The metastatic cascade through the lens of therapeutic inhibition

Summary

Metastasis is a main cause of cancer-related death, and a deeper understanding of the metastatic process will inform more targeted and mechanistic approaches that can abrogate challenges in treatment efficacy and toxicity. Several steps throughout the metastatic cascade, from angiogenesis to secondary tumor formation, offer specific vulnerabilities to therapies that can lead to the decline or cessation of metastatic progression. A deeper understanding of the metastatic cascade also allows combination systemic therapies to be used synergistically. In this review, we describe current treatment modalities in the context of multiple steps of the metastatic cascade. We highlight their mechanisms and present their efficacy across multiple cancers. This work also presents targets within the metastatic cascade in need of more research that can advance the landscape of treatments and lead to the goal of metastatic cancer remission.

Graphical abstract

In this review, Miranda et al. highlight new and current therapeutics that impede various processes in the solid tumor metastatic cascade. They also highlight vulnerabilities in steps of the cascade that present as research opportunities in pursuit of targeted therapies.

Introduction

Cancer remains a leading cause of death in the United States (US) and globally, with 9.7 million dying from the disease in 2022.¹ In 2024, 611,720 cancer deaths are projected to occur in the US, alone.² Metastasis, the spreading of cancer cells from the primary organ of origin to another part of the body, is a major cause for cancer-related death. While modern systemic treatment and occasional locoregional treatment have made metastatic cancer more clinically manageable over the last few decades, a long-lasting remission of metastasis remains an elusive goal. A deeper understanding of the metastatic cascade will offer additional targets for therapy that have the potential to yield more favorable outcomes than current treatments.^3,4

Metastasis is the culmination of detachment of cancer cells from the primary tumor and the subsequent growth of those cells in a distant organ. This complex process of coordination and dynamic adaptation on the part of the cancer cells is enabled by significant remodeling of the primary and metastatic tumor microenvironments (TMEs).⁵ The metastatic cascade is highly and inherently inefficient, with an overwhelming majority of cells dying before completing the process.^6,7 The cells that do survive one of the steps, however, can either remain dormant or progress further through the cascade. Though a relatively small percentage of cells manage to advance, those that do are considered disseminated tumor cells (DTCs) and can lead to metastatic disease progression and adverse oncological outcomes.⁸

Each step of the cascade has vulnerabilities that can prohibit advancement to the next step (Figure 1).⁹ Some of these vulnerabilities have been leveraged as treatment targets, others have yet to be explored, and still others have only been identified at the preclinical level. While clinically evident metastasis is often identified after the diagnosis of a primary tumor, for most patients, metastasis has already been initiated at the time of initial diagnosis.^4,10 Currently, systemic therapy is the primary modality of metastatic disease, but this is accompanied by a range of toxicities that can lead to significant treatment-associated morbidities.¹¹ Treatment-related adverse events are diverse and extremely relevant when considering improvements to cancer outcomes.¹² Identifying targeted therapies that exploit unique vulnerabilities of the metastatic cascade has the potential to reduce the burden of cancer outcomes. In addition, therapies effective at one stage in a patient’s treatment may become ineffective in the next because of acquired resistance, but targeting interconnected pathways shows promise in preventing or overcoming this resistance.¹³ As such, it is important to put the metastatic cascade in perspective to identify the most suitable interventions for safer, more effective treatment approaches and to highlight novel therapeutic targets (Table 1).¹⁴

Figure 1.

Metastatic and dissemination cascades of solid tumors

Solid tumors enter the metastatic cascade and reach a distant tissue for secondary tumor formation. The tumor cells must undergo a series of steps to complete this process, but each step contains a vulnerability that may be exploited by various therapeutics. Inhibition using these therapeutics could lead to the cessation of the metastatic cascade.

Table 1.

