Modulating cancer mechanopathology to restore vascular function and enhance immunotherapy

Summary

Solid tumor pathology, characterized by abnormalities in the tumor microenvironment (TME), challenges therapeutic effectiveness. Mechanical factors, including increased tumor stiffness and accumulation of intratumoral forces, can determine the success of cancer treatments, defining the tumor’s “mechanopathology” profile. These abnormalities cause extensive vascular compression, leading to hypoperfusion and hypoxia. Hypoperfusion hinders drug delivery, while hypoxia creates an unfavorable TME, promoting tumor progression through immunosuppression, heightened metastatic potential, drug resistance, and chaotic angiogenesis. Strategies targeting TME mechanopathology, such as vascular and stroma normalization, hold promise in enhancing cancer therapies with some already advancing to the clinic. Normalization can be achieved using anti-angiogenic agents, mechanotherapeutics, immune checkpoint inhibitors, engineered bacterial therapeutics, metronomic nanomedicine, and ultrasound sonopermeation. Here, we review the methods developed to rectify tumor mechanopathology, which have even led to cures in preclinical models, and discuss their bench-to-bedside translation, including the derivation of biomarkers from tumor mechanopathology for personalized therapy.

Graphical abstract

Solid tumor pathology, characterized by abnormalities in the tumor microenvironment, impairs efficacy of therapy. Mechanical factors shape the tumor’s “mechanopathology” profile and influence treatment success. Mpekris et al. review methods to rectify tumor mechanopathology, achieving preclinical cures, and discuss their clinical translation, including biomarker development for personalized therapy.

Introduction

The mechanopathology of solid tumors

The uncontrollable growth of a tumor in the confined space of a normal host tissue along with intratumoral interactions among cancer cells, stromal cells, and the fibrotic extracellular matrix (ECM) result in the accumulation of mechanical forces and the development of a dense and stiff tumor.^1,2,3,4,5 This phenomenon is particularly evident in desmoplastic cancers, which are characterized by an overproduction of ECM and include subtypes of breast, pancreatic, prostate, and colon cancer and sarcomas. In these cancer types, the stiffening of the tissue and the buildup of mechanical forces often lead to severe compression of intratumoral vessels (Figure 1A). In fact, it has been demonstrated that up to 95% of blood vessels can become compressed or completely collapsed.^3,6,7,8 In addition to compressing intratumoral vessels, mechanical forces exert a direct influence on cancer cells, elevating their apoptotic rate while predominantly diminishing proliferation, inducing apoptosis, and promoting a more invasive and metastatic phenotype.^9,10,11,12 However, it should be noted that in some other studies, it has been shown that high mechanical stresses enhance proliferation of cancer cells.^13,14 Moreover, elevated stiffness and mechanical forces compress tumor cells inducing phenotypic changes and altering gene expression patterns in tumor cells, ultimately leading to the development of resistance to chemotherapy drugs (Figure S1).^10,15,16,17 Compression conditioning of tumor cells can activate mechanosensitive signaling pathways, such as YAP/TAZ and β-catenin, which regulate gene expression programs associated with cell survival, proliferation, and drug resistance.¹⁸ Furthermore, compressive forces can change endothelial cell characteristics and signaling, hindering their capacity to create vascular networks.¹⁹ Mechanotransduction pathways, such as focal adhesion kinase (FAK) and Rho-associated protein kinase (ROCK), can be activated in response to mechanical stress, leading to cytoskeletal remodeling and changes in cell adhesion and migration.^20,21 Also, the disruption of endothelial cell interactions by compressive forces hinders vascular network formation and leads to abnormal blood vessel development or regression.^22,23 Additionally, tumor blood vessels can exhibit hyper-permeability, resulting in excessive fluid leakage from the vascular to the interstitial space of the tumor (Figure 1A).¹ Vessel hyper-permeability is defined as an increase in the pore size within the tumor vessel wall, primarily due to high levels of pro-angiogenic factors, such as vascular endothelial growth factor (VEGF), which drive tumor-induced angiogenesis.²⁴ The newly formed vessels during angiogenesis exhibit structural irregularities characterized by a chaotic vascular organization, limited intercellular connections between endothelial cells, inadequate pericyte coverage, and discontinuous or absent basement membrane.^25,26

Figure 1.

Tumor mechanopathology shapes the tumor vascular system and poses major barriers to treatment efficacy

(A) The abnormal proliferation of cancer cells within the constrained environment of the normal host tissue, along with interactions among stromal cells and the fibrotic ECM, contribute to the mechanopathology of the tumor. The hallmarks of tumor mechanopathology consist of (1) the accumulation of mechanical forces and the ECM stiffening, (2) the chaotic formation of new blood vessels via the process of angiogenesis, and (3) the increased interstitial fluid pressure (IFP) due to vessel compression and abnormal leakiness of the newly formed tumor blood vessels.

(B) Tumor stiffening and accumulation of intratumoral mechanical forces that are exerted on tumor blood vessels cause vessel compression and reduce tumor perfusion. The chaotic nature of angiogenesis contributes to a poorly perfused tumor microenvironment, affecting the overall blood circulation and influencing the responsiveness of the tumor to therapeutic interventions. Hyper-permeability of some tumor blood vessels results in fluid leakage from the vascular to the extravascular space, which increases fluid pressure in the tumors (IFP) and also contributes to hypoperfusion. Impaired blood supply and hypoxia not only reduce drastically drug delivery but also help cancer cells evade the immune system and increase their invasive and metastatic potential. Particularly, hypoperfusion hinders immune cells infiltration into the tumor, while hypoxia renders the TME immunosuppressive and promotes pro-tumor immune responses. Created with BioRender.com.

