LAG‐3 inhibitors enhance tumor immunity indicate mainly the involvement of CD8⁺ T lymphocyte activation and cytotoxicity, as well as the involvement of CD4⁺ T cells. Numerous other preclinical studies involving LAG‐3 inhibition coupled with other therapies (chemotherapies, radiotherapy, adoptive cell therapies, other immunotherapies, targeted therapies), also discussed in this review, lead to similar conclusions, amplifying the interest in LAG‐3 blockade as a potentiator of cancer therapies. It would now be interesting to study in more detail the immune response, particularly in which CD4⁺ T lymphocyte subtypes are involved in tumor regression, and what role each cytokine plays in the tumor immune response (IFNγ, TNFα, IL‐10, IL‐6, …) in different therapeutic combinations in order to flesh out the immune response mechanisms further and perhaps find even more effective therapies. Some preclinical studies on new treatment combinations involving LAG‐3 blockade have been completed, opening the door to further clinical trials. The diversity of molecules associated with LAG‐3 highlighted here and the synergistic advantages may allow a wider choice of therapies in the clinic and, toward a more personalized medicine, to define the optimal combinational treatment for each patient. It also paves the way for developing dual‐targeting drugs like the already reported bispecific LAG‐3/PD‐L1 antibodies. ⁸⁸ , ¹¹⁴

It should be noted that the anti‐PD‐1 (nivolumab)/anti‐LAG‐3 (relatlimab) combination tested in the RELATIVITY‐047 clinical trial is also being tested in another phase II clinical trial as neoadjuvant therapy (NCT02519322). The advantage of neoadjuvant treatment is not only to reduce tumor size before surgery, so as to limit the amount of healthy tissue to be removed, but also to understand the mechanisms of resistance or of response to the dual therapy and thereby to adapt the adjuvant treatment subsequently. Such studies could also provide answers concerning the immunological mechanisms of action of anti‐LAG‐3 therapy. In conclusion, LAG‐3 inhibitors, through their ability to reinvigorate anticancer immune responses, may provide a critical piece in assembling combination therapies capable of achieving durable responses against advanced cancers, with immune adverse events minimized.

AUTHOR CONTRIBUTIONS

Aude Chavanton: Conceptualization; investigation; writing – original draft. Flavie Mialhe: Investigation; resources; validation. Jimena Abrey: Investigation; resources; validation. Alvaro Baeza Garcia: Resources; supervision; validation. Carmen Garrido: Supervision; validation; writing – review and editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

ETHICS STATEMENT

Approval of the research protocol by an Institutional Reviewer Board: N/A.

Informed Consent: N/A.

Registry and the Registration No. of the study/trial: N/A.

Animal Studies: N/A.

Chavanton A, Mialhe F, Abrey J, Baeza Garcia A, Garrido C. LAG‐3 : recent developments in combinational therapies in cancer. Cancer Sci. 2024;115:2494‐2505. doi: 10.1111/cas.16205