Therapies targeting the metastatic cascade

Drug name	Target	Drug type
Angiogenesis
Bevacizumab	VEGF	mAb
Ziv-Aflibercept	VEGFA, VEGFB	fusion protein
Ramucirumab	VEGF-R2	mAb
Sorafenib	VEGF-R	small molecule
Pazopanib	VEGF-R	small molecule
Lenvatinib	VEGF-R	small molecule
Tivozanib	VEGF-R	small molecule
Ponatinib	VEGF-R	small molecule
Sunitinib	VEGF-R	small molecule
Regorafenib	VEGF-R	small molecule
Vandetanib	VEGF-R	small molecule
Epithelial mesenchymal transition
Cetuximab	EGFR	mAb
Erlotinib	EGFR	small molecule
Panitumumab	EGFR	mAb
NP137	Netrin	mAb
Pritumumab	Vimentin	mAb
Migration and invasion
AT13148	Rho	small molecule
Fasudil	Rho kinase	small molecule
Ketorolac	Rac1	small molecule
Ulixertinib	ERK1/2	small molecule
Trametinib	MEK	small molecule
Cobimetinib	MEK	small molecule
Binimetinib	MEK	small molecule
Intravasation
Rebastinib	TIE2	small molecule
PLX7486	CSF1R	small molecule
ARRY-382	CSF1R	small molecule
RG7155	CSF1R	mAb
FPA008	CSF1R	mAb
Pexidartinib	CSF1R	small molecule
Immune invasion
Lirilumab	NK cell receptor	mAb
Monalizumab	NK cell receptor	mAb
AZD5069	CXCR2	small molecule
Pulmozyme	NET	DNase
Disulfiram	aldehyde dehydrogenase	small molecule
Pembrolizumab	PD1	mAb
Sintilimab	PD1	mAb
Nivolumab	PD1	mAb
Toripalimab	PD1	mAb
Atezolizumab	PD-L1	mAb
Durvalumab	PD-L1	mAb
Avelumab	PD-L1	mAb
Ipilimumab	CTLA4	mAb
Tremelimumab	CTLA4	mAb
Colonization
Fresolimumab	TGF-β	mAb
Trabedersen	TGF-β	antisense oligonucleotide
Vigil	TGF-β	vaccine
Plerixafor	CXCR4	small molecule
Denosumab	RANKL	mAb
Zoledronic acid	osteoclasts	bisphosphonate
Clodronate	osteoclasts	bisphosphonate
Pamidronate	osteoclasts	bisphosphonate
Uproleselan	E-selectin	small molecule
Prosaposin	Thrombospondin 1	short peptide
Abemaciclib	CDK4/6	small molecule
Anoikis resistance
Defactinib	FAK	small molecule
Ifebemtinib	FAK	small molecule
PND-1186	FAK	small molecule
GSK2256098	FAK	small molecule
Etaracizumab	αvβ3	mAb
Cilengitide	integrins	small molecule
Conteltinib	integrins	small molecule
Extravasation
Heparin	P-selectin	polysaccharide
Eptifibatide	αIIbβ3	small molecule
Tirofiban	αIIbβ3	small molecule
Abciximab	αIIbβ3	mAb

Data are correct as of October 19, 2024; searches done by Ian Miranda.

In this review, we describe current treatment modalities in the context of multiple steps of the metastatic cascade, highlight their mechanisms, and present their relevance across multiple tumor types.

Angiogenesis

Angiogenesis, the formation of new blood vessels, occurs physiologically during normal development and wound healing. However, solid tumors are reliant upon aberrant angiogenesis as a means of supporting their rapid growth.^15,16 The onset of hypoxic conditions in a rapidly growing tumor mass activates molecular reprogramming by hypoxia inducible factor 1α (HIF-1α), which upregulates vascular endothelial growth factor (VEGF) and VEGF receptor (VEGF-R).^16,17 In an equilibrium between angiogenic activators and inhibitors, VEGF is a key player in tipping the scales toward the stimulation and growth of new vessels.¹⁸ Tumor cells can also over-activate the physiological angiogenic process, leading to the aberrant, pathological vascularization seen in solid tumors. These newly formed blood vessels are often dysfunctional and “leaky”¹⁹ due to increased permeability caused by promotion of nitric oxide synthase signaling by VEGF. Offering a more robust blood supply, neovascularization supports the proliferation of tumors and the possibility for metastasis.²⁰ VEGF binding causes dimerization of extracellular domain regions of VEGF-R, inducing a signaling cascade that, via the phosphorylation of phospholipase C, triggers pathways such as the extracellular signal-related kinase (ERK) signaling pathway to provoke angiogenesis.^21,22 Different types of VEGF-R serve distinct functions. While VEGF-R1 serves as a decoy to sequester VEGF and prevents excessive binding and activation of VEGF-R2, the more powerful activator of angiogenesis, VEGF3 is highly expressed in the endothelium of blood and lymphatic vessels in development and can reappear, for instance, in the setting of angiogenesis in the retina.²² These receptors have been targeted for inhibition in a number of ways (Figure 2).

Figure 2.

Tumor conditions lead to aberrant angiogenesis

As solid tumors progress, they may outgrow their existing vasculature, leading to hypoxic conditions. Hypoxia activates HIF-1⍺, which upregulates VEGF production and releases in the microenvironment for the induction of angiogenesis and creation of leaky vasculature. Various therapeutics that include mAbs and TKIs can target and inhibit proteins vital to this pathway, thus slowing the angiogenic process and mitigating the tumor’s ability to metastasize.

The first report of an antibody that inhibits tumor angiogenesis was published in 1993: a monoclonal antibody (mAb) that inhibited tumor growth in mouse xenograft models.²³ A humanized version of this mAb was created—bevacizumab—and was later approved for use in treatment of metastatic colorectal cancer, glioblastoma multiforme, renal cell carcinoma, ovarian cancer, hepatocellular cancer,²⁴ and others. Other indications for bevacizumab include cancers such as non-squamous non-small cell lung cancer, metastatic colorectal cancer, metastatic renal cell carcinoma, stage IV ovarian cancer, and hepatocellular carcinoma.²⁵ However, bevacizumab did not show any meaningful improvement of distant recurrence-free survival in the adjuvant treatment of colon and breast cancer patients.^26,27 Bevacizumab acts by directly binding to and neutralizing VEGF to prevent its binding to VEGF-R.¹⁸ Ziv-aflibercept is another anti-VEGF therapy and is a fusion protein consisting of the constant region of human immunoglobulin G1 (IgG1) fused with the extracellular domain regions of VEGF-R that bind to VEGF isoforms VEGFA and VEGFB. This binding blocks their activity and reduces downstream signaling, leading to decreased angiogenesis.²⁸

