Keywords

3.1 Pattern Recognition Receptors (PRRs) Induce Innate Inflammation

Pattern recognition receptors (PRRs) recognize rigid features known as pathogen associated molecular patterns (‘PAMPs’) or damage associated molecular patterns (‘DAMPs’) either in the extracellular space, endosome, or cytoplasm to induce appropriate inflammation during pathogen infection and/or tissue damage. Canonical PRRs include Toll-like Receptors (TLRs), of which there are 10 (TLRs 1-10) in humans; RIG-I like receptors (RLRs) including MDA5 and RIG-I; cytosolic double stranded DNA sensors (e.g., cGAS-STING); the AIM2-like receptors; the NOD-like receptors; and C-type lectin receptors. For information on PRRs, their locations, and specificities, see Fig. 3.1. Upon recognition of PAMPs or DAMPs by PRRs, signaling to the Tank Binding Kinase 1 (TBK1) and IKK-α/β kinases primarily lead to IRF3 phosphorylation and NFκB activation, respectively [1, 2], to concertedly lead to the synthesis of pro-inflammatory cytokines (e.g., type I IFNs, TNF, IL-6) and, in DCs, induce co-stimulatory ligand expression (e.g. CD86, CD80).

Fig. 3.1
figure 1

PRR cellular locations, specificities, and downstream signaling. Yellow indicates PRRs, blue indicates signaling adapters/kinases, and red indicates transcription factors that mediate transcription of inflammatory genes. PAMPs that activate PRRs are depicted in italicized red text. Toll-like Receptors 1, 2 and 4–6 are located on the cell surface and recognize bacterial features such as lipids, proteins, and lipoproteins. TLRs 3 and 7–9 are localized to endosomes, and recognize viral nucleic acids. TLR4 is both extracellular and endosomal. The specificity and function of TLR10 (not shown) is currently obscure, but in contrast to other TLRs, may negatively regulate inflammation [3]. The RIG-I like receptors, RIG-I and MDA5, recognize cytosolic viral double stranded (ds) RNA and have recently been shown to become activated at endoplasmic reticulum (ER) derived microsomes [4]. Cytosolic DNA sensing by cGAS-STING is mediated at the ER. The AIM2 inflammasome recognizes dsDNA in the cytosol to initiate cleavage of caspase-1, followed by cleavage of pro-IL-1β and pro-IL-18 to their mature, secreted state. The NLRP3 inflammasome (a NOD like receptor) recognizes various DAMP and PAMP features, including viral dsRNA and single stranded RNA, and similarly leads to caspase-1 activation. Several variations of inflammasomes recognizing diverse features are not shown. The C-type lectin receptors Dectin-1 and 2 recognize bacterial and fungal features; several additional C-type lectins with various specificities are not shown. The transcription factors IRF3 and IRF7 largely drive transcription of type I interferons (IFNs) while NFkB and AP-1 induce other pro-inflammatory cyotkines. The activated transcription factors also induce co-stimulatory ligand expression, as well as anti-viral/anti-bacterial gene products

Full size image

3.2 PRR Signaling Dictates CD8+ T Cell Priming, Recruitment, and Function During Viral Infection

Leveraging antitumor functions of CD8+ T cells to eliminate malignant cells in an antigen-specific manner is the goal of most cancer immunotherapy strategies. In a natural infectious setting, CD8+ T cells are enlisted to eliminate intracellular pathogens, e.g., viruses. Provided below, and depicted in Fig. 3.2, is an example of how a typical acute viral infection leads to priming of antiviral CD8+ T cells via the activation of PRRs. The concepts of antiviral CD8+ T cell priming and effector functions during an infectious process are analogous to events that must occur to enable priming and effector function of antitumor CD8+ T cells.

Fig. 3.2
figure 2

Innate immunity engages CD8+ T cells during viral infection. See text for stepwise details

Full size image

  • Step 1: Viral infection of epithelium occurs (Fig 3.2a).

  • Step 2: Local inflammation is induced after recognition of viral features (e.g. viral nucleic acids) by PRRs expressed on tissue resident macrophages, infected epithelium, or by other tissue resident innate immune populations (Fig 3.2b).

  • Step 3: PRR mediated inflammatory signals leads to surface expression of adhesion molecules on the local endothelium (Fig 3.2c).