Excessive fluid leakage across the hyper-permeable vessels along with the compression of lymphatics elevates tumor interstitial fluid pressure (IFP).^27,28 Indeed, IFP can rise to match the pressure within the vessels, effectively erasing the pressure differential across the tumor vessel wall and greatly impeding the transvascular transport of substances (Figure 1A). Adding to this, the abnormal elevation of IFP at the tumor interior in conjunction with the normal IFP values at the interface of the tumor with the normal tissue induces pressure gradients that ooze cells, cytokines, growth factors, small-molecule drugs, and other molecules outside of the tumor contributing further to tumor progression and metastasis.^29,30

Insufficient perfusion and elevated IFP are hallmarks of the mechanopathology of desmoplastic tumors impeding both tissue oxygenation and distribution of systemically administered cellular, nano-sized, and molecular medicines (Figure 1B). Indeed, perfusion in some regions of a tumor can be significantly lower than that in the peri-tumor normal tissue, leading to hypoxia and acidic pH. Hypoxia in turn fosters a hostile tumor microenvironment (TME) that fuels tumor progression through multiple mechanisms, including chaotic angiogenesis, immunosuppression, increased metastatic potential, drug resistance, and induction of a cancer “stem cell” phenotype.^24,31 Importantly, the uncontrolled formation of new blood vessels during tumor-induced angiogenesis lacks organizational hierarchy, inducing resistance to blood flow and contributing to compromised perfusion.

The abnormalities in the mechanopathology of the TME largely account for why standard treatments frequently encounter challenges in effectively tackling certain types of cancer, despite their significant effectiveness in targeting and eliminating cancer cells in in vitro systems. Hypoperfusion and hypoxia have been directly linked to a poorer patient prognosis³² and are regarded as unfavorable predictive markers during the diagnostic phase.^33,34

Mechanopathology induces hypoxia, triggering immunosuppression

Despite the success of immune checkpoint inhibitors (ICIs) and adoptive cell therapy (ACT) over the recent years, a very low percentage (<15%) of patients suffering from specific cancer types respond to these therapies.³⁵ This indicates that several factors impact anti-tumor immune responses. One of these factors can be attributed to the impact of tumor mechanopathology on immunosuppression, a relationship that can be elucidated through the cancer-immunity cycle (Figure 2). The cancer-immunity cycle is initiated upon release of tumor antigens from cancer cells.³⁶ These antigens are phagocytosed by antigen-presenting cells, mostly dendritic cells (DCs) which become activated and travel to lymph nodes.^37,38 In the lymph nodes, these DCs present the tumor antigens to T cells, priming and activating them.³⁷ Activated T cells then travel back to the tumor site through the circulation and identify cancer cells exhibiting the antigen, initiating an attack. However, the absence of functional tumor vessels can directly impair the trafficking of T cells to the tumor site (Figure 2, step 4), while the dense TME acts as a physical blockade, restricting T cell infiltration into the tumor (Figure 2, step 5), and contributes to an immune-excluded phenotype.^39,40 Concurrently, the resultant hypoxic conditions generate a hostile immunosuppressive TME, reprogramming tumor-associated macrophages (TAMs) from an immunosupportive phenotype toward an immunosuppressive state and obstructing T cells from recognizing cancer cells (Figure 2, step 6) and diminishing their killing efficacy (Figure 2, step 7).^41,42,43 In addition, certain chemotherapeutics can induce immunogenic cell death (ICD), which involves the release of tumor neoantigens and the training/activation of the host immune system against these antigens. Compromised delivery of such chemotherapeutics due to hypoperfusion prevents ICD and, thus, antigen presentation (Figure 2, step 1).³⁶ Therefore, tumor mechanopathology can affect most steps of the immunity cycle creating a feedback loop that hinders anti-tumor immune responses.

Figure 2.

Tumor mechanopathology compromises anti-tumor immune response

The immunity cycle against cancer is a self-perpetuating loop consisting of seven key stages, beginning with antigen release from cancer cells and culminating in their destruction. The figure details each stage, identifying the principal cell types and anatomical regions involved in each step. Abnormalities in the TME can negatively affect the trafficking and infiltration of the immune cells into the tumor, induce immunosuppression and tumor-associated macrophage polarization, and compromise antigen presentation, creating a feedback loop that hinders anti-tumor immune responses. Key abbreviations include APCs for antigen-presenting cells, CTLs for cytotoxic T cells, TAMs for tumor-associated macrophages, MHC for major histocompatibility complex, and ICD for immunogenic cell death. Created with BioRender.com.

It is crucial to acknowledge that hypoxia undermines the effectiveness of ICIs and ACT through multiple mechanisms:

• (1)Accumulation of immunosuppressive cells. Hypoxic regions within tumors can promote the accumulation of immunosuppressive cells, such as TAMs, regulatory T cells (Tregs), and myeloid-derived suppressor cells (MDSCs). TAMs constitute the primary immune cell infiltrate in the TME of solid tumors⁴⁴ and are generally classified as M1-like macrophages, which have tumor-suppressing properties, or M2-like macrophages which have tumor-promoting properties. Higher affinity of tumor cells to M2-like TAMs correlates with poor survival.⁴⁵ Polarization of TAMs is greatly influenced by the abnormal TME, whereas hypoxia can drive macrophages to polarize into the M2-like phenotype.⁴⁶
• (2)Tregs are a specialized subset of CD4⁺ T cells that are identified by the expression of the FOXP3 gene and play a crucial role in immune regulation and promoting tolerance to tumors through the secretion of transforming growth factor β (TGF-β).⁴⁷ Hypoxia-inducible factor-1 alpha (HIF-1α) binds to FOXP3 and enhances Treg abundance.⁴⁸ Such a Treg accumulation can directly contribute to VEGF pool and dampen effector T cell function and tumor angiogenesis.⁴⁹ MDSCs are a heterogeneous population of immature myeloid cells that are recruited to the TME to suppress host immunity⁴⁶ and promote the proliferation of Tregs in a TGF-β-dependent manner.⁵⁰ MDSCs may also influence the polarization of TAMs toward the M2-tumor-promoting phenotype.⁵¹
• (3)Impaired CD8⁺function. Hypoxia can delay the differentiation of CD8⁺ T cells and decrease the production of effector and proliferative cytokines, such as interferon-γ (IFN-γ) and interleukin-2 (IL-2).⁴⁴ Recently, it has been demonstrated that sustained exposure of CD8⁺ T cells to hypoxia leads to T cell exhaustion, a state of dysfunction where exhausted T cells gradually lose their ability to perform effector functions, such as cytokine production and cytotoxic activity.⁵² This significantly limits the ability of ICIs to target and activate CD8⁺ T cells within the TME, compromising anti-tumor immunity.
• (4)Altered antigen presentation. The establishment of a successful adaptive immune response heavily relies on immunogenicity, which is primarily determined by the presentation of tumor-associated antigens through major histocompatibility complex (MHC) class I molecules.⁵³ Studies have demonstrated reduced expression of MHC class I molecules in hypoxic regions within tumors.⁵⁴ Downregulation of MHC-Ι serves as a mechanism of resistance to immunotherapy in cancer patients. This phenomenon has been correlated with poorer overall prognosis and diminished response to ICIs across various cancer patient cohorts and clinical trials.^55,56
• (5)Upregulation of immune checkpoints. Hypoxia induces the upregulation of immune checkpoints and their corresponding ligands, such as programmed death-ligand 1 (PD-L1) on MDSCs, TAMs, DCs, and cancer cells through HIF-1α.⁵⁷ Specifically, HIF-1α was found to upregulate PD-L1 on breast and prostate cancer cells.⁵⁸ Subsequently, this makes it more challenging for ICIs to block the PD-1/PDL-1 pathway effectively. In the context of ACT, the upregulated expression of immune checkpoint molecules on cancer cells poses a significant obstacle, making it difficult for the infused T cells to effectively identify and attack cancer cells.

