Innate immune checkpoint inhibitors: the next breakthrough in medical oncology?

Abstract

While immunotherapy has revolutionized the treatment of many types of advanced cancer, most patients still do not derive benefit. The currently available immune checkpoint inhibitors target the adaptive immune system, generating a T cell anti-tumor response. However, an anti-tumor immune response depends on a complex interplay of both innate and adaptive immune cells. The innate immune system is a promising new target, and innate immune checkpoint inhibitors can disrupt inhibitory interactions (“don’t eat me” signals) between tumor and both phagocytes and natural killer cells. The checkpoint inhibitor may also provide a stimulatory interaction (“eat me” signal), or this can be achieved through use of combination therapy. This generates anti-tumor effector functions including phagocytosis, natural cytotoxicity, antibody-dependent effects, and synergistic activation of the adaptive immune system via antigen presentation. This is a rapidly expanding area of drug development, either alone or in combination (with anti-cancer antibodies or adaptive immune checkpoint inhibitors). Here, we comprehensively review the mechanism of action and up-to-date solid tumor clinical trial data of the drugs targeting phagocytosis checkpoints (SIRPα/CD47, LILRB1/MHC-I, and LILRB2/MHC-I) and natural killer cell checkpoints (TIGIT/CD112 + CD155, PVRIG/CD112, KIRs/MHC-I, and NKG2A-CD94/HLA-E). Innate immune checkpoint inhibitors could once again revolutionize immune-based cancer therapies.

Introduction

Immune function is governed in part by a complex array of stimulatory and inhibitory cell-surface interactions. A coordinated cancer immunity cascade must occur for immune-based therapies to be effective, including expression/release of cancer antigens, antigen presentation, immune cell activation, immune cell transport and infiltration into tumor, and cancer cell killing by both adaptive and innate mechanisms through interaction of stimulatory and inhibitory processes.¹ Inhibitory immune checkpoints exist to maintain physiologic immune responses, i.e. to induce tolerance (Figure 1). Immune checkpoint inhibitors have been successful in the treatment of many solid tumors, but the majority of advanced solid tumors do not respond and resistance usually emerges.^2,3 The currently available immune checkpoint inhibitors target cytotoxic T-lymphocyte antigen-4 (CTLA-4), programmed death-1 (PD-1), and its ligand (PD-L1), all attempting to bolster T cell activation.

Figure 1. Innate immune cell-surface regulatory interactions & therapeutic targets.

Innate immune cell anti-tumor function is regulated in part by stimulatory (“eat me”) and inhibitory (“don’t eat me”) cell-surface interactions with the cancer cell, generally involving a receptor on the immune cell and antigen on the cancer cell. Inhibitory interactions (“checkpoints”) exist to maintain physiologic immune responses, i.e. to induce tolerance. Normal human cells generally avoid immune destruction by expressing more inhibitory than stimulatory antigens. Cancer cells, likewise, can evade the innate immune system by expressing inhibitory antigens and lacking stimulatory antigens (figure left). Cancer cells may also express stimulatory antigens, activating innate immune cells and resulting in phagocytosis (phagocytes) or natural cytotoxicity (natural killer cells). Conceptually, there are four scenarios under which the balance of stimulatory and inhibitory signaling favors innate immune cell anti-tumor effects. Firstly, the cancer cell lacks inhibitory antigens and stimulatory signals are present; radiation/chemotherapy may increase cancer cell expression of stimulatory antigens (figure right, #1). Secondly, an innate immune checkpoint inhibitor can block the inhibitory checkpoint while a stimulatory signal is present (figure right, #2). Thirdly, a combination therapy approach can block the inhibitory checkpoint (with an innate immune checkpoint inhibitor lacking an active Fc region) while a second drug (anti-cancer antibody) can provide the stimulatory signal upon interacting with innate immune cell Fc receptors (figure right, #3). Finally, an innate immune checkpoint inhibitor containing an active Fc region can both block the inhibitory checkpoint and provide the stimulatory signal upon interacting with an innate immune cell Fc receptor (figure right, #4). This final strategy is often limited by hematologic toxicity. Created with BioRender.com.

There are many mechanisms of resistance to immunotherapy, including loss of tumor antigen expression, increased expression of tumor inhibitory checkpoint ligands (such as PD-L1) which bind cognate ligands on immune cells, and recruitment of suppressive cell populations by tumor-produced immunosuppressive factors.^2,4,5 Indeed, infiltration of both innate and adaptive immune cells into tumor has been shown to correlate with improved outcomes in many cancer types.^6–9 Novel treatments targeting the innate immune system are promising candidates in medical oncology.

Innate immune cells are derived from both myeloid and lymphoid lineages. Myeloid cells include monocytes, macrophages, dendritic cells (DC), polymorphonuclear cells, and mast cells. Natural killer cells (NK) are innate lymphoid cells. Innate immune cells activate the adaptive immune system via antigen presentation and execute primary effector functions, such as phagocytosis, natural cytotoxicity, and antibody-dependent cellular cytotoxicity and phagocytosis (ADCC and ADCP, respectively) upon activation of their Fc receptors for antibody.¹⁰ In fact, the innate immune system is required to generate antigen-specific T cell activity.¹¹ Innate immune checkpoint inhibitors, with or without adaptive immune checkpoint inhibitors, may overcome resistance to currently approved immune checkpoint drugs.

Innate immune cell function is regulated by a balance of stimulatory and inhibitory cell-surface interactions (Figure 1). Blocking inhibitory checkpoints (“don’t eat me” signals) between tumor and innate immune cells is a novel therapeutic target. The checkpoint inhibitor may also provide a stimulatory interaction (“eat me” signal), or this can be achieved through use of combination therapy. When the balance of stimulatory and inhibitory interactions favors activation of innate immune cells, the result is phagocytosis or natural cytotoxicity of cancer cells. Here, we will review basic science of the checkpoints and the mechanism of action, safety, and efficacy data of all innate immune checkpoint inhibitors currently being studied in solid tumor clinical trials.

Methods

All clinical data from solid tumor clinical trials on the innate immune checkpoint inhibitors is included. Studies were identified by searching for the checkpoints and drugs using the PubMed database, ClinicalTrials.gov, and abstracts from the following annual meetings: American Association of Cancer Research, American Society of Clinical Oncology, European Society for Medical Oncology, and the Society for Immunotherapy of Cancer. Objective response rate (ORR) is defined as partial response (PR) or complete response (CR), as assessed in the clinical trial. Disease control rate (DCR) is defined as stable disease (SD), PR, or CR, as assessed in the clinical trial. Table 1 provides a summary of the available solid tumor clinical safety and efficacy data of the innate immune checkpoint inhibitors. Tables 2 & 3 list the active clinical trials on ClinicalTrials.gov.

Table 1:

Results of solid tumor clinical trials of phagocytosis & NK cell checkpoint inhibitors

