Keywords

2.1 Switching the Focus from Tumor to Tumor Microenvironment

The biological implications of the tumor microenvironment (TME) on cancer progression and its spreading have begun to be suggested over the past few years. Several studies have demonstrated that TME is not just a silent bystander, but rather an active promoter of cancer progression. A milieu of immunosuppressive T-reg lymphocytes, tumor-associated macrophages, fibroblasts, and adipocytes makes up the TME, providing a real sanctuary for cancer [48]. In particular, cancer-associated fibroblasts (CAF) are key components of the TME, closely supporting cancer by secreting mitogenic growth factors such as a fibroblast growth factor (FGF) or the insulin-like growth factor 1 (IGF-1) [4]. Furthermore, CAF are centrally involved in the NF-kB inflammatory signaling pathway which promotes tumor progression, also stimulating neo-angiogenesis. The transforming growth factor beta (TGF-β) deriving from CAF induces the epithelial-to-mesenchymal transition (EMT), which is considered the key process in cancer invasion and distant spread due to the acquisition of mesenchymal stem cell features. Not only are CAF involved in such a complex crosstalk between cancer cells and TME, they are also structurally fundamental for a cancer-supporting TME. Indeed, CAF produce extracellular matrix (ECM) proteins , which are responsible for the desmoplastic reaction at the edges of a tumor and within cancer cells, protecting them from antitumor immune responses and chemotherapeutics. Recently, the interest of researchers and clinicians has focused on CAF, considered to be key mediators of cancer-stroma crosstalk, and a promising target for novel therapeutic approaches toward TME in cancer treatment.

2.2 The Heterogeneous Nature of Cancer-Associated Fibroblasts

2.2.1 Origins and Functions of CAF

Fibroblasts are the most abundant stromal cells in the TME , accounting for up to 80% of the tumor mass in certain solid tumors characterized by a desmoplastic reaction [120]. They are particularly important because of their continuous and complex crosstalk with cancer cells [51, 91]. From a quiescent state, fibroblasts can be reversibly or irreversibly activated to form myofibroblasts in response to different inputs (Fig. 2.1). Myofibroblasts, induced by TGF-β-mediated signaling, proliferate, gain contractile properties, and unleash an injury response to repair the cellular damage and to restore tissue homeostasis [23, 26, 66, 111]. When fibroblast activation persists even in absence of the initial injury (e.g., in chronic tissue damage or fibrosis), a pathological remodeling occurs, partly depending on epigenetic regulation [121], and tumor initiation is promoted [27, 29, 98], so that tumors are considered “wounds that do not heal” [30]. By supporting tumorigenesis and by interacting with cancer cells at all tumor stages, hyperactivated fibroblasts gain enhanced proliferative properties and become a functionally diverse population, called cancer-associated fibroblasts (CAF) [52]. CAF may derive from a variety of cells, including normal fibroblasts, but also surrounding endothelial cells, pericytes, stellate cells, bone marrow-derived mesenchymal cells, and adipocytes [52, 60, 92]. Depending on their origin, the function of such activated fibroblasts can be diverse and unique (Fig. 2.2). Mediators for CAF transformation are growth factors, cytokines, and micro-RNAs soaked in the tissue milieu that can modulate the cellular response through a variety of molecular mechanisms [34]. In the early stages of neoplasia, the pathological tissue remodeling may initiate tumor-promoting functions in fibroblasts through the secretion of pro-inflammatory cues, such as interleukin (IL)-1β by immune cells [36]. Later, as the tumor grows, most of the CAF-transforming factors, including TGF-β, platelet-derived growth factors (PDGF), and FGF2, derive from direct secretion by cancer or stroma cells, either as soluble factors or transported by exosomes [3, 10, 34, 63, 65]. Moreover, matrix stiffness and solid stress in the TME constitute additional physical factors that cause sustained activation of CAF, through a feedback loop involving YAP activation and Rho-associated protein kinase (ROCK) signaling pathways [14].