  • Step 4: The induction of adhesion molecules on endothelium in concert with chemokines/cytokines facilitates recruitment of additional innate immune cells (Fig. 3.2d). Much of the innate inflammation at this stage functions to limit viral replication and spread, and may lead to killing of infected cells by innate immune populations including NK cells, neutrophils, and macrophages. In addition, during this process conventional DCs consume antigens in the infected site, while receiving activating signals from inflammatory cytokines from infected cells and/or direct PRR signaling (Fig 3.2d) that induces their ‘activation’—which includes the induction of co-stimulatory signals.

  • Step 5: Antigen-bearing DCs then migrate to the draining lymph node or other lymphatic organ to prime and expand populations of CD8+ T cells recognizing viral antigen from the infection site (Fig 3.2e).

  • Step 6: Activated CD8+ T cells chemotract to recognize and kill remaining infected cells, and inflammation from the infection induces antigen presentation machinery and stress signals in infected cells that further enable T cell mediated killing. A pool of memory T cells persist after the infection is cleared for future pathogen recognition and elimination (Fig 3.2f).

Notably variations in the routes by which these steps occur are pathogen and tissue specific; alternate modes of CD8+ T cell priming have been demonstrated, e.g. via antigen transfer between migratory vs lymph node resident DCs [5, 6]; and other mechanisms of antigen transfer at sites distant from the infection may occur [7]. These processes, originally defined in the context of natural viral infection [8], have been shown to largely apply to immune surveillance, that is, the recognition and elimination of malignant cells by the immune system.

3.3 The Cancer Immunity Cycle and PRR Signaling

The host immune system recognizes malignant cells on the basis of protein coding genetic mutations; abnormal post-translational modifications; aberrantly expressed proteins (e.g., cancer-testis antigens); and in some cancer types, oncogenic viral proteins. Recognition of such antigens is mediated by cell surface MHC class I for CD8+ T cells, and surface MHC class II for CD4+ T cells, with CD8+ T cells typically being the primary antitumor effectors during immune surveillance. While co-evolution between malignant cells and the host immune system eliminates immunogenic malignant cells and results in outgrowth of ‘immunoedited’ tumors that are less immunogenic [9], and heterogeneity in the expression of tumor associated antigens is common [10], the success of immune checkpoint blockade in several tumor types implies that other potentially reversable regulatory nodes prevent immune recognition and destruction of solid tumors.

The cancer immunity cycle outlines established steps by which antitumor T cells can become activated endogenously to eliminate malignant cells [11] (Fig. 3.3). Dying cancer cells release antigens that are taken up by dendritic cells (DCs) to be loaded on to MHC-class I or II. Tumor antigen presenting DCs present antigen along with co-stimulatory signals to T cells, typically within the tumor draining lymph node. If the appropriate co-stimulatory signals along with cognate antigen are delivered to tumor antigen-specific T cells at this step, activated tumor-specific T cells may traffic to the tumor site, recognize tumor antigen-expressing malignant cells, and mediate killing of malignant cells through several mechanisms. The cytotoxic mechanisms of T cells include the release of perforin and granzymes, apoptotic signals (FAS-ligand), cytokines that mediate apoptotic signaling, e.g., TNF, as well as cytokines that induce upregulation of MHC class I and antigen processing machinery in tumor cells, particularly IFN-γ.

Fig. 3.3
figure 3

adapted from Chen and Mellman Immunity 2013 [11] with permission. Red boxes denote mechanisms by which PRR signaling impacts cancer immune surveillance. (1) Cancer cells routinely die due to genotoxic stress, chemo/radiation, nutrient deprivation, hypoxia, and other reasons; leading to the release of tumor antigens. (2) Migrating antigen presenting cells, including migratory conventional dendritic cells capable of cross presenting antigens to CD8+ T cells (see Fig. 3.4), take up antigen in the tumor microenvironment to be processed and presented on MHC-class I or II. PRRs potentiate this step by increasing antigen presentation on DCs, increasing antigen uptake (e.g. via calreticulin and HSP surface expression on apoptotic cells) and potentially recruiting additional migratory DCs into tumors. (3) Within the draining lymph node or other secondary lymphoid organ, tumor antigen loaded DCs present antigen to T cells, leading to their activation. PRR signaling induces the expression of co-stimulatory signals on DCs to potentiate T cell priming; see Fig. 3.4 for detailed explanation. (4) T cells traffic to the site of the tumor by surveying endothelial ligand (e.g. ICAM-1 and VCAM-1) expression and chemokine signals, and (5) infiltrate the tumor tissue. PRRs enhance endothelial cell T cell adhesion ligand expression and chemokine secretion from the tumor site to facilitate T cell infiltration. (6) T cells recognize cognate tumor antigen presented on tumor cells; PRRs facilitate this process by inducing inflammation that causes induction of antigen presentation machinery in cancer cells. (7) T cells kill cancer cells expressing their cognate antigen via granzymes and perforin, FAS ligand, and secretion of cytotoxic cytokines; PRRs induce inflammation that enhances antitumor T cell function and cytotoxicity