NCT	In combination with	Phase	Cancer type	ORR	DCR	mPFS (months)	Primary Toxicitya
SIRPα/CD47
ALX148 (ALX Oncology, anti-CD47 fusion (CD47 binding domain of SIRPα to inactive human Ig Fc)
0301321822	Alone, pembrolizumab, trastuzumab, or rituximab	I	Solid & NHL	0% (alone)	16% (alone)	NR	Fatigue, headache, thrombocytopenia
0301321823	Pembrolizumab, trastuzumab	I	Solid	4% (NSCLC) 18% (HNSCC) 19% (gastric)	39% (NSCLC) 41% (HNSCC) 48% (gastric cancer)	NR	Fatigue, AST elevation, anemia
0301321824	Pembrolizumab, trastuzumab, or chemo	I	HNSCC & gastric	50% (ICI-naïve HNSCC) 0% (prior ICI HNSCC) 20% (gastric cancer)	NR	4.6 (ICI-naïve HNSCC) 2 (prior ICI HNSCC) 2.2 (gastric cancer)	Fatigue, AST increase, thrombocytopenia
IBI188 (Innovent Biologics, anti-CD47 humanized IgG4 mAb)
0376314933	Alone	I	Solid & lymphoma	NR	NR	NR	Nausea, back pain, fatigue
Lemzoparlimab (TJC4, TJ011133) (AbbVie and I-Mab Biopharma, anti-CD47 fully human IgG4 mAb)
0393481432	Alone	I	Solid	NR	NR	NR	Anemia, fatigue, infusion reaction
Magrolimab (Hu5F9-G4, ONO-7913) (Gilead, anti-CD47 humanized IgG4 mAb)
0221640920	Alone	I	Solid & heme	NR	NR	NR	Anemia, fatigue, headaches
0355813993	Avelumab	I	Solid	0%	56%	NR	Headache, fatigue, infusion reaction, anemia
0295378226	Cetuximab	I/II	Solid & CRC	7% (KRAS WT CRC) 0% (KRAS MT CRC)	NR (KRAS WT CRC) 45% (KRAS MT CRC)	3.6 (KRAS WT CRC) 1.9 (KRAS MT CRC)	Acneiform rash, dry skin, fatigue
SRF231 (Surface Oncology, anti-CD47 fully human IgG4 mAb)
0351234030	Alone	I	Solid & heme	0%	NR	NR	Fatigue, headache, pyrexia
LILRB2/MHC-I
Etigilimab (OMP-313M32) (Mereo BioPharma, anti-TIGIT IgG1 humanized mAb)
0311942858	Alone or nivolumab	I	Solid	0%	22%	NR	Rash, fatigue, nausea
MK-4830 (Merck and Agenus, anti-LILRB2 fully human IgG4 mAb)
0356469138	Alone or pembrolizumab	I	Solid	13%	NR	NR	None reported
Tiragolumab (MTIG7192A, RG6058) (Genentech, anti-TIGIT humanized IgG1/kappa mAb)
0279457152	Alone	I	Solid	0% (alone) 6% (w/ atezo)	17% (alone) NR (w/ atezo)	NR	Fatigue, anemia
0279457152	Atezolizumab	I	NSCLC (PD-L1+ & ICI-naïve)	50%	79%	NR	Fatigue, anemia
0356371654	Atezolizumab	II	NSCLCb	37%	NR	5.6	Rash, infusion reactions
Vibostolimab (MK-7684) (Merck, anti-TIGIT humanized antibody)
0296401355	Alone or pembrolizumab	I	Solid	3% (alone) 19% (w/ pembro)	35% (alone) 47% (w/ pembro)	NR	Fatigue, pruritus, anemia
0296401356,57	Alone or pembrolizumab	I	NSCLC	7% (alone) 5% (w/ pembro) 29% (ICI-naïve)	NR	9 (alone) 13 (w/ pembro)	Pruritus, fatigue, rash
PVRIG / CD112
COM701 (Compugen, anti-PVRIG humanized IgG4 mAb)
0366771662-64	Alone or nivolumab	I	Solid	NR	57%	NR	Fatigue, nausea, anemia
KIRs / MHC-I
Lirilumab (IPH2102, BMS-986015) (Innate Pharma and Bristol-Myers-Squibb, anti-KIR2D fully human IgG4 mAb)
0175058070,71	Nivolumab or ipilimumab	I	Solid	24%	52%	NR	Fatigue, pruritus, infusion reaction
2009-011526-33c; 69	Alone	I	Solid & heme	0%	59%	6.6 (AML) 19.6 (CLL) 5.3 (ovarian)	Asthenia, pruritus, fatigue
NKG2A-CD94 / HLA-E
Monalizumab (IPH2201) (Innate Pharma and AstraZeneca, humanized anti-NKG2A-CD94 IgG4 mAb)
0267143575	Durvalumab	I	Solid & CRC	8%	38%	NR	Diarrhea
0267143576,77	Durvalumab, FOLFOX, and (bevacizumab or cetuximab)	I/II	CRC	41% (bevacizumab) 55% (cetuximab)	88%	NR	Fatigue, nausea, peripheral neuropathy
0308805978	Alone	II	HNSCC	0%	22%	1.72	None reported
0264355079-82	Cetuximab	I/II	HNSCC	20%	NR	4.5	Fatigue, pyrexia, headache
0245930183	Alone	I	Gynecologic	0%	41% (dose esc) 18% (dose exp)	NR	Headache, fatigue, vomiting

^a.Most common any-grade treatment-related adverse events.

^b.Chemotherapy-naïve, PD-L1+, EGFR and ALK wild-type.

^c.EudraCT (European Union Drug Regulating Authorities Clinical Trials Database) identifier.

Abbreviations: AML, acute myeloid leukemia; atezo, atezolizumab; chemo, chemotherapy; ICI, anti-PD-(L)1 immune checkpoint inhibitor; CLL, chronic lymphocytic leukemia; CRC, colorectal cancer; DCR, disease control rate; esc, escalation; exp, expansion; heme, hematologic; HNSCC, head/neck squamous cell carcinoma; mAb, monoclonal antibody; mPFS, median progression-free survival; MT, mutated; NCT, ClinicalTrials.gov identifier; NHL, non-Hodgkin lymphoma; NR, not reported; NSCLC, non-small cell lung cancer; ORR, objective response rate; pembro, pembrolizumab; PFS, progression-free survival; WT, wild-type.

Table 2.

Clinical trials of phagocytosis checkpoint inhibitors registered on ClinicalTrials.gov