Fig. 2.1
figure 1

Activation of fibroblasts in the tumor microenvironment , from quiescent resident fibroblasts to activated myofibroblasts to hyperactivated CAF, with sequential acquisition of key phenotypical and functional features

Full size image
Fig. 2.2
figure 2

CAF origins in the tumor microenvironment. The acquisition of a CAF phenotype is associated with the expression of a variety of CAF-related markers

Full size image

Multifaceted bio-functions of CAF aim to orchestrate the TME and manage the tumor-stroma interface via intercellular contacts, secretion of a number of factors, modification of the ECM, and promotion of malignant transformation of epithelial cells [67, 82]. CAF contribute to hypoxia-dependent tumor neo-angiogenesis and are key actors in the restricted penetration of drugs and nanodrugs in the tumor tissue, thus modifying tumor responsiveness and therapeutic efficacy of several drugs [52, 61]. Additionally, there is evidence that CAF promote cytotoxic T cell exclusion from the tumor and hinder antitumor immune responses [57].

2.2.2 Coexisting CAF Subsets

Unlike normal fibroblasts, CAF are characterized by an increased expression of certain biomarkers, which have been recently studied as potential targets for innovative therapeutics [16, 18, 112]. Depending on tumor type and origin, CAF express high levels of alpha-smooth muscle actin (α-SMA), fibroblast activation protein (FAP), fibroblast specific protein 1 (FSP1 or S100A4), vimentin, and platelet-derived growth factor receptor (PDGFR)-α and β [53, 85, 86, 107, 118]. Leucine-rich repeat containing 15 (LRRC15) membrane protein, CD10, and G protein-coupled receptor 77 (GPR77) were also found highly expressed in CAF in many solid tumors [24, 59, 89, 108]. Unfortunately, none of the identified markers are currently able to select CAF with a high degree of specificity, because of a high-grade heterogeneity characterizing this cell population [2]. As an example, the loss of caveolin 1 (CAV1) expression in breast tumor cases defines fibroblasts with pro-tumorigenic functions [102]; however, a high expression of CAV1 in CAF could also facilitate tumor invasion via ECM remodeling [41]. Thus, nowadays, it is becoming increasingly recognized that CAF represent a heterogeneous cell population of multiple origins [49]. Researchers have demonstrated the existence of distinct subsets of CAF with different localization within the tumor mass and specificity per tumor type [79, 109]. The existence of four CAF subsets has been demonstrated in triple-negative breast cancer (S1–4) and pancreatic ductal adenocarcinoma (subsets A-D). All subtypes have unique properties and expression profiles, as assessed by marker analysis and transcriptomic investigation [6]. Of note, a specific CAF phenotype corresponds to a prognostic impact. In breast cancer, S1-CAF are associated with immunosuppressive TME by promotion of T cell differentiation into T-reg, while S4-CAF are associated with high CD8+ T cell infiltration into the tumor [21].

2.2.3 Friend or Foe?

In many tumors CAF accumulation in the TME is often correlated with poor prognosis [7, 118]. Indeed, their presence is an effective predictor of tumor reoccurrence in colorectal cancer patients and has been highlighted as a significant prognostic factor in a number of other tumor types [12, 13]. At the same time, the functional role of CAF in cancer progression and metastasis is emerging as being complex and bimodal, with both cancer-promoting and cancer-restraining actions. Recent studies have suggested that CAF can restrain pancreatic ductal adenocarcinoma (PDAC) by reducing fibrosis and hypoxia [95]. Also, patients with high desmoplasia can have improved prognosis and overall survival in PDAC, breast cancer, and lung cancer, as demonstrated by correlation studies between CAF markers and disease outcome [38, 84]. CAF have also been suggested to play a tumor-suppressive role via the I kappa B kinase/NF-kB pathway, lowering hepatocyte growth factor (HGF) secretion and reducing tumor size and metastasis [79]. Keeping all of this in mind, CAF are not a unique population, but rather an updated description of CAF requires taking into consideration their dynamic state, with epigenetic changes and variable gene expression and functions.