PRR signaling supports the cancer immunity cycle through multiple mechanisms,

Full size image

However, several factors determine whether tumor-specific T cells will ultimately become tolerized and anergic after DC mediated antigen presentation, whether sufficient signals enable trafficking of tumor-reactive T cells to the tumor site, as well as whether antitumor T cells can function within the immune subversive tumor microenvironment (TME). These issues are dictated by that status of the innate immune system during the cancer immunity cycle, including that of DCs and tumor associated macrophages/myeloid derived suppressor cells (MDSCs). PRRs play a multifaceted role in determining the pace, and efficacy of the cancer immunity lifecycle. PRR signaling enhances antigen presentation and co-stimulatory signals on antigen presenting cells (Fig. 3.4), culminating in more effective priming of tumor antigen-specific T cells. In addition, intratumor activation of PRRs induces chemokines that facilitate the recruitment of newly primed antitumor T cells, and further supports their function by enhancing tumor antigen presentation and an inflammatory milieu that potentiates T cell effector functions. For detailed explanation of the role of PRR signaling at each step of the cancer immunity cycle, see Fig. 3.3. Accordingly, an emerging clinical strategy aimed at rectifying stalled cancer immune lifecycles in patients is that of targeting the activation of PRRs within the TME to provoke the expression of: co-stimulatory signals on DCs during antigen presentation, T cell recruiting chemokines within the tumor site, and inflammation within the TME that supports antitumor CD8+ T cell function.

Fig. 3.4
figure 4

Dendritic cell (DC) activation is dictated by PPR signaling, which enables CD8+ T cell cross-priming. An example of cross presentation is shown, which occurs via loading of engulfed exogenous antigen onto MHC-class I for presentation to CD8+ T cells, typically by cDC1s (CD103+, BATF3+ in mice; CD141+ in humans). In the context of viral infection, PRRs expressed by viral antigen presenting DCs sense PAMPs and DAMPs at the site of inflammation/infection and receive signals from locally produced cytokines. These signals lead to upregulation of antigen processing and presentation machinery (signal 1), induction of co-stimulatory ligands including the CD28 ligands CD80 and CD86 (signal 2), and secretion of cytokines like IL-12 and type I IFNs that lead to further activation and differentiation of the cognate antigen-specific CD8+ T cell (signal 3). Upon receiving these signals, CD8+ T cells become activated and can traffic to the site of infection to eliminate virally infected cells. Similarly, during the cancer immunity cycle, migratory cDCs take up tumor associated antigens, traffic to the draining lymph node, and present exogenously acquired antigen on MHC-class I to T cells. As occurs during viral infection, PRR signaling bolsters the efficiency of tumor antigen uptake, processing, and presentation on MHC-class I on DCs (Signal 1); the induction of co-stimulatory ligands expressed on the DC surface (Signal 2); as well as the induction of pro-inflammatory cytokines required for optimal antitumor CD8+ T cell priming (Signal 3)

Full size image

3.4 Endogenous Activation of PRRs in Cancer

Endogenous signals from cells that are stressed or dying necrotically can induce DC activation in the absence of foreign pathogen associated features [12]; lending explanation for how T cell priming and activation in the contexts of spontaneous antitumor immunity, transplantation rejection, and/or autoimmunity occurs. Recent work has documented the importance of several PRRs in the context of tumor biology that lead to DC activation, and generalized inflammation within tumors. These include the following PAMPs and DAMPs, which represent only a subset of relevant documented endogenous PRR ligands:

  1. 1.

    Double stranded DNA (dsDNA) from dying tumor cells can be recognized by cGAS-STING, particularly within DCs that take up debris from dead tumor cells. STING signaling culminates in IFNβ/antiviral responses that mediates activation of systemic antitumor T cell immunity and tumor regression [13, 14].

  2. 2.