NCT	Status	Phase	Cancer Type	In Combination With	Enrollmenta
SIRPα/CD47
ALX148 (ALX Oncology, anti-CD47 fusion (CD47 binding domain of SIRPα to inactive human Ig Fc)
03013218	Recruiting	I	Solid cancers and NHL	Pembrolizumab, chemo, or anti-cancer antibodies	184
04417517	Recruiting	I/II	High risk MDS	Azacitadine	63
AO-176 (Arch Oncology, anti-CD47 humanized IgG2 mAb)
03834948	Recruiting	I/II	Solid cancers	Alone or paclitaxel	132
04445701	Recruiting	I/II	Multiple myeloma	Alone, dexamethasone or bortezomib	102
BI 765063 (OSE-172) (OSE Immunotherapeutics and Boehringer Ingelheim, anti-SIRPα humanized IgG4 mAb)
03990233	Recruiting	I	Solid cancers	Alone or BI-754091	116
CC-90002 (Celgene, anti-CD47 humanized IgG4 mAb)
02367196	Not yet open	I	Hematologic neoplasms	Alone or rituximab	60
02641002	Terminated	I	AML, MDS	Alone	28
CC-95251 (Celgene, anti-SIRPα humanized mAb)
03783403	Recruiting	I	Hematologic and solid cancers	Monotherapy, rituximab, or cetuximab	230
DSP107 (Kahr Medical, SIRPα & 4-1BB ligand bi-functional fusion protein)
04440735	Recruiting	I	Solid cancers	Alone or atezolizumab	100
FSI-189 (Gilead, anti-SIRPα antibody)
04502706	Not yet open	I	NHL	Alone or rituximab	63
HX009 (Waterstone Pharmaceuticals, anti-PD-1 and -CD47 bispecific antibody fusion protein)
04097769	Recruiting	I	Solid cancers	Monotherapy	37
IBI322 (Innovent Biologics, recombinant anti-CD47/PD-L1 bispecific antibody)
04338659	Not yet open	I	Solid cancers	Alone	45
04328831	Recruiting	I	Solid cancers	Alone	218
IBI188 (Innovent Biologics, anti-CD47 humanized IgG4 mAb)
04511975	Recruiting	I	MDS	Azacitadine	32
04485052	Not yet open	I	AML	Azacitadine	126
04485065	Not yet open	I	MDS	Azacitadine	12
03763149	Not yet open	I	Hematologic and solid cancers	Alone	42
03717103	Recruiting	I	Hematologic and solid cancers	Alone or rituximab	92
IMC-002 (ImmuneOncia Therapeutics, anti-CD47 fully human IgG4 mAb)
04306224	Recruiting	I	Solid cancers and lymphoma	Alone	24
Lemzoparlimab (TJC4, TJ011133) (AbbVie and I-Mab Biopharma, anti-CD47 fully human IgG4 mAb)
04202003	Recruiting	I/II	AML and MDS	Alone	42
03934814	Recruiting	I	Solid cancers and lymphoma	Alone, pembrolizumab, or rituximab	88
Magrolimab (Hu5F9-G4, ONO-7913) (Gilead, anti-CD47 humanized IgG4 mAb)
02216409	Completed	I	Solid cancers	Alone	88
03558139	Not yet open	I	Solid cancers	Avelumab	32
03248479	Recruiting	I	AML, MDS	Alone or azacitadine	257
03922477	Recruiting	I	AML	Atezolizumab	21
02678338	Completed	I	AML, MDS	Alone	20
03527147	Recruiting	I	NHL	Acalabrutinib and Rituximab	88
04403308	Recruiting	I	Solid cancers	Alone	12
02953509	Recruiting	I/II	NHL	Rituximab	422
04435691	Recruiting	I/II	AML	Azacitadine or venetoclax	38
04541017	Not yet open	I/II	Mycosis fungoides, Sezary syndrome	Mogamulizumab	100
02953782	Completed	I/II	Solid cancers	Cetuximab	78
03869190	Recruiting	I/II	Urothelial cancer	Atezolizumab	385
04313881	Recruiting	III	MDS	Azacitadine	180
SHR-1603 (Jiangsu HengRui Medicine, anti-CD47 humanized IgG4 mAb)
03722186	Not yet open	I	Hematologic and solid cancers	Alone	128
04282070	Recruiting	I	Nasopharyngeal cancer	Alone	40
03710265	Recruiting	I	Solid cancers	Alone	112
SL-172154 (Shattuck Labs, bi-functional fusion protein SIRPα -Fc-CD40L against CD47 and CD40)
04406623	Recruiting	I	Ovarian cancer	Alone	40
04502888	Not yet open	I	SCC of the head/neck or skin	Alone (intratumoral)	18
SRF231 (Surface Oncology, anti-CD47 fully human IgG4 mAb)
03512340	Completed	I	Hematologic and solid cancers	Alone	148
TG-1801 (TG Therapeutics and Novimmune, anti-CD47/CD19 bispecific antibody)
03804996	Recruiting	I	B-cell lymphoma	Alone or ublituximab	16
TTI-621 (Trillium Therapeutics, anti-CD47 recombinant fusion (CD47 binding domain of human SIRPα fused human IgG1 Fc))
02890368	Terminated	I	Solid tumors, mycosis fungoidesb	Alone, anti-PD-(L)1, PEG-IFN-α2a, T-Vec, or RT	56
02663518	Recruiting	I	Hematologic and solid cancers	Alone, rituximab, or nivolumab	260
TTI-622 (Trillium Therapeutics, anti-CD47 recombinant fusion (CD47 binding domain of human SIRPα fused human IgG1 Fc))
03530683	Recruiting	I	Lymphoma or multiple myeloma	Alone, rituximab, anti-PD-1, or PI	156
ZL-1201 (Zai Lab, anti-CD47 humanized IgG4 monoclonal antibody)
04257617	Recruiting	I	Hematologic and solid cancers	Alone	65
LILRB2/MHC-I
MK-4830 (Merck and Agenus, anti-LILRB2 fully human IgG4 monoclonal antibody)
03564691	Recruiting	I	Solid tumors	Alone, pembrolizumab, chemo, or lenvatinib	290

^a.Projected or actual.

Abbreviations: AML, acute myeloid leukemia; chemo, chemotherapy; MDS, myelodysplastic syndrome; NCT, ClinicalTrials.gov identifier; mAb, monoclonal antibody; NHL, non-Hodgkin lymphoma; SCC, squamous cell carcinoma.

^b.Percutaneously accessible solid tumors or mycosis fungoides for intra-tumoral injection.

Abbreviations: AML, acute myeloid leukemia; MDS, myelodysplastic syndrome; chemo, chemotherapy; NCT, ClinicalTrials.gov identifier; mAb, monoclonal antibody; NHL, non-Hodgkin lymphoma; RT, radiation therapy; SCC, squamous cell carcinoma; PI, proteasome inhibitor.

Table 3.

Clinical trials of natural killer cell checkpoint inhibitors registered on ClinicalTrials.gov

NCT	Status	Phase	Cancer Type	In combination with	Enrollmenta
TIGIT/CD112 + CD155
AB154 (Arcus Biosciences, anti-TIGIT human IgG1 mAb)
03628677	Recruiting	I	Solid cancers	Alone or zimberelimab	66
04262856	Recruiting	II	Lung Cancer	Zimberelimab or AB928	150
ASP8374 (PTZ-201) (Astellas, anti-TIGIT fully human IgG4 mAb)
03945253	Completed	I	Solid cancers	Alone	6
03260322	Not yet open	I	Solid cancers	Alone or pembrolizumab	300
BGB-A1217 (BeiGene, anti-TIGIT humanized IgG1-variant mAb)
04047862	Recruiting	I	Solid cancers	Tislelizumab	39
BMS-986207 (Bristol Meyers Squibb, anti-TIGIT human IgG1 mAb with null Fcγ receptor)
04570839	Recruiting	I/II	Solid cancers	COM701 and nivolumab	100
02913313	Not yet open	I/II	Solid cancers	Alone or nivolumab	170
04150965	Recruiting	I/II	Multiple myeloma	Alone or pomalidomide and dexamethasone	104
Etigilimab (OMP-313M32) (Mereo BioPharma, anti-TIGIT IgG1 humanized mAb)
03119428	Terminated	I	Solid cancers	Alone or nivolumab	33
SGN-TGT (SEA-TGT) (Seattle Genetics, anti-TIGIT human mAb)
04254107	Recruiting	I	Hematologic and solid cancers	Monotherapy or pembrolizumab	111
Tiragolumab (MTIG7192A, RG6058) (Genentech, anti-TIGIT humanized IgG1/kappa mAb)
04584112	Recruiting	I	Triple negative breast cancer	Atezolizumab or chemo	60
04045028	Recruiting	I	Multiple myeloma, NHL	Alone, daratumumab, or rituximab	52
02794571	Recruiting	I	Solid cancers	Alone, atezolizumab, or chemo	400
04524871	Recruiting	I/II	HCC	Atezolizumab and bevacizumab	100
03281369	Recruiting	I/II	Gastric or esophageal cancer	Atezolizumab or chemo	410
03869190	Recruiting	I/II	Urothelial cancer	Atezolizumab	385
03193190	Recruiting	I/II	Pancreatic cancer	Atezolizumab or chemo	260
04300647	Recruiting	II	Cervical cancer	Atezolizumab	160
04543617	Recruiting	III	Esophageal SCC	Atezolizumab	750
04540211	Recruiting	III	Esophageal SCC	Atezolizumab or chemo	450
03563716	Not yet open	III	NSCLC	Atezolizumab	135
04294810	Recruiting	III	NSCLC	Atezolizumab	500
04513925	Recruiting	III	NSCLC	Atezolizumab	800
04256421	Recruiting	III	NSCLC	Atezolizumab or chemo	400
Vibostolimab (MK-7684) (Merck, anti-TIGIT humanized antibody)
02964013	Recruiting	I	Solid cancers	Alone, pembrolizumab, or chemo	432
04305054	Recruiting	I/II	Melanoma	Pembrolizumab	135
04305041	Recruiting	I/II	Melanoma	Pembrolizumab and MK-1308	200
04303169	Recruiting	I/II	Melanoma	Pembrolizumab	65
04165070	Recruiting	II	Non-small cell lung cancer	Pembrolizumab or chemo	90
PVRIG/CD112
COM701 (Compugen, anti-PVRIG humanized IgG4 mAb)
03667716	Recruiting	I	Solid cancers	Alone or nivolumab	140
04570839	Recruiting	I/II	Solid cancers	BMS-986207 and nivolumab	100
KIRs/MHC-I
IPH2101 (1–7F9) (Innate Pharma and Bristol-Myers-Squibb, anti-KIR2D fully human IgG4 mAb)
01217203	Completed	I	Multiple myeloma	Lenalidomide	15
01256073	Completed	I	AML	Alone	21
00552396	Completed	I	Multiple myeloma	Alone	32
00999830	Completed	II	Multiple myeloma	Alone	27
01222286	Completed	II	Smoldering myeloma	Alone	30
01248455	Terminated	II	Smoldering myeloma	Alone	9
Lirilumab (IPH2102, BMS-986015) (Innate Pharma and Bristol-Myers-Squibb, anti-KIR2D fully human IgG4 mAb)
01750580	Completed	I	Solid cancers	Ipilimumab	22
02252263	Completed	I	Multiple myeloma	Elotuzumab	44
03203876	Not yet open	I	Solid cancers	Nivolumab and ipilimumab	21
03532451	Not yet open	I	Bladder cancer	Nivolumab	43
03347123	Not yet open	I/II	Hematologic and solid cancers	Epacadostat and nivolumab	11
01714739	Completed	I/II	Solid cancers	Nivolumab and ipilimumab	337
02813135	Recruiting	I/II	Solid cancers	Nivolumab	397
01592370	Not yet open	I/II	Hematologic neoplasms	Nivolumab	375
01687387	Completed	II	Acute myeloid leukemia	Alone	152
02599649	Terminated	II	MDS	Alone, nivolumab, and/or azacitadine	10
02399917	Terminated	II	AML	Azacitadine	36
02481297	Completed	II	Leukemia	Rituximab	7
03341936	Recruiting	II	Head and neck cancer	Nivolumab	58
NKG2A-CD94/HLA-E
Monalizumab (IPH2201) (Innate Pharma and AstraZeneca, humanized anti-NKG2A-CD94 IgG4 mAb)
02459301	Completed	I	Gynecologic cancers	Alone	59
02921685	Recruiting	I	Hematologic neoplasms	Alone	18
02671435	Not yet open	I/II	Colorectal cancer	Durvalumab and cetuximab	383
02643550	Recruiting	I/II	HNSCC	Alone, cetuximab, or anti-PD(L)1	140
02557516	Terminated	I/II	Chronic lymphocytic leukemia	Alone	22
02331875	Terminated	I/II	HNSCC	Alone	3
03088059	Recruiting	II	HNSCC	Alone, durvalumab	340
04307329	Not yet open	II	Breast cancer	Trastuzumab	38
03822351	Not yet open	II	NSCLC	Durvalumab	189
03794544	Not yet open	II	NSCLC	Durvalumab	80
03833440	Recruiting	II	NSCLC	Durvalumab	120
04590963	Recruiting	III	HNSCC	Cetuximab	600