2.3 Fibroblasts and Tumor Progression: A Key Role in Tumor Architecture Remodeling and Desmoplasia

Over time, researchers have progressively realized that initiation, proliferation, invasion, and metastases of tumors do not rely on tumor cells properties alone, but they are influenced by the pathological stroma. From the “seed and soil” hypothesis, it has been recognized that the dynamic crosstalk between cancer cells (“seed”) and TME (“soil”) has a pivotal relevance in a variety of processes such as proliferation, migration, invasion, survival, angiogenesis, and EMT [83]. Through EMT, cancer cells gradually lose their epithelial hallmarks and acquire mesenchymal properties related to invasiveness and the remodeling of surrounding ECM [58]. The final result of EMT is the capability of cancer cells to reach blood circulation and metastasize at distant sites, making cancer progress from an in situ lesion to an invasive disease [46]. CAF have been shown to support cancer cell growth and metastatic dissemination in several ways [11, 51, 97] (Fig. 2.3). Their effects are mediated through both paracrine and autocrine stimulation by a variety of growth factors and cytokines, including TGF-β, bFGF, vascular endothelial growth factor (VEGF), PDGF, and interleukins (IL) [75]. TGF-β/TGF-βR signaling is required for advanced carcinogenesis via EMT induction, angiogenesis, and the modification of the stromal compartment [22, 42]. CAF-derived TGF-β1 was identified as a central molecular regulator of mesenchymal stem cells as well as a tumor-promoting factor in prostate cancer and other types of carcinoma [68, 71, 96]. Other important cues that drive the gaining of mesenchymal traits include HGF, stromal-derived factor-1α (SDF-1), osteopontin (OPN) and key cytokines released by CAF able to reprogram cancer cells through activation of the Wnt/β-catenin signaling pathway, which fosters migration and metastasis [110]. HGF and IL-6 are also considered drivers of tumor initiation and progression, through their interaction with MAPK, PI3K/Akt, and JAK/STAT signaling pathways, along with the subsequent induction of c-MET expression as positive feedback regulation [25, 116]. The coordination of these pathways controls tumorigenic progression in response to CAF’s paracrine activity. CAF-derived SDF-1, also known as CXCL12, is also able to induce an angiogenic response in synergy with the chemokine ligand CXCL -8 and enhances the invasiveness of pancreatic cancer cells [69, 81]. A gene knockdown assay and gain- and loss-of-function assays revealed that CAF secrete TGF-β and SDF-1, which promote the formation of capillary-like structures, participate in vascular endothelial cells migration, tube formation, and angiogenesis via interaction with TGF-βR1 and CXCR-4 in tumor cells [37, 119].

Fig. 2.3
figure 3

Multivalent activity of CAF and their secretome for shaping the tumor microenvironment

Full size image

The ability to control the local remodeling of ECM is another critical function of CAF and a feature of paramount importance during the desmoplastic reaction occurring in many carcinomas. Activated fibroblasts are an important source of ECM-degrading proteases, including matrix metalloproteinases (MMP), zinc-dependent endopeptidases that facilitate cancer cell migration across ECM [101, 105, 114]. MMP-3, produced by CAF, promotes EMT by cleavage of E-cadherin and induces invasiveness of cancer cells [64, 104]. MMP-13 promotes angiogenesis by releasing VEGF and increasing the invasive capabilities of squamous cell carcinoma cells [56]. Additionally, other stromal MMP, such as MMP-1, MMP-9, and MMP-14, are able to induce cancer invasiveness, and their expression has been associated with tumor progression in several carcinomas [8, 106].

2.4 Cancer-Associated Fibroblasts in Immunosuppression and Chemoresistance

Generally, CAF are known to promote an immunosuppressive TME. Fibroblasts are a significant source of immunomodulatory cytokines and chemokines, notably interferon-γ, IL-6, CCL2, and tumor-necrosis factor-α, which can influence the mobilization of cytotoxic T lymphocytes, natural killer cells, and macrophages [43, 93, 100] (Fig. 2.3). Paracrine CAF–immune cell signaling may induce differentiation of immunosuppressive myeloid cells and affect macrophage recruitment to the tumor [55, 72, 113]. T cell recruitment and activation also involves cytokines that are found in the CAF secretome, such as CXCL9, CXCL10, and SDF-1 [5]. A recent study has shown that programmed cell death 1 ligand (PDL) 1 and 2 in a subset of CAF derived from patients with lung cancer may carry an immunosuppressive effect on T cell activation ex vivo [77, 87].