Here, we provide a summary of established methods developed to modulate mechanical forces and rectify vascular abnormalities within the TME. These approaches aim to augment perfusion and oxygenation, optimize drug delivery, and ultimately enhance therapeutic outcomes (Figure 3).

Figure 3.

Therapeutic approaches targeting tumor mechanopathology

(A) Schematic of proposed mechanism of action of normalization strategies to improve cancer therapy. The normalization of the tumor microenvironment involves the restoration of vascular functionality and the decompression of tumor blood vessels through the targeting of ECM components. This process leads to enhanced tumor perfusion and oxygenation, fostering the activation of effector immune cells and reducing the presence of immune system regulator cells. Furthermore, it causes a shift in TAM polarization from the immunosuppressive M2-like phenotype to the M1-like phenotype. As a result, there is an increase in drug delivery efficiency and the potency of cancer cell elimination.

(B) Grade of the effect of various proposed therapeutic strategies in vascular and stroma normalization. Vascular normalization restores in a great manner the abnormalities in the tumor vasculature to a more functional phenotype using anti-angiogenic agents targeting VEGF, PDGF-B, HIF, and Ang1/2.²⁴ Mechanotherapeutics normalize tumor stroma and vasculature by targeting ECM components, through reprogramming of CAFs and/or by targeting VEGF⁵⁹ and interferon-γ.⁶⁰ Metronomic therapy induces moderate effects in vessels and stroma components via the frequent administration of chemotherapeutic drugs at lower doses than MTD. It can induce vascular normalization by increasing the levels of the endogenous angiogenesis inhibitor thrombospondin-1.²⁴ Administration of ICIs can improve pericyte coverage by overexpressing interferon-γ.⁶¹ Engineered bacteria are alternative agents than mechanotherapeutics, which induce stroma normalization via ECM degradation and vascular normalization by targeting VEGF or matrix metalloproteinases.^62,63 Created with BioRender.com.

Therapeutic strategies to restore perfusion and improve therapeutic outcomes

Vascular normalization with anti-angiogenic agents

Vascular normalization aims to change the phenotype of the abnormal tumor vasculature to a more functional phenotype that resembles the structure of normal vessels. By employing judicious doses of anti-angiogenic agents, it becomes possible to restore vessel hyper-permeability and rectify the irregular vascular network. This intervention enhances pericyte coverage and reinforces the integrity of the permeable vessels while avoiding excessive vessel pruning. The vascular normalization strategy was introduced by Rakesh Jain in 2001,¹ and since then it has shown to improve tumor perfusion and thereby increase oxygenation and delivery of medicines as well as treatment efficacy.^64,65 Vascular normalization has also an immune-supportive role in improving anti-tumor immunity via opening the route and entry of T lymphocytes and upregulation of adhesion molecules on the luminal surface of endothelial cells, respectively, and by polarizing immunosuppressive cells toward immune-stimulatory phenotypes.⁶⁶

Fortification of tumor blood vessels decreases IFP and restores the pressure difference across the vessel wall, which, in turn, enhances the extravasation of therapeutic agents.^{67,68,69,70,71} Conversely, a reduction in pore size in the vessel walls leads to a decrease in the maximum particle size that traverses across the vessel wall. In a previous study, we have shown that vascular normalization has the ability to improve the delivery of nanoparticles less than 60 nm in diameter.⁶⁷ Studies involving vascular normalization have been conducted on both animal and human tumors, yielding compelling evidence that this approach has the potential to enhance cancer treatment.^{1,72,73,74,75} However, normalization is dose and time dependent, and, thus, high doses or prolonged use of the anti-angiogenic drug can prune the vessels and drastically reduce perfusion.^29,76 Therefore, vascular normalization might have transient effects, which result in a “normalization window” within which perfusion is improved.^29,64,65,77 During angiogenesis, the main angiogenic signaling player is VEGF that promotes the growth of new blood vessels.⁷⁸ VEGF forms part of the mechanism that restores blood supply to cells and tissues when they are deprived of oxygenated blood due to compromised blood circulation. The overexpression of VEGF is a contributing factor to the development of the disease. For example, solid tumors require an increased blood supply if they are to continue growing beyond a certain size and tumors that express VEGF are able to continue growing because they can develop this enhanced blood supply, through angiogenesis. Hence, cancers that exhibit VEGF expression can proliferate and spread to distant organs and areas within the body. Drugs have been used to induce vascular normalization by blocking VEGF-A or inhibiting the three VEGF receptor (VEGFR-1, -2, -3) tyrosine kinases (e.g., cediranib, sunitinib).^{72,73,74,75,77,79} Additionally, a number of other molecules, such as PDGF-B, HIF, and Ang1/2 presented in both cancer and host cells can be targeted to induce vascular normalization.⁶⁵ Dual inhibition of VEGF/VEGF receptor (with cediranib) and Ang2 (with MEDI3617) inhibited tumor growth and prolonged vessel normalization compared with VEGFR inhibition alone, resulting in improved survival in murine glioblastoma models.⁸⁰ Specifically, dual therapy increased tumor microvessel density, perivascular cell coverage, and basement membrane attachment fortifying the vessel wall and leading to a normalized vessel structure.