^a.Projected or actual.

Abbreviation: AML, acute myeloid leukemia; chemo, chemotherapy; HCC, hepatocellular carcinoma; HNSCC, head/neck squamous cell carcinoma; NCT, ClinicalTrials.gov identifier; mAb, monoclonal antibody; NHL, non-Hodgkin lymphoma; NSCLC, non-small cell lung cancer; SCC, squamous cell carcinoma.

Results

Phagocytosis checkpoint inhibitors

The phagocytes are macrophages, neutrophils, and DCs. Upon activation, phagocytes engulf target cells. Macrophages and neutrophils then directly destroy the target cell (macrophages by lysosomal proteolysis and neutrophils by oxidative burst).¹² DCs, on the other hand, process antigens and load them onto major histocompatibility complex class I (MHC-I) or II (MHC-II) to present to T cells, thereby activating the adaptive immune system.¹² Antigen presentation is required for T cell anti-tumor activity.¹⁰ The amount of tumoral conventional DCs (a subpopulation of DCs even more specialized in antigen presentation) predicts survival across multiple tumor types and predicts response to anti-PD-L1 therapy.^9,13

Thus, phagocytosis checkpoint inhibitors directly destroy tumor cells and activate the adaptive immune system (via antigen presentation), providing rationale for use with adaptive immune checkpoint inhibitors (targeting CTLA-4 or PD-(L)1), which strengthen T cell anti-tumor effects. Additionally, while the PD-1-PD-L1 axis is primarily characterized as a T cell checkpoint, there is evidence of innate immune regulation. Specifically, blockade of the PD-1-PD-L1 axis in mice lacking a functional immune system, other than intact macrophages, generated anti-tumor response.¹⁴ Similarly, PD-L1 knockout tumor cells significantly increased phagocytosis by macrophages.¹⁴

Phagocytosis of cancer cells is regulated by pro- and anti-phagocytic receptor-ligand interactions (Figure 1). The primary prophagocytic (“eat me”) signals on cancer cells are calreticulin (interacts with LRP1 on phagocytes), SLAMF7 (interacts Mac-1 on phagocytes), and tumor antigen bound by antibody (interacts with Fcγ receptor on phagocytes); these antigens are minimally expressed on most normal cells.^4,15 Radiation therapy and chemotherapy may also increase cancer cell expression of prophagocytic signals, synergizing with phagocytosis checkpoint inhibitors.^4,16 The primary inhibitory checkpoints between phagocytes and tumor cells, respectively, are SIRPα/CD47, LILRB1/MHC-I, and LILRB2/MHC-I (Figure 2). Drugs targeting these checkpoints are typically designed to be given with a prophagocytic signal such as an opsonizing anti-cancer antibody, radiation, or chemotherapy (Figure 1).

Figure 2. Innate immune checkpoints and inhibitors.

Tumor cells express ligands (“don’t eat me” signals) which interact with phagocyte (Panel A) and natural killer (Panel B, NK) cell-surface receptors, stimulating signaling pathways which inhibit phagocytosis and natural cytotoxicity, respectively. The phagocytosis checkpoints (Panel A) include (a) SIRPα/CD47, (b) LILRB1/MHC-I, and (c) LILRB2/MHC-I. Drugs are being developed to interrupt these signaling pathways, thereby increasing phagocytosis of tumor cells; drugs with presented/published clinical trial data are shown. (a) The anti-CD47 drugs are ALX148 (engineered fusion protein - two high affinity CD47 binding domains of SIRPα linked to an inactive Fc), magrolimab (anti-CD47 humanized IgG4 mAb), SRF231 (anti-CD47 fully human IgG4 mAb), lemzoparlimab (anti-CD47 humanized IgG4 mAb), and IBI188 (anti-CD47 humanized IgG4 mAb). No compounds blocking the LILRB1/MHC-I axis have available clinical trial results (b). MK-4830 is an anti-LILRB2 fully human IgG4 mAb (c). The NK checkpoints (Panel B) include (a) TIGIT/CD155+CD112, (b) PVRIG/CD112, (c) Inhibitory KIRs/MHC-I, and (d) NKG2A-CD94/HLA-E. Drugs are being developed to interrupt these signaling pathways, thereby increasing natural cytotoxicity of tumor cells; drugs with presented/published clinical trial data are shown. Tiragolumab, vibostolimab, and etigilimab are anti-TIGIT humanized IgG1 mAbs (a). COM701 is an anti-PVRIG humanized IgG4 mAb (b). Lirilumab is an anti-KIR2D fully human IgG4 mAb (c). Monalizumab is an anti-NKG2A-CD94 humanized IgG4 mAb (d). Created with BioRender.com.

Abbreviations: B2MG, beta-2-microglobulin; Fc, immunoglobulin fragment crystallizable region; HLA, human leukocyte antigen; Ig, immunoglobulin; KIR, killer cell immunoglobulin-like receptor; LILRB1, leukocyte immunoglobulin-like receptor B1; LILRB2, leukocyte immunoglobulin-like receptor B2; mAb, monoclonal antibody; MHC, major histocompatibility complex; NKG2A, NK group 2 member A; PVRIG, poliovirus receptor related immunoglobulin domain containing; SIRPα, signal-regulatory protein-α; TIGIT, T cell immunoglobulin and immunoreceptor tyrosine-based inhibition motifs domain.