Ramucirumab is an IgG1 mAb that binds to VEGF-R2 to prevent VEGF-ligand binding.²⁹ This therapy is approved by the Food and Drug Administration to treat a number of metastatic tumors, including colorectal cancer, non-small cell lung cancer, gastric adenocarcinoma, and gastroesophageal junction adenocarcinoma.³⁰

VEGF-R can also be targeted by small-molecule inhibitors called VEGF tyrosine kinase inhibitors (VEGF-TKIs), many of which target the different homologs of VEGF-R. This class of drugs seems to be particularly useful in treating metastatic renal carcinomas (RCCs), since metastatic RCC is often accompanied by a mutation in the von Hippel-Lindau gene that leads to overexpression of HIF proteins and a subsequent release of VEGF.^31,32 VEGF-TKIs also find use in treatment of other cancers: for example, sorafenib is also used to treat unresectable hepatocellular carcinoma and thyroid cancer,³³ and pazopanib is used to treat soft tissue sarcoma.³⁴ The development of resistance to these therapies highlighted the need for next-generation VEGF-TKIs. Second- and third-line drugs, such as lenvatinib and tivozanib, can avoid this resistance.³² These therapies, along with several first-line therapies, form a class of drugs known as multikinase inhibitors that, as the name suggests, hinder multiple kinase-related pathways that are involved in many diseases, including tumorigenesis and/or metastasis.³⁵ In addition to the ones named, this class of next-generation VEGF-TKIs includes ponatinib, sunitinib, regorafenib, and vandetanib.³⁶

A major drawback, counterintuitively, to the long-term use of anti-angiogenesis treatment is the risk of increased local and distant metastasis. This is due to hypoxia-induced changes in the tumor during treatment as well as revascularization after the cessation of treatment.³⁷ The drug-induced hypoxic environments can lead to an upregulation in HIF-1α, which is shown to drive an epithelial-mesenchymal transition (EMT; discussed in the following section) and promote metastasis. Some studies show that anti-angiogenic treatment may also lead to a compensatory rise in pro-angiogenic molecules alternative to the ones that are being targeted by the treatment.³⁸

Although anti-angiogenic therapies are effective in treating established metastasis, they showed minimal to no efficacy in the adjuvant settings to prevent metastatic recurrence or improve overall survival.^39,40,41

EMT

Following angiogenesis, some tumor cells undergo EMT and morph from an epithelial phenotype to a more mesenchymal phenotype, a process that occurs physiologically in embryonic development and wound healing. In addition to apicobasal and lateral polarity, epithelial cells are held together by cell-to-cell interactions via several junction proteins and desmosomes. Epithelial cells express proteins such as E-cadherin, claudins, and occludins that maintain cell polarity. The transition to a mesenchymal-like phenotype is accompanied by a gain in the ability to migrate and invade into neighboring tissues, enabled by cytoskeletal actin and vimentin. The process is regulated by transcription factors (TFs) such as Twist, Snail, Slug, and FOXC2 that reduce the expression of epithelial genes, such as E-cadherin, induce the expression of mesenchymal-related genes, such as N-cadherin and vimentin, and upregulate signaling pathways including NOTCH, Wnt, and transforming growth factor β (TGF-β).^11,42 EMT is a highly dynamic and reversible process, and tumors exist on a continuum between epithelial and mesenchymal states, a byproduct of which is resistance to host defenses and therapeutics. Mobilized tumor cells often skew to mesenchymal states, and those that undergo the reverse of EMT—mesenchymal-epithelial transition—are likely able to seed at the metastatic site. However, despite advancements in the understanding of molecular pathways, therapy resistance, and other mechanisms involved with EMT, there are no specific EMT-targeting therapeutic agents.⁴³

TGF-β is one of the main inducers of EMT; thus, inhibiting this pathway can impede EMT. Receptor tyrosine kinases (RTKs) engage in crosstalk with TGF-β, and inhibition of these RTKs has been fruitful in inhibiting EMT and decreasing tumor progression. Inhibitors of one type of RTK implicated in EMT, epidermal growth factor receptor (EGFR), include cetuximab, erlotinib, and panitumumab.⁴⁴ Netrin-1, a secreted laminin-associated protein, has also been shown to be involved in inducing EMT through activation of the phosphoinositide 3 kinase pathway, which is upregulated in various cancers.^45,46 Downregulation of netrin-1 abrogated the effects of TGF-β-induced EMT.⁴⁵ In phase 1 clinical trial testing,⁴⁷ the netrin-targeting mAb NP137 showed reduction in lesions in over half of the clinical trial participants after 6 months. An analysis of these patients’ biopsy samples revealed a significant decrease in EMT-related features, including a decrease in tumor-expressed vimentin. The effectiveness of NP137 in advanced and metastatic solid tumors evokes the clinical relevance of EMT and its role in enabling metastatic potential.⁴⁸