Beyond CAF secretome which switches off anticancer immunity , CAF-orchestrated ECM plays a crucial role in restricting access of immune cells to cancer, by generating a physical barrier to tumor infiltration and unmasking cryptic binding sites that could promote immune cell adhesion [33, 50]. In orthotopic tumor grafts, targeting FAP+ CAF with a DNA vaccine showed antitumor effects via suppression of collagen synthesis and intratumoral recruitment of CD8+ T cells, with the subsequent immuno-control of tumor growth [62, 80]. CAF distribution at the interface between blood vessels and tumor cells contributes to increasing the tumor interstitial fluid pressure, which represents a physical barrier to several drugs [7]. Moreover, dynamic ECM alterations may induce tissue stiffening and increased tension, which have been associated with poor outcome in patients with many solid tumors [14]. The immunosuppressive and poorly accessible TME drastically limits the potential of effective therapeutics, which have raised new hopes for the treatment of several malignant tumors. Therefore, favoring ECM remodeling and overcoming immunosuppression in the tumor is of fundamental importance for effective anticancer treatment .

2.5 Targeting Cancer-Associated Fibroblasts: Current Clinical Evidence

Considering the central role of CAF in cancer progression and diffusion, it is quite surprising that TME-targeted treatments have been so poorly explored in clinical trials to date. A main reason for the lack in clinical data is the relatively difficult specific targeting of CAF. A promising candidate for CAF targeting is FAP, a cell surface glycoprotein expressed in over 90% of these stromal cells while normally not expressed in most healthy tissues. In 1994 a first phase I study evaluated the clinical use of a monoclonal antibody toward FAP for imaging purposes, labeling it with iodine 131, to detect liver metastases from colorectal cancer [115]. In accrued patients, iodine 131-labeled anti-FAP antibodies were administered 1 week before liver surgery or regional chemotherapy, demonstrating a high accumulation within liver metastases but not in liver normal parenchyma, and no significant toxicities. Therefore, a first proof-of-concept on selective overexpression of FAP in metastatic colorectal cancer was provided, together with the usefulness of focusing on TME for clinical purposes. Subsequently, the anti-FAP antibody named sibrotuzumab was clinically assessed for anticancer efficacy in further trials. First, a phase I clinical study evaluated sibrotruzumab in FAP+ metastatic colorectal and non-small cell lung cancer [99]. After 12 weeks, treatment with sibrotuzumab showed no significant toxicities and was overall well tolerated. However, on the other hand, cancer progression was observed in all included patients, and no objective tumor response was reported. In another phase II trial, sibrotuzumab was administered in metastatic colorectal cancer patients: unfortunately, all patients still experienced cancer progression except for 2 cases, where a stable disease was observed [44]. Despite, yet again, the fact that no significant toxicities were reported, the trial failed to provide a benefit from sibrotuzumab and it was terminated. Furthermore, although no severe adverse events were reported, it should be noted that FAP is overexpressed also in bone marrow, further making clinical translation difficult. The discouraging findings from the above-mentioned trials have resulted into a long-lasting abandonment of the interest toward sibrotuzumab ,; however it has also produced a number of lines of thought. Targeting the TME could probably be a winning strategy in preventing reactivation and progression of dormant metastatic tumor cells, rather than arresting the metastatic storm once the TME has elicited its promoting activity [19]. Indeed, once cancer progression has started and metastatic disease occurs, a large amount of cancer-promoting forces are activated, making it difficult to be effectively counteracted by targeting tumor stroma only. Targeted therapy for TME might therefore be preferred as an ancillary treatment to support conventional chemotherapy in the first-line therapy of cancer, since its anticancer efficacy as a stand-alone treatment is limited, as demonstrated in preclinical studies on FAP inhibition [20] or in clinical trials targeting other TME actors, such as metalloproteinases [78]. More recently, another approach to target CAF activity has been proposed, based on inhibition of FAP enzymatic activity rather than targeting FAP itself. In a phase II clinical trial, Talabostat , an orally