SIRPα/CD47

Signal-regulatory protein-α (SIRPα) is an inhibitory receptor expressed on myeloid phagocytes. SIRPα contains an extracellular immunoglobulin domain for ligand binding and intracellular immunoreceptor tyrosine-based inhibition motifs (ITIM).^4,10 CD47, the ligand for SIRPα, is widely expressed on cancer cells.¹⁷ The SIRPα/CD47 interaction results in phosphorylation of the ITIMs and downstream disruption of the cytoskeleton, resulting in inhibition of phagocytosis.¹⁷ Blockade of the SIRPα/CD47 axis increases phagocytosis, resulting in cell death and antigen presentation. Tumor killing can be further stimulated if an active Fc domain is supplied, either on the anti-CD47 agent or on a separate anti-cancer antibody, both of which generate ADCP or ADCC.¹⁰ In addition to directly blocking the SIRPα/CD47 interaction, pre-clinical work has demonstrated an evolving potential drug target. Glutaminyl-peptide cyclotransferase-like protein (QPCTL) is responsible for the formation of pyroglutamate on CD47 at the SIRPα binding site. Disruption of QPCTL reduces binding of CD47 to SIRPα, enhancing ADCP and ADCC.¹⁵

While CD47 is widely expressed among normal human cells (including hematopoietic cells), SIRPα/CD47 blockade preferentially results in phagocytosis of tumor, as normal human cells typically lack stimulatory “eat me” signals.^4,15 A notable exception is red blood cells (RBC), which are subject to dose-dependent phagocytosis in the setting of SIRPα/CD47 blockade.¹⁸ Strategies to minimize anemia include administering a priming dose of the SIRPα-CD47 blocker (resulting in RBC phagocytosis then reticulocytosis, which express higher levels of CD47 and are thus resistant to phagocytosis) or selecting the Fc domain of the anti-CD47 monoclonal antibody (mAb) to minimize interactions with the phagocyte Fcγ receptor (either by using an inactive Fc domain or using an Fc domain of IgG2/4, which more weakly interact with Fcγ receptors than do IgG1/3).^19,20

ALX148

ALX148 (ALX Oncology) is an engineered fusion protein containing two high affinity CD47 binding domains of SIRPα linked to an inactive Fc region of human immunoglobulin.²¹ ALX148 is being evaluated in combination with anti-cancer antibodies and/or with anti-PD-1 agents in both solid (with development focused on non-small cell lung cancer (NSCLC), head/neck squamous cell carcinoma (HNSCC), and gastric/gastroesophageal junction carcinoma) and hematologic malignancies. Phase I efficacy data is encouraging, and treatment appears to be tolerable (including minimal hematologic toxicity).

In a phase I study, 25 patients with advanced refractory solid tumors or non-Hodgkin lymphoma were treated with ALX148 alone (NCT03013218).²² ORR was 0% and DCR was 16%. The most common TRAEs were headache (16%), fatigue (12%), dizziness (8%), rash (8%), and thrombocytopenia (8%). Grade ≥3 TRAEs were infection, pancreatitis, thrombocytopenia, and neutropenia (1 each).

Seventy-nine patients with advanced refractory solid tumors were treated with ALX148 in combination with pembrolizumab (NSCLC or HNSCC, N=50) or in combination with trastuzumab (HER2⁺, primarily gastric/gastroesophageal junction carcinoma, N=29) (NCT03013218).²³ NSCLC patients previously received pembrolizumab or had PD-L1 tumor proportion score <50%, HNSCC patients had progressed on platinum-based therapy, and HER2⁺ patients had previously received trastuzumab. ORR and DCR, respectively, were 4% and 39% for NSCLC, 18% and 41% for HNSCC, and 19% and 48% for gastric/gastroesophageal junction carcinoma. Treatment was well-tolerated. Any grade TRAEs included fatigue (11%), AST increase (9%), ALT increase (8%), anemia (8%), and thrombocytopenia (6%).

An overlapping patient cohort was subsequently presented (NCT03013218).²⁴ Patients included 52 with HNSCC that had progressed on platinum-based therapy (with or without prior anti-PD-1 therapy) and were treated with ALX148 and pembrolizumab, one with previously untreated HNSCC treated with ALX148, pembrolizumab, and platinum doublet, and 30 with HER2⁺ gastric/gastroesophageal junction cancer that had progressed on platinum doublet with trastuzumab and were treated with ALX148 and trastuzumab with or without ramucirumab/paclitaxel. ORR, median progression-free survival (PFS, months), and median overall survival (OS, months) were 40%, 4.6, and not reached in the previously treated anti-PD-1 naïve HNSCC arm treated with ALX148/pembrolizumab; 0%, 2.0, and 7.4 in the anti-PD-1 experienced HNSCC arm treated with ALX148/pembrolizumab; and 20%, 2.2, and 8.1 in the gastric/gastroesophageal junction cancer arm treated with ALX148/trastuzumab. TRAEs in the arms treated with ALX148 and either pembrolizumab or trastuzumab were fatigue (18%), AST increase (11%), thrombocytopenia (10%), ALT increase 9%), anemia (9%), and pruritus (9%).

Magrolimab (Hu5F9-G4, ONO-7913)

Magrolimab (Gilead, recently acquired from Forty Seven) is an anti-CD47 humanized IgG4 mAb.¹⁸ It is being evaluated in solid tumors and hematologic malignancies. Anemia is common but can be mitigated by administration of a priming dose of magrolimab with intra-patient dose-escalation.²⁰ Thrombocytopenia and leukopenia are usually not severe. There are many ongoing phase I-III trials evaluating magrolimab in both solid and hematologic cancers (Table 2).

In a phase I study, 60 patients with refractory advanced solid tumors (plus 2 patients with diffuse large B cell lymphoma) were treated with magrolimab (NCT02216409).²⁰ ORR was 3% (ovarian and fallopian tube carcinoma). The most common hematologic toxicities were anemia (57%, typically with a hemoglobin decrease of 1–2 g/dL after the priming dose with rapid subsequent improvement; 4 patients required transfusions), lymphopenia (18%), and thrombocytopenia (11%). The most common non-hematologic toxicities were fatigue (64%), headache (50%), fever and chills (45% each), and hyperbilirubinemia (34%, usually transient unconjugated hyperbilirubinemia associated with anemia during the priming period). Thus, efficacy of magrolimab monotherapy appears to be limited.

In a subsequent phase Ib study, 34 patients with refractory solid tumors (primarily ovarian cancer) were treated with magrolimab and avelumab (NCT03558139).²⁵ Among the 18 response-evaluable ovarian cancer patients, ORR was 0% and DCR was 56%. TRAEs were similar.

Magrolimab was evaluated with cetuximab in patients with solid tumors (phase Ib, N=32) and previously treated colorectal cancer (CRC; phase II, N=46) (NCT02953782).²⁶ The most common any-grade TRAEs were aceniform rash (36%), dry skin (33%), fatigue (32%), infusion reaction (31%), headache (30%), diarrhea (23%), nausea (23%), chills (23%), and anemia (22%); 4% of patients discontinued treatment due to an AE. Among KRAS wild type CRC (N=30), ORR was 7% (all cetuximab-experienced), mPFS was 3.6 mo, and mOS was 10.1 mo. Among KRAS mutant CRC (N=40), ORR was 0%, DCR was 45%, mPFS was 1.9 mo, and mOS was 10.4 mo. Tumor biopsies showed increases in tumor infiltrating macrophages.

SRF231

SRF231 (Surface Oncology) is an anti-CD47 fully human IgG4 mAb.^27–29 In pre-clinical models, SRF231 binds to RBC CD47 but does not cause RBC phagocytosis.²⁷ It is being developed in both solid and hematologic malignancies. Among 37-response evaluable patients with advanced refractory solid (N=45) and hematologic malignancies (N=1) treated with SRF231, ORR was 0% with SD reported (phase I, NCT03512340).³⁰ The most common TRAEs were fatigue (43%), headache (35%), and fever (30%).