The previously mentioned EMT transcription factors (EMT-TFs), such as the ZEB, SNAIL, and TWIST family of proteins, upregulate EMT and, therefore, would theoretically be ideal targets for therapy. However, EMT-TFs are involved in highly interconnected pathways with redundancies, complementary functions, and feedback mechanisms, leading to difficulties in registering an effect when targeting one TF alone.⁴⁴

Tumor cells that have transitioned to the mesenchymal phenotype can also be targeted. Mesenchymal cells are marked by the surface protein vimentin, which is important for cell migration. Pritumumab is a mAb therapy against vimentin that is currently in phase 2 clinical trials for brain cancers. While the efficacy of Pritumumab is still being tested clinically, preliminary studies from Japan indicate that the drug will likely be an effective therapeutic.^49,50,51,52

Migration and invasion

One of the first steps in distant metastasis is the intrusion of the primary tumor cells into the surrounding stroma via migration and invasion.⁵³ Invasion involves breach of the basement membrane that surrounds the primary tumor. The invading cells detach from the primary tumor in one of two ways: single cells or clusters. In collective invasion, a cluster of cells undergoes EMT, migrates away from the primary tumor, and breaches the basement membrane.⁵⁴ Though they undergo EMT, the collectively migrating cells do not completely lose their epithelial phenotype. In fact, they are considered to enter a hybrid-EMT phenotype whereby they maintain their E-cadherin expression and cell-to-cell adhesion but are also able to migrate through surrounding tissue.⁵⁵

The cluster of migrating and invading cells is driven by leader cells that are trailed by follower cells. The cytoskeletal organization of the leader cells is rearranged to facilitate membrane protrusion. The Rho signaling pathway has been heavily implicated in this cytoskeletal rearrangement, yet there are few clinical therapeutics that target Rho.⁵⁶ The drug AT13148 advanced to clinical trials as a potential therapeutic Rho inhibitor, but it did not progress to phase 2 because of its narrow therapeutic window.⁵⁷ Fasudil, a Rho kinase inhibitor used for cerebral vasospasms, has shown some effectiveness in treating cancers and reducing migration, but clinical trials for cancer have yet to be initiated.⁵⁸ Leader cells also upregulate cytoskeletal components such as Rac. Targeting Rac with ketorolac was shown to reduce the metastatic burden in preclinical models, presumably due its ability to decrease invasion. Clinical trials are underway to measure its efficacy in ovarian cancer in human patients.⁵⁶

In contrast to collective migration and invasion, single cells have decreased metastatic potential.⁵⁹ Migrating and invading single cells must rely on their own individual cellular processes to overcome the many obstacles to metastasis. They cannot benefit from the interconnected network of communication and specialization that is observed among the leader and follower cells of collective migration.⁵⁴ Single cells that survive either find paths via gaps in the extracellular matrix (ECM) or secrete matrix metalloproteases (MMPs) to breakdown or remodel the ECM.⁶⁰

MMPs are integral to the invasive process for both collective and individually invading cells.⁶¹ However, creating clinically feasible inhibitors of this large class of proteins has proven difficult. Various toxicities and issues with bioavailability have prevented MMP inhibitors from successfully completing clinical trials. There has been some recent interest in the use of nanoparticles as potential regulators of MMPs, although these are still in the early research stages.⁶²

A slightly more attractive and achievable target for therapeutics is the ERK family of proteins in the mitogen-activated protein kinase (MAPK) cascade. ERK upregulates expression of MMPs and induces reorganization of the cytoskeleton, increasing the capacity for migration and invasion.⁶³ Ulixertinib, an ERK1/2 inhibitor currently in clinical trials, has shown benefits in reducing the tumor burden.⁶⁴ MAP/ERK kinase (MEK) is an activator of the MAPK pathway. MEK inhibitors trametinib, cobimetinib, and binimetinib have demonstrated anti-tumor activity. Trametinib, when used either alone or in combination with dabrafenib, was shown to increase overall survival in patients with metastatic melanoma and with low-grade gliomas.^65,66,67 B-type Raf kinase (BRAF) is a potent MAPK pathway activator.⁶⁸ When BRAF was inhibited in combination with cobimetinib and vemurafenib, progression-free survival increased in patients with BRAF-mutated metastatic melanoma.⁶⁹ A combination therapy using binimetinib and encorafenib, a BRAF inhibitor, showed similar results.⁷⁰

Intravasation

Intravasation is facilitated by the “leakiness” of the neo-angiogenic blood vessels that are more permeable to cells.^19,71 Cells can enter the vascular endothelium and into circulation either as single cells or as clusters,⁷² and through either ameboid invasion or disruption of the endothelial junctions. For whichever method used, macrophages, specifically tumor-associated macrophages (TAMs), have shown to be extremely important to the success of intravasation.⁷³