available amino boronic dipeptide which competitively inhibits the dipeptidyl peptidase activity, has been administered as a stand-alone therapy in metastatic colorectal patients previously treated with conventional chemotherapy [76]. Although 21% of patients maintained a stable disease for up to 25 weeks, no objective responses were observed, demonstrating a minimal clinical activity of Talabostat. However, since it was tolerated well by patients, Talabostat was further assessed in non-small cell lung cancer patients in combination with docetaxel; however, only 3 patients out of 42 reported an objective response [31]. Since Talabostat has been related to increased production of cytokines leading to enhanced antitumor immunity [1], this FAP inhibitor represented a hope for new treatment approaches in highly immunogenic malignancies, such as melanoma. Inspired by the intriguing discovery that Talabostat with cisplatin makes mice resistant to rechallenge with melanoma cells, a phase II trial evaluating Talabostat and cisplatin as a second-line therapy for metastatic melanoma was conducted [32]. A partial response was observed in less than 10% of included patients, similarly to treatment with cisplatin alone: thus, Talabostat added no clinical benefit. Furthermore, about one-third of patients experienced severe side effects related to the use of Talabostat , mainly anemia, thrombocytopenia, and neutrophilia. Regarding the minimal clinical effect, it should be noted that Talabostat acts by inhibiting the peptidase activity of FAP only; however, it was recently demonstrated that FAP promotes cancer growth and progression also through non-enzymatic activities, such as stimulating ECM remodeling by MMP-9 [47]. Phase III clinical trials on Talabostat combined with docetaxel or pemetrexed for treatment of late-stage non-small cell lung cancer were initiated, but these studies were prematurely stopped at the interim evaluation due to the observation of a lower survival rate in the Talabostat group compared to the placebo group [9]. The current difficulty in targeting FAP or in inhibiting its enzymatic activity has not decreased the great interest in implementing an effective strategy toward TME in cancer management. Indeed, while the targeting of cancer cells must follow their evolving wide heterogeneity with frequent onset of resistance, TME and interactions between cancer and TME are much more universal and common to different types of cancer, making targeting TME a promising approach. Therefore, an innovative strategy was proposed which focuses on growth factors deriving from CAF, such as FGF. Nintedanib is a pan-tyrosin kinase inhibitor, acting toward receptors for FGF, VEGF, or PDGF, overexpressed in cancer cells. By inhibiting the activity of the above-mentioned growth factors, the downstream support from CAF to cancer cells could theoretically be reduced or abolished, avoiding stimulation of tumor proliferation, migration, and survival. In 2010 Nintedanib was evaluated in a phase I clinical trial on 61 patients affected by advanced solid malignancies, demonstrating to be limited by G3-G4 reversible liver enzyme elevation but substantially showing a decent level of tolerability on behalf of patients. Despite the fact that only 3 clinical responses were reported, in 55% of patients, a significant reduction in tumor blood flow was observed, suggesting that targeting a CAF-derived growth factor may significantly impact on TME and its neoangiogenesis [74]. A further clinical trial of Nintedanib in advanced or metastatic relapsed non-small cell lung cancer administration achieved disease stabilization in 46% of patients, with a median progression-free survival of 6.9 weeks [94]. These encouraging findings warranted further clinical exploration of this strategy, and after the finding that Nintedanib in addition to docetaxel improves the overall survival rate, it is currently an established second-line treatment for non-small cell lung cancer [88]. Under the new perspective of targeting the signaling network of CAF, a monoclonal antibody toward TGF-β has been developed and named Fresolimumab . Recent clinical trials have evaluated Fresolimumab in previously treated melanoma, renal cell cancer [73], or metastatic breast cancer [39], but a limited clinical response was conjugated with the occurrence of secondary cutaneous malignancies, stopping any further clinical trial with TGF-β antagonists. Indeed, TGF-β may stimulate cancer in advanced stages making its inhibition a potential anticancer treatment; on the other hand TGF-β could mediate inhibition of cancer development in normal tissues [35].