Lemzoparlimab (TJC4, TJ011133)

Lemzoparlimab (AbbVie and I-Mab Biopharma) is an anti-CD47 fully human IgG4 mAb with a novel conformational epitope, resulting in minimal binding to human RBCs and platelets in pre-clinical models.³¹ It is being evaluated in both and solid tumors. A phase I trial of lemzoparlimab in patients with solid tumors and lymphoma is ongoing (NCT03934814). Thus far, 20 patients with advanced refractory solid tumors were treated with lemzoparlimab in dose escalation (part 1).³² The most common any grade TRAEs were anemia (30%, primarily transiently during cycle 1 with average reduction in hemoglobin 1.5 g/dL), fatigue (25%), infusion reaction (20%), and diarrhea (15%); no grade ≥3 TRAEs were reported. Part 2 is planned, which will evaluate lemzoparlimab with pembrolizumab or rituximab.

IBI188

IBI188 (Innovent Biologics) is an anti-CD47 humanized IgG4 mAb that is being evaluated in both solid and hematologic malignancies.¹⁰ Among 20 patients with refractory/advanced solid tumors or lymphoma treated with IBI188, the most common any-grade TRAEs were nausea (35%), back pain (35%), fatigue (30%), vomiting (20%), elevated bilirubin (20%), and anemia (15%). Grade ≥3 TRAEs were elevated bilirubin, thrombocytopenia, and anemia (5% each) (NCT03763149).³³

LILRB1/MHC-I

Leukocyte immunoglobulin-like receptor B1 (LILRB1) is expressed on both innate (monocytes, macrophages, eosinophils, basophils, DCs, and certain NK cells) and adaptive (certain T and B cells) immune cells.⁴ MHC-I, expressed on nucleated cells (including cancer cells), is a heterodimer composed of a heavy α-chain and a β2-microglobulin chain.^4,34 Antigen-bound MHC-I on cancer cells is recognized by cytotoxic T cells, and engagement of MHC-I by the T cell receptor and CD8 results in cytotoxicity.^4,34 However, expression of MHC-I on cancer cells has been correlated with resistance to phagocytosis, likely due to an inhibitory interaction between LILRB1 on phagocytes and the β2-microglobulin subunit of MHC-I on cancer cells.^4,35 While cytotoxic T cell anti-tumor activity is dependent upon intact interaction with MHC-I, specifically blocking the LILRB1/β2-microglobulin axis is a possible innate immune drug target. There are currently no active clinical trials targeting this checkpoint, which may be explained by 1) the relatively recent discovery that the LILRB1/MHC-I interaction correlates with resistance to phagocytosis in cancer cells (in 2018)⁴ and 2) drugs cannot interfere with the MHC-I/T cell receptor interaction).

LILRB2/MHC-I

LILRB2 is also expressed on both innate (monocytes, macrophages, basophils, and DCs) and adaptive (CD4⁺ T cells) immune cells and interacts with MHC-I on nucleated cells (unknown specific ligand).⁴ LILRB2 is also referred to as ILT4, LIR2, monocyte/macrophage immunoglobulin-like receptor 10, MIR-10, and CD85d. It has been shown pre-clinically that LILRB2 antagonism promotes macrophage maturation and activation.³⁶

MK-4830

MK-4830 (Merck and Agenus) is an anti-LILRB2 fully human IgG4 monoclonal antibody.³⁷ In a phase I study, 84 patients with advanced refractory solid tumors were treated with MK-4830 alone (N=50) or in combination with pembrolizumab (N=34) (NCT03564691).³⁸ Among all patients, ORR was 13%; most (91%) were in the combination group, 45% were anti-PD-1 naïve, and some were durable (>1 year). TRAEs occurred in 52% of patients and most were grade 1–2.

NK cell checkpoint inhibitors

NK cells contain both stimulatory and inhibitory receptors, serving as possible drug targets (Figure 2). NK cell stimulatory receptors include DNAM-1 (also called CD226), NKG2C, NKG2D, and natural cytotoxicity receptors (NKP30, NKP44, and NKP46); these interact with cell-surface and soluble markers of stressed cells, including tumor cells, such as cell-surface NK ligands.^10,39–41 The result is natural cytotoxicity, via secretion of cytokines (IFN-γ, TNFα, and GM-CSF), and dendritic cell recruitment and activation via production of chemoattractants (such as CCL5, CXL1, and CXL2).^10,39,40 Thus, NK cell inhibitory checkpoint inhibitors, like phagocytosis checkpoint inhibitors, strengthen the adaptive immune anti-tumor response. NK cell-mediated recruitment and activation of DCs has been demonstrated in pre-clinical melanoma models and may predict response to anti-PD-1 therapy.^13,42

Healthy cells express low levels of stimulatory ligands and high levels of inhibitory ligands; NK cells identify cancer cells by the opposite pattern, however cancer cells can evade detection by altering ligand expression.⁴³ Intratumoral NK cell activity and infiltration has been correlated with improved outcomes in multiple solid tumor types and stages.^44–49 The primary inhibitory checkpoints between NK cells and tumor cells, respectively, are TIGIT/CD112 + CD155, PVRIG/CD112, KIRs/MHC-I, and NKG2A-CD94/HLA-E (Figure 2). NK cell inhibitory checkpoint inhibitors are frequently being evaluated in combination with anti-PD-(L)1 therapy.

TIGIT/CD112 + CD155

T cell immunoglobulin and ITIM domain (TIGIT) is an ITIM-containing inhibitory receptor expressed on NK cells, and also T cell subsets (CD4⁺ T cells, CD8⁺ T cells, and regulatory T cells).¹⁰ The ligands for TIGIT are CD112 (also called PVRL2 and nectin-2) and CD155 (also called PVR), both of which are expressed on tumor cells.¹⁰ TIGIT competes for binding to CD112 and CD155 with DNAM-1, a stimulatory NK cell receptor.⁴³ In pre-clinical models, TIGIT blockade has been shown to increase anti-tumor NK cell activity and CD8⁺ T cell cytokine production/cytotoxicity; anti-tumor T cell effect was NK-cell dependent.^43,50,51 Pre-clinical data suggests synergy between TIGIT blockade and PD-(L)1 blockade, a strategy implemented in clinical trials.^43,50,52

Tiragolumab (MTIG7192A, RG6058)

Tiragolumab (Genentech) is an anti-TIGIT humanized IgG1 mAb designed to inhibit the interaction of TIGIT with CD155.^52,53 In a phase I study, 73 patients with refractory advanced solid tumors were treated with tiragolumab alone (phase Ia, N=24) or in combination with atezolizumab (phase Ib, N=49) (NCT02794571).⁵² In phase Ia, ORR was 0% and DCR was 17%. In phase Ib, ORR was 6% (N=3, all with PD-L1⁺ tumors, 2 with NSCLC and 1 with HNSCC) and DCR was not reported. Given promising efficacy, an expansion cohort was initiated in PD-L1⁺, checkpoint inhibitor naïve patients. This included 14 patients with NSCLC treated with tiragolumab and atezolizumab, in whom ORR was 50% and DCR was 79%. TRAEs occurred in 67% of patients in phase Ia (most common: fatigue in 38%) and 59% of patients in phase Ib (most common: anemia in 31%). Grade 3–4 TRAEs were uncommon.

In a prospective, randomized, double-blind, placebo controlled phase II study, 135 patients with chemotherapy-naïve, PD-L1⁺, EGFR and ALK wild-type, locally advanced/metastatic NSCLC were randomized 1:1 to atezolizumab with or without tiragolumab (CITYSCAPE, NCT03563716).⁵⁴ After a median follow-up of 10.9 months, tiragolumab + atezolizumab significantly improved ORR (37% vs 21%) and median PFS (5.6 vs 3.9 months). Patients in the tiragolumab arm experienced similar AEs as the placebo arm. TRAEs, grade ≥3 TRAEs, and AEs leading to treatment withdrawal occurred in 81%, 15%, and 8% of patients treated with tiragolumab + atezolizumab and 72%, 19%, and 10% of patients treated with placebo + atezolizumab. Tiragolumab is currently being studied in multiple phase III clinical trials, focused on esophageal squamous cell carcinoma and NSCLC (Table 3).