As far as the immune TME is concerned, vascular normalization has been shown to skew TAM polarization away from the immunosuppressive M2 phenotype to M1 phenotype, which is tumoricidal, resulting in a less immunosuppressive TME.^69,80,81 Consequently, this led to the activation of DCs, cytotoxic T lymphocytes, and natural killer (NK) cells.⁸¹ Specifically, it has been shown that low doses of anti-VEGFR2 antibody treatment induce normalization of breast tumor vessels and produce a uniform distribution of perfused tumor vessels, promoting perivascular TAMs changing from M2-like to M1-like phenotype. Elevated CXCL9 expression in M1-like TAMs further promotes T cell infiltration into tumors, while inhibiting the number of tumor-infiltrating MDSCs in tumors.⁷⁶ Rolny et al.⁸¹ demonstrated that the host-produced histidine-rich glycoprotein (HRG), which is an anti-angiogenic and immunomodulatory factor, promotes anti-tumor immune responses and vessel normalization and inhibits tumor growth and metastasis, while improving chemotherapy. In murine models of breast cancer and colorectal cancer,⁸² thalidomide induced a regular monolayer of endothelial cells in tumor vessels, inhibited vascular instability, and normalized tumor vessels by increasing vascular maturity, pericyte coverage, and endothelial junctions. Enhanced tumor perfusion led to increased delivery of cisplatin chemotherapy, culminating in significant anti-tumor and anti-metastatic effects. In a pertinent study, tumor vascular stabilization of the phosphatidylinositol 3-kinase (PI3K) inhibitor HS-173 resulted in a decrease in tumor vessel tortuosity and vessel thinning and improved vessel function and blood flow in pancreatic cancer.⁸³

Clinical studies (Table S1) have verified in humans that anti-angiogenic therapy (AAT) can normalize tumor vasculature and the patients whose tumor blood perfusion increases survive longer.^29,75,79 The agents that have been used to induce vascular normalization include the monoclonal antibody bevacizumab in patients with rectal cancer and tyrosine kinases in patients with recurrent or newly diagnosed glioblastomas.^{72,73,74,75,77,79} VEGF blockade had a direct and rapid anti-vascular effect in human rectal cancer tumors.⁷³ The use of VEGF-specific antibody bevacizumab inhibited tumor perfusion, vascular volume, microvascular density, IFP, and the number of viable, circulating endothelial and progenitor cells and improved the pericyte coverage of blood vessels.⁷³ Combination of bevacizumab with ICI durvalumab improved the levels of effector T cells in patients with HER2-negative metastatic breast cancer.⁸⁴ Furthermore, it has been shown that treatment with cediranib inhibited levels of VEGF, restoring tumor vasculature and improving tumor perfusion in glioblastoma patients.^75,79 These studies further highlighted that tumor perfusion can be used as a biomarker to distinguish responders from non-responders to AAT. In patients with hepatocellular cell carcinoma the combination of AAT with immunotherapy, but not immunotherapy alone, outperformed the efficacy of anti-angiogenic monotherapy.⁸⁵

Even though preclinical studies have shown promise, the translation of vascular normalization strategies from laboratory settings to clinical applications poses significant challenges.⁸⁶ Factors including patient heterogeneity, dosage protocols, and the identification of suitable biomarkers can profoundly influence the outcomes of clinical trials. Additionally, the absence of easily accessible non-invasive techniques for direct monitoring of normalization status further complicates the assessment of treatment efficacy. Although AAT holds the potential to reinstate the aberrant vascular state of tumors to normalcy, it is noteworthy that the observed normalization typically endures for merely a few days following the commencement of treatment. If normalization occurs too early or too late in the treatment process, it may not have the desired effect on tumor growth or metastasis. As a result, many patients experience a lack of sustained clinical responses.^87,88 Moreover, vascular normalization treatments might have unintended side effects, such as changes in blood pressure, increased risk of bleeding, or impaired wound healing. It is also of note that vascular normalization is not expected and has not shown any benefit in desmoplastic tumors with abundant compressed vessels, such as pancreatic ductal adenocarcinomas.⁸⁹ In such tumors, normalization of the tumor stroma with the aim to alleviate mechanical forces and stiffness in order to decompress tumor vessels should be considered.

Stroma normalization with mechanotherapeutics

Mechanotherapeutics is a novel class of therapeutics that aims to normalize tumor stroma by reducing tumor stiffness and intratumoral mechanical forces either via the direct depletion of ECM components or through reprogramming of cancer-associated fibroblasts (CAFs), ultimately leading to vessel decompression. Mechanotherapeutics involves usually generic drugs that have been used for decades in other diseases and are now repurposed to remodel the TME. One such agent is paricalcitol—an agonist targeting the vitamin D receptor on CAFs.⁹⁰ Another agent, currently being explored as a mechanotherapeutic, is ketotifen, an anti-histamine that inhibits mast cell activation, suppressing CAF proliferation and ECM components.^60,91 Notably, a phase 2 clinical trial is underway to evaluate ketotifen’s potential in enhancing chemotherapy for sarcoma patients (EudraCT number: 2022-002311-39). Additional mechanotherapeutics includes agents that inhibit TGF-β signaling in CAFs, such as the anti-histamine drug tranilast,⁹² the anti-fibrotic pirfenidone,⁹³ the anti-hyperglycemic agent metformin,⁹⁴ the corticosteroid dexamethasone,^93,95,96,97 and the endothelin receptor antagonist bosentan.⁹⁸ Bosentan is being tested in a phase 1 clinical trial in patients with unresectable pancreatic cancer (ClinicalTrials.gov identifier: NCT04158635). Furthermore, angiotensin system inhibitors, exemplified by losartan, not only block the angiotensin system but also reduce the activation of TGF-β signaling and ECM production by CAFs.⁹⁵ Consequently, this reduction in tumor stiffness and mechanical forces leads to the decompression of tumor blood vessels to some extent, ultimately enhancing vascular perfusion and reducing intratumoral hypoxia.⁹⁵ Importantly, losartan was the first mechanotherapeutic to successfully progress to clinical trials (Table S2). It was found that, in combination with radiation and FOLFIRINOX, losartan made 60% of previously unresectable, locally advanced pancreatic ductal adenocarcinoma tumors resectable.⁹² Currently, losartan is under investigation in combination with chemoradiation and the ICI nivolumab for the treatment of pancreatic cancer (NCT03563248).