    High-mobility-group box 1 (HMGB1) is a nuclear protein that is released during cell death, including after chemotherapy/radiation, to engage TLR4 on DCs [15].

  3. 3.

    Cell surface calreticulin facilitates phagocytosis of apoptotic tumor cells by DCs and macrophages and determines the immunogenicity of phagocytosed cells [16,17,18].

  4. 4.

    Heat Shock Proteins (HSPs) released from dying cancer cells are widely reported to bind to TLRs 2 and 4 to induce inflammation [19,20,21,22,23,24].

  5. 5.

    Endogenous retroviruses have been shown to reactivate in some cancers presumably due to epigenetic dysregulation or loss of innate signaling in malignant cells [25], and can induce TLR and RLR signaling due to cytoplasmic presence of replicating retroviral RNA [26].

  6. 6.

    Uric acid/monosodium urate crystals, a byproduct of purine metabolism that causes gout, have been shown to induce activation of the NLRP3 inflammasome [27], TLR2, and TLR4 [28]. Uric acid mediates DC activation [29].

  7. 7.

    The tumor microbiome has recently been defined, demonstrating evidence of microbial presence (e.g., intracellular bacteria and viruses) in various tumor types at baseline [30, 31]. The presence of such microbes is likely to impact endogenous PRR signaling in tumors, though this remains to be determined.

Thus, PRRs in tumors are not inert, but rather may recognize features associated with cell death; tissue damage; and in some cases, endogenous pathogens. Collectively, their activity in cancer likely explains how spontaneous antitumor T cells are primed to eliminate malignant cells. In contrast, a lack of co-stimulation from DCs to antitumor T cells is frequent in cancer [32, 33], causing tolerance and suppression as opposed to activation. This contradiction may be due to insufficient PRR signals from endogenous PAMPs/DAMPs, tolerance/desensitization of PRR signaling due to chronic PAMP exposure, or other mechanisms of tumor mediated immune suppression. Thus, given their roles in orchestrating co-stimulatory signal expression in antigen presenting cells, targeting PRRs to induce inflammation compatible with T cell priming and co-stimulation is a therapeutically viable strategy to engage antitumor CD8+ T cell immunity.

3.5 Engaging PRRs for Cancer Immunotherapy

3.5.1 Inducing Innate Inflammation in Tumors: A Historical Perspective

The first widespread medical use of a PRR activator in cancer is that of Coley’s toxin [34, 35] in the late 1800s, though the use of pathogens for cancer therapy, as well as anecdotal correlations of pathogen infection and spontaneous tumor regression, was documented much earlier [34, 36, 37]. Based upon clinical case review of a patient that experienced sarcoma tumor regression after bacterial infection at the tumor site, William Coley tested if bacterial infection of sarcomas may mediate tumor regression in patients. After initially using live bacterium, Coley switched to inactivated bacterium; which caused regression in some patients. This cocktail of inactivated bacteria became known as ‘Coley’s toxins’, but the approach was ultimately overshadowed by advances in radiation therapy, and suffered from limitations in standardizing the treatment [34].

In the early 1900s Mycobacterium bovis was isolated from a cow with tuberculosis mastitis. Laboratory passaging in bovine bile (to prevent clumping) led to a loss of virulence, and the strain of M Bovis was named Bacillus of Calmette and Guerin (BCG) after the scientists that developed the strain [38]. Coincidentally, Tuberculosis infection was noted to be associated with a lower frequency of cancer [39], raising the possibility for using M bovis for cancer therapy. In 1969 the first report of BCG’s use as a cancer therapy was reported by Mathe et al.in the treatment of lymphoblastoid leukemia where encouraging results were reported [40]. The first clinical trial of BCG for bladder cancer was published in 1976 [41] where a decrease in recurrence of superficial bladder cancer was observed. These observations were confirmed in 1980 [42], spurring widespread use of BCG as a intravesicular therapy for bladder cancer. BCG was FDA approved for the treatment of bladder cancer in 1990, and represents the first approved cancer immunotherapy. BCG mediates innate inflammation that engages CD4+ and CD8+ T cells with several TLRs being shown to mediate the initial innate response, including TLRs 2, 4, and 9 [43]. The success of BCG, along with discoveries on the role of PRR signaling in mediating immune surveillance, led to further studies applying intratumor PRR activators for cancer immunotherapy in several cancer types.