Vibostolimab (MK-7684)

Vibostolimab (Merck) is an anti-TIGIT IgG1 humanized mAb which is being evaluated in early phase clinical trials in solid tumors, with a focus on NSCLC.⁵⁵ In a phase I study, 34 patients with refractory advanced solid tumors were treated with vibostolimab alone and 47 patients in combination with pembrolizumab (NCT02964013).⁵⁵ In the monotherapy and combination therapy groups, respectively, ORR was 3% and 19% and DCR was 35% and 47%. AEs occurred in 53% (6% grade ≥3) of patients treated with monotherapy and 65% (12% grade ≥3) of patients treated with combination therapy. The most common AEs were fatigue, pruritus, infusion reaction, anemia, nausea, and rash.

A 120 patient NSCLC dose-expansion was initiated (NCT02964013).^56,57 Patients were treated with either vibostolimab alone or in combination with pembrolizumab and analysis was stratified by prior anti-PD-(L)1 therapy. Thus, the cohorts included anti-PD-(L)1 refractory NSCLC treated with vibostolimab alone (N=41) or in combination with pembrolizumab (N=38), and anti-PD-(L)1 naïve refractory NSCLC (N=41). ORR was 7% in anti-PD-(L)1 refractory NSCLC treated with vibostolimab monotherapy, 5% in anti-PD-(L)1 refractory NSCLC treated with vibostolimab in combination with pembrolizumab, and 29% in anti-PD-(L)1 naïve NSCLC treated with vibostolimab in combination with pembrolizumab (with a trend toward higher ORR among patients with PD-L1 Tumor Proportion Score ≥ 1%). Many responses were durable. Across all patients, TRAEs occurred in approximately 70% of patients and most frequently were pruritus, fatigue, rash, fever, arthralgia, and anorexia. Grade 3–4 TRAEs occurred in approximately 15% of patients and included lipase elevation, hypertension, and pneumonitis (the latter after receiving combination therapy and resulted in death).

Etigilimab (OMP-313M32)

Etigilimab (Mereo BioPharma) is an anti-TIGIT IgG1 humanized mAb.⁵⁸ Pre-clinical studies of intra-tumor immune cells demonstrated increased NK cell cytotoxicity and T cell (both CD4⁺ and CD8⁺) activation/infiltration.⁵⁹ However, ORR was 0% and DCR 22% among 18 patients with refractory advanced solid tumors treated with etigilimab in a phase I study reported in 2018 (NCT03119428).⁵⁸ This study was terminated August 2020 and no additional clinical trials are currently registered.

PVRIG/CD112

Poliovirus receptor related immunoglobulin domain containing (PVRIG, also called CD112 receptor) is an inhibitory receptor expressed on NK cells and CD8⁺ T cells which recognizes CD112 but not CD155 on tumor cells.^41,51 Like TIGIT, PVRIG also competes with the NK cell activating receptor DNAM-1 for binding to CD112. In pre-clinical studies, anti-PVRIG therapy has been shown to increase NK cell cytotoxicity and CD8⁺ T cell cytokine production/cytotoxicity.^51,60

COM701

COM701 (Compugen) is an anti-PVRIG humanized IgG4 mAb.^61,62 Twenty-eight patients with advanced refractory solid tumors were treated with COM701 with (N=16) or without (N=12) nivolumab in a phase I study (NCT03667716).^62–64 In the entire population DCR was 57%. No patients stopped treatment due to toxicity. The most common treatment-emergent AEs (TEAE) were fatigue (46%), nausea (31%), and anxiety (23%) in patients with COM701 alone; with combination therapy, 88% of TEAEs were grade 1–2 (anemia, edema, rash, and fatigue most commonly).

KIRs/MHC-I

Killer cell immunoglobulin-like receptors (KIRs) are expressed on most NK cells and a minority of T cells (primarily CD8⁺ memory T cells) and recognize MHC-I on tumor cells.^10,65 There are numerous individual KIRs, with nomenclature determined by the number of extracellular Ig-like domains (“2D” or “3D” following “KIR”) and the size of the cytoplasmic tail (long/“L” or short/“S”).^65,66 KIRs can either be activating or inhibitory, and inhibitory KIRs typically have a long cytoplasmic tail. The inhibitor KIRs include KIR2DL1, KIR2DL2, KIR2DL3, KIR2DL4, KIR2DL5, KIR3DL1, KIR3DL2, and KIR3DL3.^11,39,65,66 Individual KIRs bind HLA class I subgroups.

In fact, all NK cells express at least one inhibitory receptor (either a KIR or NKG2A) and normal tissues express low levels of NK cell activating ligands, allowing for self-tolerance.⁶⁷ Loss or mutation of tumor MHC-I generates an anti-tumor NK cell response by responding to missing self.^11,65 On the contrary, MHC-I molecule upregulation allows cancer cells to evade immune destruction. In 1999 it was shown that KIR-HLA mismatches between donor and recipient during hematopoietic stem cell transplant for AML generated a T cell independent graft-versus-leukemia effect, by preventing the inhibitory interaction between KIRs on donor NK cells and MHC-I on recipient cancer cells.⁶⁸ Hence, the development of many anti-KIR drugs has focused on hematologic malignancies.

Combination approaches include blocking KIRs (or NKG2A-CD94, see below) with the addition of an anti-PD-(L)1 agent, anti CTLA-4 agent, or anti-cancer antibody (the latter increasing NK cell ADCC).⁶⁶ Blockade of CTLA-4 or PD-(L)1 can stimulate cytokine secretion from T cells (such as IL-2), enhancing NK cell function. Conversely, blockade of KIRs can stimulate cytokine secretion from NK cells (such as IFN-γ), enhancing myeloid and T cell anti-tumor function.⁴⁰

Lirilumab (IPH2102, BMS-986015, BMS-986015-01)

Lirilumab (Innate Pharma and Bristol-Myers-Squibb), is a fully human IgG4 mAb directed against a common epitope shared by KIR2D (i.e. anti-KIR2DL1–3 effect).⁶⁹ This is identical to the structure of IPH2101 (previously developed in malignant hematology) except for one mutation introduced into the constant region of the heavy chain to increase yield and prevent half-antibody formation that can occur with human IgG4.⁶⁹ Lirilumab is being evaluated in both solid and hematologic malignancies (leukemia, lymphoma, and multiple myeloma).

In a phase I study, 136 patients with refractory advanced solid tumors were treated with lirilumab in combination with nivolumab (N=136, NCT01714739) or ipilimumab (N=22, NCT01750580).^70,71 TRAEs occurred in approximately 70% of patients in each cohort and most commonly were fatigue, pruritus, infusion reaction, nausea/vomiting, rash, and diarrhea. Grade 3–4 TRAEs occurred in 13% of the nivolumab cohort and 9% of the ipilimumab cohort, including grade 4 thrombocytopenia (N=1), grade 3 pancreatitis (N=1), and grade 3 radiation skin injury (N=1). Overall toxicity in the nivolumab cohort was similar to nivolumab monotherapy, with the exception of manageable infusion reactions. Among 29 response-evaluable patients with platinum-refractory HNSCC, ORR was 24% and DCR was 52%.

A separate phase I trial included patients with both hematologic (N=22) and solid malignancies (breast N=6, ovarian N=7, pancreatic N=1, and endometrial N=1) treated with lirilumab (EudraCT 2009–011526-33).⁶⁹ Among all patients (including those with hematologic malignancy), any grade TRAEs were primarily mild and occurred in 68% of patients. Most commonly these were pruritus (19%), asthenia (16%), fatigue (14%), infusion reaction (14%), and headache (11%). Grade 3–4 TRAEs occurred in 19% of patients and included elevated lipase, lymphopenia, presyncope, elevated bilirubin, abnormal LFTs, urticaria, and angioedema. Among only solid tumor patients, ORR was 60% (4 breast cancer, 4 ovarian cancer, and 1 pancreas cancer). mPFS was not reached for breast cancer and 5.3 months for ovarian cancer. Additional phase I-II trials are currently recruiting patients with solid tumors (Table 3).