Another mechanotherapeutic with reported TME normalization effects which made it to the clinic is the αvβ3 and αvβ5 integrin inhibitor cilengitide.⁹⁹ However, when tested in a phase 3 clinical trial, it failed to show any improvements in patient outcomes when combined with chemoradiotherapy.¹⁰⁰ ECM-targeting enzymes including the pegylated human hyaluronidase (Hy), PEGPH20, advanced to late-stage clinical trials for the treatment of pancreatic ductal adenocarcinomas. PEGPH20 administration enhanced the delivery of gemcitabine and nab-paclitaxel, with the most significant advantages being apparent in patients with hyaluronan-rich tumors who received the combination therapy.¹⁰¹ Despite these promising results, the phase 3 trial revealed that the combination had no effect on the progression-free and overall survival of patients, ultimately resulting in its retraction.¹⁰²

Decompression of tumor vessels with mechanotherapeutics would be most beneficial for tumors with abundant compressed but poorly or moderately permeable vessels, such as subtypes of sarcomas and pancreatic and breast tumors. However, decompression of tumor vessels is reversible and vessels can revert to their compressed state unless stroma normalization treatment is continued. The TME is highly complex and dynamic, consisting of various cell types, signaling molecules, and ECM components. Targeting stromal cells and signaling pathways may inadvertently affect normal tissue function or promote tumor progression through unintended mechanisms. Despite promising preclinical results, translating stroma normalization approaches into clinically effective therapies remains challenging. Factors such as patient variability, treatment resistance, and complex interactions within the TME can hinder successful clinical translation.

Apart from AAT and mechanotherapeutics, other therapeutic strategies that have shown to induce vascular or stroma normalization or both are discussed in the following paragraphs.

Engineered bacteria enhance anti-tumor therapy by normalizing the tumor stroma

Engineered bacteria have recently emerged as a more targeted intervention to modulate the TME than mechanotherapeutics. Several studies have demonstrated that tumor-targeting bacteria, whether utilized as a monotherapy or in combination with other anticancer treatments, can elicit an anti-tumor response and lead to enhanced clinical outcomes. In contrast to conventional nanoparticle carriers, of which a significant amount is rapidly eliminated by the hepatobiliary and renal route of systemic elimination, live biotherapeutics and especially anaerobes are retained in the TME where they proliferate often to numbers greatly exceeding the colony-forming units of initial administration. By exploiting the acidic and hypoxic conditions commonly found within tumors, these engineered bacteria can be utilized as delivery vehicles that respond to specific signals, such as hypoxia or low pH, and subsequently release their cargo. This is the case for Clostridium spp., which produces dormant spores that lead to its persistence and dissemination. Clostridium spores exclusively germinate in low oxygen and necrotic areas of tumor causing hemorrhagic necrosis, cell lysis, and tumor regression. To date, the attenuated Clostridium novyi-NT strain has successfully passed phase 1 and 2 of clinical trials.¹⁰³

Moreover, these bacteria can be exploited to target TME mechanopathology by disrupting the tumor’s ability to create a supportive niche for its growth, thereby weakening its resilience and potentially sensitizing it to other therapeutic interventions.¹⁰⁴ Live bacteria can be engineered to target angiogenesis in tumors. A study reports that using an orally administered DNA vaccine encoding murine VEGFR2, also known as FLK-1, carried by attenuated Salmonella typhimurium caused tumor blood vessel collapse by evoking a T cell-mediated immune response in murine models of colon carcinoma.⁶² In addition to tumor vasculature, other components of the TME can be targeted such as hyaluronic acid. For instance, the commensal Escherichia coli Nissle was engineered to generate outer membrane vesicles (OMVs) packed with a therapeutic payload allowing for its spatial and temporal release, similarly to a nanoparticle carrier. Bacteria-derived OMVs are compatible with the mammalian cell-derived exosomes since they are both nano-sized vesicles penetrating deep in the tissue, releasing their payload and participating in cellular communication. A recent study,¹⁰⁵ leveraging the ability of E. coli to localize and proliferate in the hypoxic TME, has designed a hypervesiculating strain generating OMVs (ΔECHy) to express the enzyme Hy fused to cytolysin A (ClyA), a pore-forming toxin. The resulting fused protein (Chy) was tested in mouse models for breast and colon cancer in combination with lapatinib tyrosine kinase inhibitor. Notably, combination of lapatinib with ΔECHy caused a greater reduction in tumor growth rate compared to lapatinib monotherapy suggesting that the combination of stroma degradation and the cytolytic activity of ClyA permits the penetration of drugs inside the tumor and thus potentiates their effect. Furthermore, ΔECHy treatment combined with an anti-PD-L1 antibody significantly improved the survival of mice with colorectal cancer but did not benefit mice with breast cancer, possibly due to the known immunosuppressive nature of the particular tumor model or the subtherapeutic dose of ICI used in the study. These findings suggest the use of live biotherapeutics as a promising strategy for modulating the TME to overcome chemoresistance by improving therapeutic delivery and immune cell infiltration and ultimately achieving greater control of desmoplastic tumors. Nevertheless, it is important to evaluate not only the effect of the bacteria but also the dosage, treatment frequency, and delivery mode to ensure the safest treatment that will have the least adverse health outcomes later in life.

TME normalization with metronomic chemotherapy or nanomedicine administration with metronomic schedule

Currently, traditional chemotherapy serves as the cornerstone of the standard treatment for human cancer. This therapeutic strategy involves the delivery of the highest dose of a chemotherapy drug, followed by periods of drug-free intervals to allow recovery. This regimen, known as the maximal tolerated dose (MTD) protocol, is designed to eliminate cancer cells by disrupting their cell cycle¹⁰⁶ and has demonstrated success in effectively managing various forms of human cancer in clinical practice.¹⁰⁷ However, given the genetic diversity commonly found within tumors and the fact that conventional chemotherapy primarily targets the subset of tumor cells sensitive to the treatment, MTD might lead to the promotion of the proliferation of drug-resistant tumor cell populations during the breaks between treatment cycles. This could contribute to the development of resistance to therapy and the recurrence of the disease in a subset of cancer patients following conventional chemotherapy.