TAMs secrete epidermal growth factor (EGF) and attract EGFR-expressing tumor cells to the perivascular region. Tumor cells in turn express colony stimulating factor 1 (CSF1) and attract more TAMs, thus, creating a positive feedback loop.⁷³ As a result, the combination of endothelial cells, TAMs, and tumor cells sets up the TME of metastasis (TMEM). Some of the TAMs in the perivascular TMEM that robustly express the angiopoietin-1 receptor TIE2 (tyrosine kinase with immunoglobulin-like loops and EGF homology domains-2) secrete VEGFA, which causes transient vascular permeability in the local environment thereby facilitating tumor invasion.⁷⁴ TIE2, therefore, serves as a therapeutic target because of its central role in intravasation. To this end, rebastinib was created as an orally administered, potent inhibitor of TIE2. In a phase 1b/2 clinical trial, the combination of rebastinib with the chemotherapeutic paclitaxel demonstrated an adequate safety profile as well as effective anti-tumor activity.⁷⁵ The recruitment of TAMs to the TMEM via the described EGF-CSF1 loop can also be inhibited by small molecules, such as PLX7486 and ARRY-382, and monoclonal antibodies, such as RG7155 and FPA008, that are in their very early stages of investigation. They act by targeting the receptor for CSF1 on TAMs, leading to TAM depletion and a probable decrease in tumor cell intravasation.⁷⁶ Despite the potential for positive effects on metastatic cancers, inhibitors of the CSF1/CSF1R pathway have been limited in their use. Pexidartinib, an inhibitor of CSF1R, for example, has so far shown efficacy only in tenosynovial giant cell tumors.⁷⁷

In addition to increasing the intravasation capacity of tumor cells indirectly, TAMs in the perivascular region and TMEM also directly increase intravasation. Upon physical contact with TAMs, human triple-negative breast cancer cells (TNBCs) showed an increase in the RhoA signaling pathway leading to the development of invadopodia.⁷⁸ However, as previously discussed, inhibition of Rho signaling has proved challenging.^56,57 Additional studies are needed to increase the breadth and effectiveness of these therapies, which are stifled by the heterogeneity of cancer and cancer types.

Immune system evasion

Tumor cells must evade the body’s immune system in order to reach distant sites.⁷⁹ Tumors and circulating tumor cells (CTCs) express neoantigens as a result of mutations that occur during tumor progression, and these neoantigens can help drive the immune system’s anti-tumor activity. However, tumors are able to evade immune cells.⁸⁰ Therapeutically targeting this evasion, therefore, augments the effectiveness of the immune system.

Natural killer cells

Natural killer (NK) cells recognize tumor cells that have downregulated HLA1 on the cell surface in an effort to evade the immune system.⁸⁰ This evasion starts with EMT, where mesenchymal tumor cells display immunosuppressive markers, like programmed death ligand 1 (PD-L1), and low levels of immune activators, such as HLA1.⁸¹ CTCs can evade NK cells by maintaining or even upregulating HLA1, while downregulating NK-activating molecules, such as NK group 2D ligand (NKG2DL). NKG2DL is the ligand for NKG2D on NK cells that is needed for their activation. Therefore, downregulation of this ligand on tumor cells leads to reduced activation of NK cells and increased metastasis.^82,83 Conventional chemotherapies, such as cyclophosphamide, gemcitabine, paclitaxel, and 5-fluorouracil, increase the recruitment and efficacy of NK cells by elevating the NK-activating ligands on the surface of metastatic and malignant cells.⁸⁴ In addition, mAbs such as trastuzumab, rituximab, and cetuximab, which target surface antigens, enhance antibody-dependent cellular cytotoxicity (ADCC) by NK cells in addition to their direct effects on cancerous cells.⁸⁵ Lirilumab, a mAb against killer cell immunoglobulin-like receptor (KIR), and monalizumab, a humanized mAb targeting NKG2A, increase ADCC by NK cells via their blockade of NK cell surface receptors.^86,87 There are a few therapies in the preclinical stages that enhance NK cell function. For example, an antibody against CD96—a negative regulator of NK cells—and antibodies against ligands for NKG2D that lead to NKG2D internalization demonstrate effectiveness in decreasing NK cell suppression to allow for increased NK cell-mediated tumor cell killing.^85,88,89

Neutrophils

Neutrophils, unlike NK cells, assist in the metastasis of CTCs, both directly and indirectly. Neutrophils can directly bind to CTCs, thereby protecting the CTCs from shear force in the circulation and enable their extravasation by mediating binding to the vascular endothelium. Neutrophils can also increase metastatic ability via the formation of neutrophil extracellular traps (NETs) that capture and shield CTCs from immune surveillance.⁹⁰