2.6 Future Trends for Cancer Therapy Through Fibroblasts

2.6.1 CAF Reprograming

As suggested by the major concerns emerged from tout-court CAF-inhibiting strategies, TME might play several different roles in cancer progression, including both cancer-promoting and cancer-suppressing pathways. TME was classically depicted as a stable and universal feature of cancer, while it is increasingly recognized that it is highly heterogeneous. The coexistence of different subpopulations of CAF has been proposed, ranging from cancer-inhibiting to cancer-enhancing fibroblasts [51]. Therefore, a precision medicine approach should also be preferred in targeting CAF, and turning CAF from a cancer-enhancing profile to one that is cancer-inhibiting might be a more suitable strategy than the total depletion of CAF. Two recently proposed specific surface biomarkers of tumor-enhancing CAF are CD10 and GPR77, and a monoclonal antibody toward the latter receptor has shown reduced chemoresistance in a patient-derived breast cancer xenograft [108]. Beyond a precise targeting of cancer-supporting CAF, the main challenge is how to reprogram them in order to convert an immunosuppressive into an immune-permissive TME. An interesting approach has been proposed to block those signals fueling fibroblast activity, such as the angiotensin II-angiotensin II receptor type-1 axis. Indeed, angiotensin II transforms quiescent fibroblasts into CAF; therefore angiotensin receptor blockers (ARBs) should hypothetically reverse the process and reprogram CAF. A clinical concern is represented by the potent antihypertensive effects of ARBs, making them useless as anticancer treatment in clinical practice. However, ARBs have been recently nano-conjugated with pH-dependent degradable polymers in order to selectively direct ARBs into the acidic TME in a murine model of metastatic breast cancer [15]. Intriguingly, this strategy allowed for the reprogramming of CAF without hypotensive effects, deleting the immunosuppression promoted by TME and improving the T lymphocyte antitumor response, thus extending survival of mice with concurrent administration of immune checkpoint blockers. Another original approach for CAF reprograming is based on epigenetic regulation. The use of a selective inhibitor of histone deacetylases (HDACs) has been successfully used to interfere with TGF-β-mediated CAF differentiation, thus reversing CAF activation and delaying cancer growth [54].

2.6.2 Immunotherapy

Following the increasing interest toward antitumor immunity and strategies based on enhancement of T cell responses to cancer cells, a similar approach may be translated as anti-CAF treatment. In particular, combined treatments toward cancer cells and CAF are particularly promising. A specifically engineered T-cell engager for both FAP and human CD3 has been inserted into an oncolytic virus: the binding with CD3+ effector T lymphocytes and with FAP-expressing CAF lead to T cells activation and cytotoxicity toward CAF, while the oncolytic activity of the viral vector exerted its well-known anticancer effect [103]. This oncolytic approach not only results in CAF depletion, but it may also mediate a reversal of TME from immunosuppressive to immune-permissive, as shown by the repolarization of M2 macrophages toward a proinflammatory profile [40] in fresh prostate cancer tissue derived from biopsy samples. In other words, tumor-infiltrating lymphocytes could be reeducated to kill CAF, leading to TME remodeling and cancer suppression. Beyond oncolytic viruses, an elegant solution for priming the natural intratumoral immune response toward CAF is the use of specific vaccines. Tolerance toward FAP can be broken by specific DNA vaccines to exploit the cytotoxic activity of CD8+ and CD4+ T lymphocytes toward CAF. Interestingly, the T cell-mediated CAF depletion also decreased macrophage infiltration and increased intratumoral lymphocytes; furthermore this strategy was improved by adding tumor-specific DNA vaccines in different cancer models [28]. As previously stated, targeting TME as a stand-alone therapy might be ineffective, especially in aggressive cancers or where metastatic spread has already occurred. A combination strategy toward both cancer and TME could maximize the outcome. Therefore, other DNA vaccines to prime cytotoxic T lymphocytes toward FAP-positive CAF have been developed and tested in combination with chemotherapeutics with immunomodulatory activity, such as cyclophosphamide [117], demonstrating enhanced anticancer efficacy. An original sort of FAP-specific vaccination has been proposed by fusing dendritic cells, which normally present antigens to start the immune response, with CAF [90]. The resulting hybrid cells effectively activated T cells to generate a specific cytotoxic immune response toward CAF, inhibiting cancer growth.