An alternative method that was introduced for chemotherapy administration is metronomic chemotherapy, which is based on the frequent administration of chemotherapeutic drugs at lower doses than the MTD and with minimal drug-free breaks.^108,109 This approach has been shown to yield better survival than MTD in several preclinical tumor models and is currently being evaluated in multiple clinical trials (Table S3).^{108,109,110,111,112,113} Despite the initial belief that it works as an AAT,^114,115 it has been later shown to also affect the immune system as well as directly cancer cells, including the more chemoresistant stem-like cancer cells. The impact of metronomic chemotherapy on the tumor’s blood vessel formation might occur through the elevation of levels of the endogenous angiogenesis inhibitor thrombospondin-1 (TSP-1).²⁴ TSP-1 is produced by not only stromal cells but also cancer cells⁶⁸ and induces apoptosis in circulating endothelial cells.^116,117 Later studies showed that metronomic chemotherapy could indeed normalize tumor blood vessels.^118,119 Consistent with the effect of normalization on the immune system, metronomic chemotherapy has also been shown to reprogram the TME from immunosuppressive to immunostimulatory.^24,69 It is widely demonstrated that low and more frequent doses of cyclophosphamide have a strong impact on Treg abundance locally and systemically, in preclinical models and clinical studies. In vivo studies showed a block in Treg renewal after low-dose cyclophosphamide administration.^120,121 Several clinical trials reported suppression of the inhibitory functions and depletion of Tregs after metronomic cyclophosphamide treatment in advanced cancer patients.^122,123 Also, recently, it has been shown that low-dose metronomic cisplatin treatment reduced tumor growth and bone metastasis, induced vascular normalization, and suppressed inflammatory changes of the TME enhanced by MTD chemotherapy.¹²⁴ Chemotherapy-induced normalization effects of metronomic therapy also result in the increase in cancer cell death with particular effects on stem cell-like cancer cells.^117,125 Metronomic administration of cyclophosphamide and methotrexate has been employed in metastatic breast cancer in multiple clinical trials.^110,111,112 Based on these studies, the combination of two drugs administered in a metronomic schedule has shown comparable or enhanced efficacy in treatment outcomes. Recently, Mayer et al.¹¹³ demonstrated that the addition of bevacizumab in metronomic treatment with cyclophosphamide and methotrexate improved progression-free survival in patients with advanced breast cancer. Despite its pleiotropic effects, the metronomic regimen of cytotoxic drugs has not yet gained approval from the Food and Drug Administration (FDA). A key hurdle with metronomic therapy is that the enhancement of the immune response is influenced not only by the dosage adjustment, specifically transitioning from MTD to a metronomic approach, but also by the type of chemotherapy used. Certain chemotherapeutic agents have been observed to amplify the immune response, whereas others have contrasting effects.¹²⁶

Nanoparticle formulations seem to be more advantageous in comparison to traditional chemotherapy due to their propensity for selective accumulation within tumor tissues as a result of the enhanced permeability and retention effect and their prolonged circulation in the blood.^127,128 Interestingly, in preclinical studies, CRLX101 nanoparticles, which are composed of cyclodextrin-containing polymer conjugated to the highly potent cytotoxic drug camptothecin, have shown to improve therapeutic outcomes on both cancer cells and stem cell-like cancer cells and improve tumor perfusion and hypoxia^129,130 similar to the normalization effects of metronomic chemotherapy. Even though CRLX101 failed in clinical trials, similarities between nanomedicine and metronomic therapy imply that these approaches can be evaluated within a shared theoretical framework that aims to induce the normalization of the TME, ultimately enhancing therapeutic outcomes. A key to the effectiveness of both approaches compared to MTD is that they can potentially maintain effective drug levels in the blood for a long duration. Interestingly, apart from CRLX101, a few other nanoparticle formulations, which have not reached clinical trials yet, have also shown to be similar to metronomic therapy normalization effects in preclinical studies.¹²⁶ We previously observed that low doses of anthracycline-based nanocarrier Doxil or NC-6300 (epirubicin micelle) induced vascular normalization.^131,132,133 Both nanoparticles prompted the maturation of blood vessels by elevating the proportion of pericytes linked to tumor-associated blood vessels, leading to reduced IFP.

ICIs to induce vascular normalization

Immunotherapy and particularly ICIs have been recently reported as a new mechanism of vascular normalization. Beyond modulating the stimulatory-inhibitory axis and promoting T cell activation,¹³⁴ ICI alone can upregulate pericyte coverage. Multiple studies demonstrate that ICI mediates its anti-angiogenic effects via an overexpression of IFN-γ, while the cells responsible for IFN-γ production vary between studies. Dual inhibition of PD-1 and CTLA-4 in transplant mammary tumors led to a significant increase in the activation of IFN-γ-secreting type 1 helper T (Th1) cells, a subset of CD4⁺ T cells, accompanied by a reduction in the expression levels of VEGFA and an increase in the expression of CXCL9, CXCL10, and CXCL11, which encode chemokines acting both as chemoattractants for effector T cells¹³⁵ and pericytes¹³⁶ and as angiostatic molecules.¹³⁷ Interestingly, a different study⁶¹ demonstrated that anti-CTLA-4 and anti-PD-1 administration can enhance tumor vessel perfusion via a CD8⁺ T cell-dependent mechanism in breast and colon tumor models. Nevertheless, our studies in syngeneic models of triple-negative breast cancer and sarcoma subtypes^{60,132,133,138} did not reproduce the previously reported effect of ICIs in facilitating vascular normalization and have not been reported to have any additive effect when administered along with a mechanotherapeutic agent. These data suggest that the impact of immunotherapy in promoting vascular perfusion is likely tumor type and dosage dependent and perhaps not as strong as the use of mechanotherapeutics or AAT.