The CXCR2 chemokine receptor on neutrophils orchestrates neutrophil recruitment to the tumor. AZD5069 is an inhibitor of CXCR2 and has been shown in clinical trials to prolong patient survival and reduce metastasis.⁹¹ Inhibitors for other relevant chemokines are still in preclinical testing.⁹¹ There are currently two methods for direct NET inhibition, but their use in cancer would be considered off-label. Pulmozyme is an inhaled DNase 1 that is currently being used to treat cystic fibrosis. This current formulation localizes to the lungs and, as such, lacks systemic effects, thereby limiting its use in patients with metastasis. Disulfiram is an aldehyde dehydrogenase inhibitor used for substance abuse disorders. It can block gasmedrin D, which is required for NET release, but the interactions with aldehyde dehydrogenase make its use in patients with metastatic disease somewhat challenging.⁹²

T cells

T cells are important in controlling metastasis, and the abundance of CD8⁺ T cells is an important prognostic marker.⁹³ Tumor cells avoid T cell immunity via the expression of PD-L1 and cytotoxic T lymphocyte antigen 4 (CTLA4). PD-L1 is expressed physiologically by vascular endothelial cells, pancreatic islet cells, T cell subsets including regulatory T cells, and antigen-presenting cells. PD-L1 reduces T cell activity by binding to programmed death protein 1 (PD1) on T cells. Tumors often hijack this system and express PD-L1 themselves. Likewise, CTLA4 on tumor cells competes with the T cell costimulatory molecule CD28 and mitigates its binding to CD80 and CD86, both T cell activators, leading to negative regulation of the T cell.^94,95 Immune checkpoint inhibitors combat both of these tumor escape mechanisms. PD1 is targeted by therapies such as pembrolizumab, sintilimab, nivolumab, and toripalimab that all act by binding to PD1. PD-L1 can be inactivated by therapies such as atezolizumab, durvalumab, and avelumab. These are often combined with radiotherapy, systemic chemotherapy, and angiogenesis inhibitors in order to attack the tumor and metastatic cells from multiple vantage points.⁹⁶ CTLA4, for its part, can be inhibited by ipilimumab and tremelimumab. CTLA4-targeting agents can be combined with PD1/PD-L1 inhibitors in order to provide a more robust response.^97,98

Colonization

Malignant cells within the primary tumor establish a pre-metastatic niche at distant sites long before frank metastasis becomes evident radiologically, where mesenchymal and hematopoietic stem cells create a microenvironment conducive for colonization by extravasated CTCs or DTCs.⁹⁹ TGF-β has been shown to be highly involved not only in aiding the formation of the pre-metastatic niche but also in supporting secondary tumor survival.¹⁴ At the pre-metastatic niche, TGF-β signaling in stromal cells helps to remodel the microenvironment to promote tumor progression, and inhibiting TGF-β receptor 2 with a mAb in a murine cancer model led to a decrease in metastasis.¹⁰⁰

Several TGF-β inhibitors have progressed to the clinic. Fresolimumab is a human mAb that can bind to and inhibit different isoforms of TGF-β. Preclinical studies demonstrated its usefulness in reducing tumor growth in vivo, prompting its progress to clinical trials. One clinical trial (NCT01401062) testing a combination treatment of fresolimumab with radiation in patients with metastatic breast cancer seeks to register safety of this combination and its ability to achieve tumor regression.^101,102 The antisense oligonucleotide trabedersen hybridizes to the TGF-β mRNA, causing its degradation and thereby decreasing migration of tumor cells in vitro. This subsequently led to a phase 2b clinical study that showed increased safety and efficacy in patients with high-grade gliomas who received chemotherapy with trabedersen.^101,103 Vaccine technology is also being used to target TGF-β. Vigil is a vaccine targeting furtin convertase, which is involved in processing the TGF-β precursor into activated TGF-β. While not approved yet, Vigil has shown success with metastatic Ewing’s sarcoma and is in an ongoing phase 2 study in patients with high-grade ovarian, fallopian tube, or peritoneal cancer.^101,104

Chemokines are closely involved in the homing and colonization of DTCs to certain tissues.¹⁰⁵ For example, cancer-associated fibroblasts (CAFs) linked to breast cancer have been shown to prime tumor cells to the cytokine CXCL12, whose receptor CXCR4 is expressed by tumor cells. Importantly, CXCL12 is also highly expressed in bone marrow, which can explain metastatic breast cancer’s preference for seeding in this environment.¹⁰⁶ CAFs are also relevant in many other types of cancers: pancreatic, colorectal, head and neck, lung, melanoma, ovarian, bladder, prostate, and cholangiocarcinoma.¹⁰⁷ Plerixafor is an inhibitor of CXCR4 that has shown effectiveness in preclinical studies to reduce metastasis and has completed a phase 2 clinical trial in metastatic pancreatic cancer.^14,108 Due to upregulation of immunosuppressive mechanisms as an unintended effect and a narrow therapeutic window, these therapies need to be further investigated to improve their efficacy in patients.¹⁰⁹