3.5.2 Non-Infectious Engagers of PRRs for Cancer Immunotherapy

PRR activating PAMPs have shown preclinical promise in engaging systemic antitumor immunity. In addition, therapies that evoke PRR signaling through indirect means are also being explored. Clinically tested approaches to engage immune surveillance through targeted PRR activation using non-replicating, non-infectious PAMPs are described below according to their PRR specificities.

3.5.2.1 TLR Agonists

The toll like receptors were first discovered in Drosophila, and later confirmed to induce innate inflammation in mammalian systems [8]. TLR signaling culminates in activation of TBK1 and IKKα/β to induce type I IFNs and NFkB dependent gene expression, respectively (Fig. 3.1).

TLR3

The double stranded RNA mimetics Poly I:C and Poly A:U, and derivatives thereof, have been widely tested as cancer immunotherapies in several solid tumor indications. Poly A:U was tested in the 1980–90 s wherein it was shown to extend relapse free survival after systemic delivery in breast cancer patients [44], but showed minimal efficacy in melanoma [45], and was associated with less favorable survival after systemic delivery in colorectal cancer patients [46].

The double stranded RNA mimetic, Poly I:C, effectively induces type I IFN in several tumor associated cell types and mediates generation of Th1 responses in mice [47, 48]. A poly-L-lysine stabilized version of Poly I:C in carboxymethyl-cellulose, Poly ICLC (Hiltonol), also engages MDA5 activation [49, 50] and has been tested in several trials. Poly ICLC has been delivered intramuscularly to boost systemic type I IFN responses [51, 52], as well as via intratumoral routes [53,54,55]. Poly ICLC was well tolerated in early trials, but limited efficacy as a monotherapy was reported overall. Ongoing work demonstrating potential clinical benefit focuses on combining Poly ICLC with other modalities including FLT3L, radiation, and PD-1 blockade [56]. Poly ICLC is also being used as a personalized cancer vaccine adjuvant, where sustained antitumor T cell responses were demonstrated [57].

TLR4

After the clinical use of Coley’s toxin in sarcomas, it was proposed that TLR4 activation via bacterial polysaccharides mediated that anti-sarcoma effects of Coley’s toxin in mice [58]. Lipopolysaccharide (LPS), also known as endotoxin, is the canonical bacterial polysaccharide used to activate TLR4 signaling in laboratory studies. Inducing TLR4 signaling leads to robust myeloid cell activation, particularly macrophages, and in preclinical models has been shown to induce robust antitumor effects [59,60,61]. The first clinical trial of LPS in cancer patients occurred via intravenous injection concomitant with ibuprofen to prevent inflammatory side effects, and showed induction of pro-inflammatory cytokines in the sera (TNF, IL-6 and MCSF) with moderate antitumor activity observed in 2 patients with colorectal cancer [62]. A follow-up trial of systemic LPS delivery showed only moderate antitumor efficacy [63]. Systemic toxicities associated with inflammation were a common issue for trials using LPS.

Usage of the lipid A subunit of LPS was later shown to induce antitumor activity with a more favorable toxicity profile [59, 64, 65]. Lipid A isolated from Salmonella, called monophosphoryl lipid A (MPL) was subsequently tested in cancer patients intravenously and was found to have minimal antitumor efficacy [66]. Several derivatives of MPL have been clinically tested in cancer patients with inconsistent or lacking indication of antitumor efficacy [67,68,69]. Despite limited efficacy of various MPL based strategies delivered intratumor, subcutaneous, or intravenously, MPL succeeded as an adjuvant for the HPV vaccine Cervarix® and was FDA approved for this use in 2009 [70]. Recent clinical efforts include the use of the TLR4 activating glycolipid (GSK1795091) in combination with an activating OX40 antibody, and a synthetic MPL mimetic (GLA-SE) in combination with radiation therapy [71, 72].

TLR7/8

Imiquimod is a small non-nucleoside TLR7/8 activator that originally demonstrated utility as an antiviral agent in preclinical models in the 1980s [73,74,75]. The antiviral effects observed were dependent upon mediating inflammation as opposed to direct action on viruses [74, 76,77,78,79,80,81,82]. Initial clinical trials delivering oral imiquimod in cancer patients failed to demonstrate efficacy beyond induction of inflammation [83, 84]. However, topical application of imiquimod cream for actinic keratosis [85,86,87,88,89,90,91] and basal cell carcinoma [92,93,94,95,96,97,98,99,100] was efficacious, and was FDA approved in 2004 for these indications. Several studies in various topical pre-cancerous and cancerous disease have since been conducted [67], with more recent testing occurring in breast cancer, melanoma, and other solid tumors alone or in combination with immune checkpoint blockade.