Besides T lymphocytes, other immune cells are likely to contribute to immunotherapy-induced vascular normalization. Eosinophils accumulate in the TME in response to IFN-γ production by CD8⁺ T cells.¹³⁹ Upon anti-CTLA-4 treatment, intratumoral CD8⁺ T cells produce high amounts of the CCL5 chemokine whereas CD4⁺ T cells express CCL11, acting as chemoattractants for T lymphocytes and eosinophils, respectively. Activated eosinophils can induce M1 skewing of TAMs, and at least in part, facilitate vascular normalization by reducing VEGF production, upregulating VCAM1, and further promote adhesion of eosinophil and CD8⁺ T cells to the endothelium, thereby establishing a positive feedback loop.¹⁴⁰ Eosinophils may also contribute to vascular normalization by impeding Treg accumulation, a common characteristic of hypoxic tumors.⁴⁹ Furthermore, tumor necrosis factor alpha (TNF-α) is another cytokine secreted by T lymphocytes and macrophages with reported vascular normalization properties. In contrast to IFN-γ which acts as an anti-vascular agent, application of low doses of TNF-α cytokine enhances anticancer immunotherapy by increasing vascular functionality, rather than destroying angiogenic vessels.¹⁴¹

Sonopermeation for instant mechano-modulation of TME

Ultrasound imaging has been used in the clinic for many years as a safe and widely applied real-time diagnostic imaging modality. Recently, it has also been increasingly employed for therapeutic purposes.¹⁴² The use of ultrasound waves directed at tissues in the presence of microbubbles leads to localized mechanical forces generated by the oscillation and/or collapse of the microbubbles. As a result, transient nanometer- to micrometer-sized openings are formed in the vessel walls. This phenomenon is known as ultrasound cavitation or sonopermeation.¹⁴³ The temporary openings created by sonopermeation allow therapeutic agents to enter the tumor more effectively. It is important to recognize that the ultrasound parameters can be adjusted based on the desired therapeutic result, the precise tissues under focus, and safety concerns. These adjustments can make the mechanical effects resulting from the collapse of microbubbles more aggressive, potentially causing significant disruption to microvessels, cessation of blood flow, and even potential cell death.

At the early stages, sonopermeation was used for the temporal disruption of the blood-brain barrier to enhance delivery of drugs of various sizes.^144,145,146 The promise of sonopermeation to improve perfusion and enhance delivery of chemotherapeutics or nanoparticles to cancer has been extensively investigated in several preclinical studies (Figure S2). Combination of sonopermeation and gemcitabine in pancreatic adenocarcinoma mouse models was found to have superior effects in cancer treatment compared to gemcitabine alone.¹⁴⁷ More recently,¹⁴⁸ another study in pancreatic tumors showed that the employment of sonopermeation in conjunction with FOLFIRINOX treatment significantly increased the uptake of platinum, in contrast to untreated tumors or single-agent therapy. Moreover, compelling evidence has emerged from studies conducted on subcutaneous models of colorectal cancer¹⁴⁹ and prostrate adenocarcinoma,¹⁵⁰ illustrating that sonopermeation can enhance the uptake of Doxil. Also, when exposed to sonopermeation, delivery of free and encapsulated cabazitaxel increased in prostate cancer.¹⁵¹

Although the main emphasis of sonopermeation lies on enhancing drug delivery and therapeutic efficacy by facilitating the penetration of therapeutic agents into cells, it can also influence the TME. It has been reported that sonopermeation can reduce intratumoral mechanical forces and improve perfusion in prostate adenocarcinoma and osteosarcoma murine models, indicative of stroma normalization.¹⁵² Also, sonopermeation significantly increased perfusion measured with contrast-enhanced ultrasound imaging and decreased tumor microvascular density, demonstrating vascular normalization in colon cancer.¹⁵³ Enhanced perfusion resulting from sonopermeation-induced alterations in the TME is a relatively new finding and could potentially modify the immune-suppressive factors commonly found within tumors. This modification could potentially facilitate the increased infiltration of immune cells into hypo-vascular tumors, offering new avenues for effective cancer treatment.

Sonopermeation efficacy has been also investigated in clinical trials. It has demonstrated improved effects on conventional chemotherapeutics in patients suffering from pancreatic cancer, liver metastasis resulting from primary colon cancer, and hepatic metastases from tumors of the digestive system.^154,155 Dimcevski et al.¹⁵⁴ showed that combination of gemcitabine and sonopermeation extended survival in pancreatic cancer patients without any additional toxicities. Recently, it has been reported that sonopermeation failed to improve efficacy of chemotherapy in patients with liver metastases.¹⁵⁵ Finally, several other studies incorporating sonopermeation have been initiated in patients with breast cancer, glioblastoma, pancreatic cancer, liver metastasis from breast and colorectal cancer, and brain metastases from melanoma (NCT03322813, NCT03477019, NCT04146441, NCT04021420, and NCT03385200).

Biomarkers of tumor mechanopathology to predict response to therapy

The crucial role of tumor mechanopathology in the efficacy of cancer therapy and the strong evidence that making tumors softer and improving perfusion can boost significantly immune responses and anti-tumor effects of cancer therapies have been well recognized. It is thus reasonable to argue that specific measures of stiffness and perfusion could be developed as biomarkers predictive of therapy response and for separating patients as responders and non-responders to a particular treatment. Therefore, apart from the usual practice of identifying biomarkers based on genomic/biological analysis, specific aspects of the mechanical TME could be also employed as predictive biomarkers to complement biologically derived markers and assist clinicians in the decision-making process. Tumor stiffness, perfusion, and hypoxia could serve as such mechanical biomarkers.