In bone metastasis, the metastatic niche may be primed when tumor cells induce osteoblasts to secrete the receptor activator of nuclear factor κB ligand (RANKL), which activates osteoclasts to degrade surrounding bone mineral. This degradation releases growth factors that prime the metastatic niche for seeding.⁴ Denosumab is a mAb that binds to RANKL and is approved for the treatment of bone metastases. Though the long-term effects of denosumab are still being studied, clinical trials have shown efficacy in this therapy’s ability to reduce skeletal-related events, a measure of adverse events related to bone fragility.¹¹⁰ Likewise, bisphosphonates that also inhibit bone resorption are taken up by osteoclasts and subsequently inhibit osteoclast function through disruption of signaling pathways that promote bone resorption.¹¹¹ Common bisphosphonates include zoledronic acid, clodronate, and pamidronate.¹¹² Further, studies have shown the importance of E-selectin in metastatic seeding of bone. E-selectin, though usually expressed on the surface of endothelial cells, is shown to be highly expressed in the bone marrow and can bind prostate and colon tumor cells.¹¹³ Uproleselan, a small-molecule inhibitor of E-selectin, has been shown decrease bone metastasis in preclinical models.^114,115

Some DTCs are able to proliferate in their metastatic niche and develop into secondary tumors, while other DTCs become dormant. This dormancy can occur when single DTCs become quiescent or when macrometastatic lesions either are targeted by immune cells or outgrow their blood supply.¹¹⁶ Dormant DTCs and micrometastases are widely known to be the major cause of recurrence in cancer patients whose primary tumors have been treated, yet little is known about their mechanisms for becoming and/or remaining dormant.¹¹⁷ Therefore, the treatment and/or inhibition of dormant metastases is the subject of current research that will hopefully see advances in the near future. Currently, there are two well-described strategies for managing these dormant cells. The first is to mobilize them into the bloodstream where they can be inhibited by systemic or other therapies.¹¹⁸ This mobilization can be accomplished by therapies such as plerixafor,¹¹⁹ which is mostly successful in hematological tumors but can hopefully provide an outline for using similar therapies in solid tumor metastases.¹²⁰ On the other hand, mobilizing the DTCs could lead to quicker relapse. Once the DTCs are in circulation, they can develop drug resistance and acquire and inherit mutations, leading to worse prognosis. Therefore, it may be fruitful to explore options that lengthen the time that DTCs are dormant.¹²¹ Dormancy can be maintained by promoting the metastatic niches through the delivery of components beneficial to seeding and colonization. For example, thrombospondin 1 (TSP1) is implicated in the quiescence of DTCs in the lung and bone marrow. Its expression can be induced systemically by the short peptide prosaposin in mouse models, where this therapeutic strategy has shown to reduce lung metastasis.¹²⁰ Clinical dormancy can also be maintained by targeting the DTCs themselves, as is the case in patients with estrogen receptor positive (ER+) breast cancer. Adjuvant endocrine therapy using anti-estrogen significantly reduced distant recurrence and improve overall survival in early-stage breast cancer patients.^122,123 This effect could be further enhanced with extension of endocrine therapy for up to 10 years^124,125 In combination with anti-estrogen therapy, CDK4/6 inhibitors maintain cell senescence by inhibiting cyclin-dependent kinase (CDK) 4/6, important for the transition of the cell cycle from G1 to S phase.¹²⁶ Guided by trials demonstrating significant improvement of distant metastasis-free survival, the CDK4/6 inhibitor abemaciclib is approved for treatment of early-stage high-risk ER-positive and HER2-negative breast cancer.¹²⁷

Future directions

Despite recent advances in understanding the metastatic cascade, there is a need for continued investigations to fully understand nuanced aspects of mechanisms to develop knowledge-guided targeted therapeutics. Here we have described key aspects of the metastatic cascade that represent opportunities for research and drug development.

Anoikis resistance

Anoikis is the programmed cell death of epithelial cells caused by the loss of integrin attachment to the ECM, an attachment that is unavailable while in circulation. Resistance to anoikis is seen almost as a prerequisite to aggressive metastasis. An important mechanism of anoikis is the cleavage of focal adhesion kinase (FAK) that mediates the final steps of cell death. Circulating single cells need to circumvent this process in order to survive.¹²⁸ Clusters of tumor cells in circulation seem to more easily resist anoikis because they maintain some degree of attachment to each other while in circulation.¹²⁹ FAK inhibitors have been the subject of recent clinical trials of combination therapies with immunotherapy, targeted therapy, chemotherapy, or radiotherapy. Defactinib, ifebemtinib, PND-1186, and GSK2256098 are among the small-molecule FAK inhibitors that are being tested in multiple tumor types including pancreatic cancer, metastatic melanoma, non-small cell lung cancer, ovarian cancer, mesothelioma, and gastric cancer.¹³⁰