R848 (Resmiquimod) and motolimod (VTX-2337) are TLR7/8 agonists that are currently being clinically explored for cancer immunotherapy, in pre-cancerous actinic keratosis and head and neck squamous cell carcinoma, respectively [101]. Other agonists targeting TLR7 and TLR8 are currently being tested alone or in combination with immune checkpoint blockade in various solid tumors [102].

TLR9

TLR9 agonists induce potent type I IFN responses from plasmacytoid DCs (pDCs), and generally mimic unmethylated CpG DNA. Importantly, murine cell-type expression patterns of TLR9 is distinct from that of humans, with TLR9 largely being expressed in human pDCs and B cells, while murine expression of TLR9 is more ubiquitous in macrophage and DC populations [103]. Numerous clinical trials using TLR9 agonists have been conducted as monotherapy studies in solid tumors, exhibiting manageable safety profiles despite association with cytokine release syndrome related to IFN mediated inflammation [104]. Clinical efficacy signals in monotherapy trials have been limited, with more promising signals being observed when combined with other modalities [104].

Lefitolimod (MGN1703) was tested in two phase II trials in small cell lung cancer and metastatic colorectal cancer with subcutaneous delivery and did not meet survival endpoints [105, 106]. A phase III trial was conducted in a subgroup of metastatic colorectal cancer patients identified in the phase II trial, where negative results were posted.

Tilsotolimod (IMO-2125) has been tested in multiple solid tumors, most extensively in melanoma. A phase I/II trial in anti-PD-1 refractory melanoma showed evidence of efficacy of Tilsotolimod in combination with ipilimumab (anti-CTLA-4) or pembrolizumab (anti-PD-1) with an objective response rate (ORR) of 22% (News release, Idera Pharmaceuticals, April 21, 2020 press release). A follow-up phase III trial of Tilsotolimod in combination with ipilimumab in anti-PD-1 refractory melanoma was conducted but did not meet objective response rate endpoint (Idera pharmaceuticals, March 18, 2021 press release). A phase III trial in microsatellite stable colorectal cancer in combination with ipilumimab and nivolumab (anti-PD-1) is ongoing.

SD-101 has shown abscopal effects in indolent lymphoma patients after intratumor administration [107] and demonstrated an objective response rate (ORR) of 78% in treatment-naïve and 15% ORR in PD-1 refractory melanoma in combination with pembrolizumab [108]. Intratumor SD-101 in combination with pembrolizumab and paclitaxel in HER2-negative breast cancer showed non-significant improvement in pathological complete responses [109]. Several trials of SD-101 as an intratumor therapy are ongoing in combination with other modalities in melanoma, breast cancer, prostate cancer, and lymphoma.

CMP-001 is a virus like particle comprised of bacteriophage capsid with a CpG oligodeoxynucleotides. The drug is taken up by pDCs via FCgamma receptor anti-bacteriophage antibodies that bind the virus like particle leading to robust type-I IFN induction [110, 111]. In anti-PD-1 refractory melanoma patients, intratumor CMP-001 in combination with pembrolizumab achieved an ORR of 25% associated with abscopal effects noted [112, 113]. Ongoing clinical trials are testing CMP-001 in melanoma, head and neck squamous cell carcinoma, and lymphoma.

TLR9 agonists remain a very active area of clinical pursuit, particularly with newer routes of delivery, e.g., in the aforementioned case of bacteriophage-antibody mediated delivery via CMP-001; NZ-TLR, which uses a cold isostatic pressing to encapsulate a TLR9 agonist that permits extended release following intratumor injection; and AST-008, a spherical nucleic acid-based nanomaterial TLR9 agonist [104].

3.5.2.2 RLR Activation

The RLRs MDA5 and RIG-I are cytoplasmic sensors of viral RNA that have recently gained attention as potential targets for cancer immunotherapy.

MDA5

MDA5 recognizes long dsRNA in the cytosol, culminating a distinct type I IFN dominant activation of macrophages and DCs and cell death signaling in cancer cells [114,115,116]. Poly ICLC activates both TLR3 and MDA5 as mentioned in Sect. 3.5.2.1, with MDA5 activation being linked to the potent Th1 antitumor activity observed by Poly ICLC ([49, 50] (see TLR3 agonist description above for clinical status of Poly ICLC). An emerging route to target MDA5 activation is via synthetic RNA viruses and indirect reactivation of endogenous retroviruses (ERVs) using epigenetic modulators.