Tumor stiffness has been measured from biopsies using atomic force microscopy (AFM), which was able to identify unique elastic modulus distribution defined as “nanomechanical fingerprints” for tumor and normal tissues, and it is tested as a marker for tumor detection.^156,157 Our research supports the hypothesis that AFM-derived nanomechanical fingerprints are sensitive to TME modifications and could thus be considered for predicting and monitoring TME normalization and efficacy of cancer therapy.¹⁵⁸ Additionally, non-invasive, ultrasound-based methods have been developed for acquisition of elasticity maps using ultrasound shear wave elastography (SWE) both in preclinical studies and in the clinic.^{133,138,159,160} SWE-derived stiffness measures by accounting for average values of the elastic modulus over the entire tumor region or by applying machine learning methods to identify complex patterns and subvisual features that have been used not only for cancer detection¹⁶¹ but also for the prediction of tumor response to therapy.^159,162

Perfusion is measured in oncology using contrast-enhanced ultrasound, and several measures have been proposed based on the analysis of the produced time intensity curves (e.g., rise time, mean transit time, area under the curve, and peak intensity).¹⁶³ Perfusion has been related to the efficacy of some cancer therapies, including immunotherapy,⁶¹ and recently we found that the perfusion area fraction identified with contrast-enhanced ultrasound at the time of peak intensity can be used as a biomarker for the prediction of chemo-immunotherapy in preclinical tumor models.¹⁵⁹ Finally, hypoxia has been considered as a marker for several cancer therapies,^164,165 and it can be measured with non-invasive, clinically applied methods, such as photoacoustic imaging¹⁶⁶ and mainly with PET/scan imaging.¹⁶⁷

Image analysis from histopathology samples offers valuable insights into mechano-dysregulation within tissues, providing potential biomarker readouts of various abnormalities, such as local fibrosis, compressed tissue phenotypes, and disrupted tissue architectures.^168,169

It is important to note that, despite the tremendous scientific efforts in the field, only few biomarkers have been developed to predict treatment response, which are mostly based on the human genome (e.g., breast cancers positive to the HER2 protein are treated with chemotherapy and the HER2-targeted antibody herceptin). However, tumor types might have a significantly different genetic background, and, thus, the identification of common predictive markers based on their genome is an extremely difficult task. On the other hand, tumor stiffening and hypoperfusion are common for all fibrotic tumor types and their effects on reduced oxygenation are well documented. Therefore, mechanical biomarkers could be applied to more than one desmoplastic tumor types. There are also additional benefits from the development of biomarkers. Specifically, cancer patients who will be classified as non-responders (Figure 4) will not undergo therapies with severe adverse effects, and these biomarkers can be used to guide patient-specific treatment optimization that involves the strategies reviewed here for improvement of tumor perfusion (e.g., AAT, mechanotherapeutics, etc.).

Figure 4.

Biomarkers predictive of response can be used to guide therapy and benefit non-responder patients

Ultrasound shear wave elastography, atomic force microscopy, and magnetic resonance elastography-derived measures of tumor stiffness and ultrasound contrast-enhanced ultrasound and PET scan-derived measures of perfusion and oxygenation can predict response to cancer therapy and guide the use of normalization agents to potentiate therapy in non-responders.

Conclusions

The application of physical sciences and engineering in oncology represents a multidisciplinary domain, necessitating the amalgamation of insights from tumor mechanopathology and drug delivery. The outcomes of this collaborative endeavor have made their way into clinical applications. A key takeaway from years of research in this area is that the modification of the physical microenvironment within solid tumors can effectively surmount the challenges posed by the vascular and non-vascular components of these tumors. The encouraging aspect is that some drugs with the capacity to normalize the TME are already clinically approved, with many of them employed in clinical practice for treating other medical conditions.

In this review, we summarized the strategies that can be used to normalize the TME, increase perfusion and drug delivery, and improve therapeutic outcomes. While these strategies are starting to benefit cancer patients, there is still substantial work ahead due to their limitations. The goal of vascular normalization is to create an environment that enhances the effectiveness of AAT by improving drug delivery and creating conditions more conducive to the immune system’s activity. It is worth noting that, while this strategy holds promise, achieving successful and consistent vascular normalization in clinical settings remains a challenge due to the complex and dynamic nature of tumor angiogenesis. Mechanotherapeutics demonstrate a promising ability to improve drug delivery in preclinical and clinical studies, but they cannot decompress all collapsed tumor blood vessels for all cancer patients nor alleviate all signaling barriers that anti-tumor immune cells face. Moving to metronomic strategies, preclinical and clinical evidence supports metronomic chemotherapy as an efficient tool to fight certain types of cancer. Nevertheless, the advancement of metronomic chemotherapy faces limitations that should be overcome. Upcoming preclinical and clinical investigations will be required to determine the most suitable agents for specific tumor types, the optimal combination of agents, the appropriate individual and combined dosages, as well as the timing of drug administration. The duration of treatment and the ideal method for discontinuing therapy should also be fine-tuned. Biotherapeutics offer a promising approach to modify the TME, overcoming chemoresistance by enhancing drug delivery and immune cell infiltration, thereby gaining better control over desmoplastic tumors. However, it is crucial to not only assess the impact of the bacteria but also consider factors, such as dosage, treatment frequency, and the mode of delivery to ensure the safest treatment with minimal long-term adverse health effects. As for the use of sonopermeation, its effectiveness depends on several factors including the type of cancer, the choice of therapeutic agents, the ultrasound parameters, and the delivery methods. In addition, sonopermeation requires an effective vascular network so that microbubbles enter uniformly the tumor region. Therefore, vessel compression compromises the efficacy of sonopermeation as the microbubbles cannot effectively and uniformly be delivered to the entire tumor, leading to local effects and compromised efficacy of this therapeutic strategy.^143,147 Decompression of tumor vessels through pre-treatment with a mechanotherapeutic would increase delivery of microbubbles and enhance the efficacy of sonopermeation. This suggests that sonopermeation could be combined and have additive effects with mechanotherapeutics on improving tumor perfusion.

The strategies presented in this review are promising, and some of them have already reached clinical trials. To turn these strategies into standard-of-care clinical practice, one needs to identify which tumors have leaky or compressed vessels, both, or neither. This is a challenging task. Although we can make some broad statements, such as pancreatic ductal adenocarcinomas have abundant compressed vessels, there are many tumors, such as breast cancers, in which the degree of desmoplasia and vessel compression is highly variable from one tumor to the next and potentially from the primary site to the metastatic site, and thus it could be hard to choose an appropriate strategy until the state of that individual tumor is known. Emerging imaging approaches and biomarkers have the potential to help in this selection.

Finally, it is conceivable that opening blood vessels could bring more nutrients to the tumor, which might increase its progression. Also, improved perfusion could allow more metastatic cells to leave the primary tumor. Therefore, drugs that improve vessel functionality should be given with concurrent cytotoxic treatments, such as chemotherapy, immunotherapy, or another cancer-cell-targeted treatment.