Cell adhesion molecules (CAMs), such as integrins and selectins, play a vital role in anoikis resistance. Cells adapt their integrin expression in order to survive in their new microenvironment.¹²⁸ For example, during EMT, collagen-binding integrins normally seen in epithelial cells are downregulated in favor of fibronectin-binding integrins that assist in migration. This activates pro-survival signaling in the face of cellular detachment.¹³¹ For example, integrin αvβ3 is implicated in increasing the metastatic potential in a variety of cancer types. Integrin αvβ3 assists with anoikis resistance by activating pro-survival kinases in an FAK-independent manner.¹³² Importantly, integrins such as αvβ3 and others can be targeted therapeutically.¹³¹ One method of targeting integrins is through mAb therapy. Etaracizumab is an αvβ3 mAb and has been evaluated in clinical trials for metastatic melanoma and metastatic prostate cancer, but it failed to register benefit over other therapies in the trials.^133,134,135 Other small-molecule inhibitors of integrins, such as cilengitide and conteltinib, have shown promising results in clinical trials.¹³¹ Cilengitide is a potent αvβ3 (and αvβ5) selective inhibitor that has proven safe and effective in phase 1 and phase 2 clinical trials.^133,136 Though its use in metastatic cancers is limited, combination therapies of cilengitide with cisplatin, 5-fluorouracil, and cetuximab have shown adequate safety in clinical trials for patients with metastatic squamous cell carcinoma of the head and neck.¹³⁷ Conteltinib is being studied in combination with toripalimab and gemcitabine for patients with advanced pancreatic cancer in an ongoing phase 2 clinical trial (NCT05580445).¹³⁸ However, the efficacy of integrin inhibitors relies heavily on their use in combination therapy since these inhibitors, as a class, have not shown much apoptotic potential when used on their own.¹³¹ While the effects of integrin therapies on metastasis are currently limited, mAbs and small molecules targeting αvβ3 should be viewed as an exciting future direction to be pursued.

Extravasation

As CTCs travel through the bloodstream, they may be physically trapped within small capillaries. Here, the CTCs can form transient adhesions to the endothelium by expressing ligands that bind to selectins and neuronal cadherins on endothelial cells. A stronger binding between integrins, such as αvβ3 on CTC surfaces and CAMs on endothelial cells, replaces the transient selectin interactions, setting the stage for migration across the endothelial membrane.¹³⁶ This final step in extravasation of CTCs occurs with the help of invadopodia that permit the CTCs to extend across the endothelial membrane and migrate into the extravascular space¹³⁹ at the secondary site. While there are few reports on extravasation inhibition, therapies to mitigate expression by way of αvβ3 inhibition and other integrins have been covered in the previous section on “anoikis resistance.”

Efficient extravasation is aided by platelets. CTCs aggregate with platelets for protection against shear forces, and this aggregation also helps the CTCs bind to the vascular endothelium. Aggregation occurs via crosslinking when cell receptors on CTCs, such as integrin αvβ3, bind to integrin αIIbβ3 on platelets.¹⁴⁰ These platelets release granule-derived ATP that directly modulates the permeability of the endothelium to allow for CTC extravasation. Importantly, platelets can release chemokines to attract leukocytes that can further modulate permeability. For example, the release of the chemokine CCL2 by platelets attracts monocytes that differentiate into macrophages and subsequently release VEGF, a potent mediator of vascular permeability.⁷⁹

Targeting platelets may, therefore, be a fruitful endeavor in inhibiting extravasation. Currently, there is great interest in repurposing anti-platelet therapies for use in patients with cancer. Heparin has shown to reduce experimental metastases in preclinical trials because of its inhibition of platelet P-selectin, which mediates the binding between platelets and CTCs.¹⁴¹ Other therapies have not yet been explicitly tested in cancer patients but target key mediators in the metastatic cascade. Eptifibatide, tirofiban, and abciximab are inhibitors of the platelet integrin αIIbβ3 and are currently used to treat acute coronary syndrome, unstable angina, ischemic cardiac complications with cardiac procedures, and other platelet-related disorders.¹⁴² In clinical trials, these therapies were relatively safe and had side effects of increased bleeding or thrombocytopenia, though these could be controlled by careful selection of patient population.^143,144,145 Interestingly, a prospective study found that the use of aspirin was associated with a decreased risk of death and distant recurrence in a cohort of breast cancer patients.¹⁴⁶ Though not yet tested in patients with cancer, the efficacy and safety of anti-platelet therapies in hematological disorders indicate their possible success in inhibiting metastasis.¹⁴¹

Conclusion

Once cancer progresses to becoming metastatic, the goal is often to reduce the burden of metastasis for the patient, often by using combination therapies.¹⁴⁷ To be able to inhibit metastasis would mean being able to deliver life-saving care to the 19.2 million people that are diagnosed with cancer every year,¹⁴⁸ but more studies and better insight to relevant treatment options may be needed to achieve this goal. In this review, we have presented therapies that inhibit many steps of the metastatic cascade. We have described current therapies in practice, those in development, and those that need further exploration. Our expectation is that clinicians and scientists can learn of the different options available to treat and prevent cancer spread and that we can work toward the goal of metastasis no longer being the final chapter in a patient’s battle with cancer.

Acknowledgments

The figures in this review were created using BioRender. The authors extend gratitude to the Breast Cancer Research Foundation of Alabama and to METAvivor for their support.

Author contributions

Courtney Swain, Brandon Metge, and Amr Elhamamsy were involved in editing and contributing suggestions for this work.

Declaration of interests

The authors declare no competing interests.