2.6.3 Nano-strategies to Target CAF

Nanoparticles have been profoundly explored as an excellent drug delivery system in tumors, first exploiting their natural intratumoral delivery due to extravasation from leaky vasculature (the so-called enhanced permeability and retention effect, EPR). Then, by conjugation with specific antibodies, nanoparticles have been increasingly evaluated for actively targeting cancer. In both cases a high anticancer efficacy combined with a significantly lower toxicity have been reported, thanks to the specific action of drugs loaded inside cancer cells, thus avoiding off-target adverse effects in healthy tissues. Despite nanomedicine demonstrating great potential for cancer treatment, its clinical translation is a slow process, due to production costs and safety concerns. A special interest in nanomedicine has recently been developed also for targeting TME. Nano-liposomes conjugated with a peptide recognizing tenascin C, overexpressed in CAF, have been demonstrated to adequately address the anti-apoptotic drug Navitoclax in TME [17]. As a consequence, downregulation of ECM deposition, decreased interstitial fluid pressure, and increased blood perfusion with a subsequent improvement in chemotherapeutics penetration have been observed. The reduction in the high intratumoral interstitial pressure due to TME has been observed also with gold nanoparticles in xenograft of colorectal cancer [123]. Interestingly, after treatment with naked gold nanoparticles, CAF and pro-fibrotic signals decreased as well as TME stiffness, leading to increased penetrance and activity of cisplatin, which was subsequently administered. Similar findings were reported also for ovarian cancer , where gold nanoparticles were demonstrated to affect the VEGF signaling, thus blocking neoangiogenesis by disrupting the cancer cell-TME crosstalk [122]. The innate capability of untargeted gold nanoparticles to inhibit the interaction between cancer cells and TME has been more deeply studied: not only do they act on AKT pathways and VEGF signaling, they also modulate cancer cell secretome to reduce the desmoplastic feature in pancreatic cancers [70]. A more intriguing feature of gold nanoparticles might explain their natural anti-TME effects not only affecting cell crosstalk but also finely modulating the CAF profile. As recently demonstrated, gold nanoparticles increase lipid intracellular content by inducing an expression of lipogenesis genes in CAF, which use endogenously synthetized lipids to convert into quiescent fibroblasts [45]. Also, actively targeted nanoparticles toward CAF have been evaluated. A biocompatible ferritin-based nanocage has been engineered with a FAP-specific single-chain variable fragment to provide a prompt targeting and internalization into CAF, for subsequent photoirradiation exploiting the photosensitizing feature of ferritin [124]. By this nano-based photoimmunotherapy, CAF were efficiently depleted, enhancing T cell infiltration and tumor suppression in immunocompetent mice, again providing a proof-of-concept on the usefulness of targeting TME to increase antitumor immunity.

An increased interest toward implementation of anti-TME treatments for cancer therapy is expected over the next years. After the failure of clinical trials to demonstrate a significant benefit provided by anti-FAP monoclonal antibodies, it appeared clearer that, beyond merely killing CAF, other strategies aiming at reeducating CAF to modulate TME merit further exploration. Promising therapies in reaching this goal are selective inhibitors of CAF signaling, DNA vaccines toward CAF, and targeted nanodrugs; however, further characterization of CAF molecular biomarkers is needed in order to exploit suitable targets and thus avoid a tout-court action on all fibroblasts, including those providing anti-cancer activity, and avoid off-target toxicities. Finally, a selective modulation of TME could be an optimal treatment to prevent the invasive features of primary cancer or, in the best case, to prevent metastatic cancer cells in distant niches, but its potential efficacy for advanced/metastatic cancers is much less clear and combination strategies with cytotoxic drugs could maximize the outcome in these cases.