NKG2A-CD94/HLA-E

NK group 2 member A (NKG2A) and CD94 form a heterodimeric ITIM-containing (on NKG2A) inhibitory receptor, which is expressed on cytotoxic lymphocytes (NK cells and a subset of CD8⁺ T cells).^10,72–74 NKG2A is the first inhibitory receptor expressed during NK cell development.⁶⁶ After differentiation, NK cells may co-express NKG2A with KIRs or NKG2A may be lost.⁶⁶ The NKG2A-CD94 heterodimer recognizes the nonclassical MHC-I HLA-E, which is upregulated on both hematologic and solid tumors. ^10,39,43,72 In contrast, classical MHC-I molecules are often downregulated on tumor, allowing immune evasion.⁷³ Increased tumor expression of HLA-E has been associated with decreased NK cell cytotoxicity and worse prognosis in both hematologic and solid cancers.^39,43 Synergistic anti-tumor effect with PD-(L)1 blockade has been noted in pre-clinical models, an effect which is dependent on both NK and CD8⁺ T cells.⁷²

Monalizumab (IPH2201)

Monalizumab (Innate Pharma and AstraZeneca) is a humanized IgG4 mAb which binds to the NKG2A-CD94 heterodimeric receptor, blocking its interaction with HLA-E and promoting effector NK and CD8⁺ T cell functions.^39,73 In pre-clinical studies, monalizumab alone promoted NK cell cytotoxicity, monolizumab with durvalumab promoted NK and CD8⁺ T cell effector functions, and monalizumab with cetuximab enhanced NK-cell mediated ADCC.^72,73 Monalizumab is being evaluated in solid tumors (primarily CRC, HNSCC, NSCLC, and gynecologic cancer), often in combination with anti-cancer antibodies and/or anti-PD-(L)1 agents.

Initially, 55 patients with refractory advanced solid tumors (including 40 patients with microsatellite stable (MSS) metastatic CRC in dose-expansion) were treated with monalizumab and durvalumab (NCT02671435).⁷⁵ In the CRC expansion cohort, among 37 response-evaluable patients, ORR was 8% and DCR was 38%. In the entire population, 48% had any grade TRAEs (diarrhea most common) and grade ≥3 TRAEs were rare.

Development in CRC was continued. In a phase I/II study of treatment-naïve MSS mCRC, 18 patients were treated with monalizumab + durvalumab + FOLFOX + bevacizumab and 17 patients were treated with monalizumab + durvalumab + FOLFOX + cetuximab (RAS/BRAF wild-type and left-sided tumor only) (NCT02671435).^76,77 ORR and DCR, respectively, were 41% and 88% in the bevacizumab cohort and 53% and 88% in the cetuximab cohort. Any TEAE occurred in nearly all patients. In the bevacizumab cohort, the most common TEAEs were fatigue, nausea, and peripheral neuropathy (any TEAE grade ≥3 78%). In the cetuximab cohort, the most common TEAEs were peripheral neuropathy, rash, and dermatitis acneiform (any TEAE grade ≥3 71%).

Monalizumab monotherapy was also evaluated in a phase II trial in refractory advanced HNSCC, where 27 patients (100% prior platinum, 59% prior anti-PD-(L)1 agent) (UPSTREAM, NCT03088059).⁷⁸ The site of primary tumor included oral cavity, oropharynx, hypopharynx, and larynx. ORR was 0%, DCR was 22%, median PFS was 7.4 weeks, and median OS was 28 weeks. Grade 3 or higher TEAEs occurred in 59% of patients, however none were attributed to treatment. Pre-specified ORR was not met and the study was closed at interim analysis for futility.

Development in HNSCC was continued, but in combination with cetuximab. In a multicenter phase I/II trial, 40 patients with recurrent or metastatic HNSCC (100% prior platinum, 45% prior anti-PD-(L)1 agent, 13% prior cetuximab, and no more than 2 prior lines of therapy for advanced disease) were treated with monalizumab and cetuximab (NCT02643550).^79–81 ORR was 28% (36% in anti-PD-(L)1-naïve patients and 17% in anti-PD-(L)1-experienced patients). Among all patients, median duration of response was 5.6 months, median PFS was 4.5 months, and median OS was 8.3 months. Treatment was well tolerated. The most common monalizumab-related AEs were fatigue (17%), pyrexia (13%), and headache (10%). Six percent of TRAEs were grade 3–4. In a separate expansion cohort, 40 patients with recurrent or metastatic HNSCC previously treated with both platinum and an anti-PD-(L)1 agent were treated with monalizumab and cetuximab, and early results show ORR 20% (all PR).⁸² A randomized phase III trial of monalizumab and cetuximab is recruiting in this population (Table 3).

Additionally, 58 patients with refractory advanced gynecologic malignancies (platinum-sensitive ovarian, platinum-resistant ovarian, squamous cervical, and epithelial endometrial carcinoma) were treated with monalizumab monotherapy (NCT02459301).⁸³ Anti-tumor efficacy was poor. ORR and DCR, respectively, in the dose ranging cohort were 0% and 41% and in the dose expansion cohort were 0% and 18%. Overall treatment was well tolerated and none of the AEs were felt to be immune mediated. The most common TRAEs were headache, fatigue, nausea/vomiting, and abdominal pain (>15% each). Grade ≥3 TRAEs included lymphopenia (19%), anemia (16%), vomiting (5%), fatigue (3%), and neutropenia (2%).

Discussion

While immune checkpoint inhibitors targeting the PD-(L)1 and CTLA-4 axes have revolutionized cancer treatment, novel immune-based therapies are desperately needed. Innate immune checkpoint inhibitors alone or in combination (with anti-cancer antibodies or adaptive immune checkpoint inhibitors) are a promising novel anti-cancer strategy. These drugs can unleash the innate immune system against tumor, generating phagocytosis and natural cytotoxicity (both of which can be enhanced with the addition of an anti-cancer antibody), and synergistically activate the adaptive immune system via antigen presentation (often combining well with adaptive immune checkpoint inhibitors).

As repeatedly noted, there is considerable overlap in checkpoint expression and effects of checkpoint modulation between the innate and adaptive immune systems. TIM-3 and LAG-3 are inhibitory receptors expressed on both innate and adaptive immune cells.^10,11,84,85 While they serve primary as adaptive immune checkpoints (with blockade resulting in T cell stimulation), there is evidence of innate immune effects.^86–89 There are many areas of ongoing and future research, and studies may elucidate a larger role of these checkpoints in anti-tumor efficacy.

Phagocytosis and NK cell checkpoint inhibitors are currently being evaluated in clinical trials with solid tumor safety and efficacy data summarized here. In specific solid tumor types, several agents have demonstrated promising early efficacy signals (ALX148 in HNSCC and gastric cancer, tiragolumab and vibostolimab in NSCLC, and monalizumab in CRC and HNSCC). Currently, phase III trials are underway for tiragolumab (esophageal SCC and NSCLC) and monalizumab (HNSCC). While clinical development is still early and ongoing monitoring is needed, innate immune checkpoint inhibitors (as monotherapy) do not appear to frequently cause immune-related adverse events, as inhibitors of PD-(L)1 and CTLA-4 do, likely due to low expression of prophagocytic signals on normal tissues.^90,91 Additionally, strategies have been developed to mitigate on-target toxicities, such as cytopenias seen with anti-CD47 agents.

The development of innate immune checkpoint inhibitors continues at a rapid pace. Tables 2 & 3 list the active clinical trials on ClinicalTrials.gov and there are many additional compounds in pre-clinical development. The use of novel genetic screening techniques may identify additional cell-surface regulators of innate immune cell function, serving as new therapeutic targets.⁹² Innate immune checkpoint inhibitors are an exciting new therapeutic class in both solid and hematologic malignancies that have the potential to expand the use of immunotherapy in many cancer types, significantly improving patient outcomes.