Synthetic positive-sense RNA viruses and replicons are commonly engineered from Semliki Forest virus, Sinbus virus, or Venezuelan equine encephalitis virus and delivers self-replicating RNA into the cytosol of cells, which can be recognized by both MDA5 and RIG-I [117]. Results from a phase I clinical trial testing a Simliki Forest virus-based HPV vaccine in HPV induced cancers efficiently induced HPV antigen-specific T cells and was well tolerated [118]. Synthetic coxsackievirus A21 RNA that engage MDA5 are currently in development for clinical use [119]. How well replicons/synthetic viral RNA engage MDA5 and other PRRs relative to oncolytic viruses/natural virus infection (see Sect. 3.5.3 below) remains unknown.

ERVs comprise up to 8% of the human genome [120] where they typically remain inactive, but have been shown to be reactivated in various cancer types [25]. Usage of demethylating agents was initially proposed to mediate antitumor effects by inducing expression of tumor suppressor genes [121]. However, several studies have shown that 5-aza-2-deoxycytidine, a DNA methylation inhibitor, causes re-expression of ERV gene products that induce dsRNA recognized by MDA5 in cancer cells, sensitizing tumor-bearing mice to anti-CTLA-4 therapy [122, 123]. Given that immunotherapy success is associated with ERV gene expression in tumors [124,125,126], it is possible that optimizing DNA demethylating agents for induction of ERV mediated MDA5 signaling will enhance immune checkpoint blockade therapy. DNA demethylating agents have been tested extensively in the clinic [127, 128], however it is unclear whether MDA5 engagement occurred and/or contributed to therapy effect.

RIG-I

In contrast to MDA5, RIG-I recognizes short dsRNA as well as 5’-ppp-RNA that lacks a 7-methylguanosine cap on the 5’ end of RNA commonly added to endogenous mRNAs. Several RIG-I agonists have been developed and are in clinical testing.

MK-4621 is a 5’-ppp synthetic RNA oligonucleotide that was delivered intratumoral in various solid tumor types where an interim analysis showed a favorable safety profile and induction of serum chemokine levels [129]. A second study testing MK-4621 complexed with JetPEI™ and pembrolizumab is also ongoing [130]. CV8102 is a single stranded, uncapped RNA complexed with cationic peptides that activates RIG-I along with TLR7/8. This drug was tested by intratumor injection in solid tumors alone or in combination with PD-1 blockade wherein the drug was well tolerated and early responses were observed [131]. GEN0101 is a drug composed of inactivated Sendai virus particles that engage RIG-I and have shown immunological responses to treatment, declines in prostate-specific antigen, and potential disease stabilization after intratumoral and subcutaneous injection in castration-resistant prostate cancer patients [132, 133]. Several other studies of GEN0101 have been conducted in melanoma and mesothelioma, however results from these trials have not been reported [134].

3.5.2.3 STING Agonists

Due to the role of endogenous STING signaling in tumors leading to spontaneous antitumor immunity [13], STING agonists have gained considerable attention as intratumoral agonists for clinical cancer immunotherapy.

DMXAA was originally developed as an anti-vascular drug that was later found to activate TBK1-IRF3 signaling [135] via STING [136]. A large phase III study was conducted in non-small cell lung cancer patients, but was discontinued [137]. Other clinical efforts with this agent have failed to show compelling clinical responses. However, despite preclinical data indicating its ability to engage antitumor CD8+ T cell immunity, it was later found to only induce mouse STING signaling, and not that of humans [138], possibly explaining its lack of clinical activity.

Other clinical trials of intratumor delivered STING agonists are ongoing and include GSK3745417, MK-2118, MK-1454, BMS-986301, IMSA-101, ADU-S100 and E7766; most of which are being combined with immune checkpoint blockade. At the time of writing, biological activity of STING agonists has been reported in patients [139, 140], but the clinical efficacy of these agents remains to be reported. Further development of modified versions of STING agonists in preclinical settings are ongoing that include the development of orally available STING agonists [141, 142], as well as higher potency STING agonists [143, 144].

3.5.3 Infectious Agents as Engagers of PRRs for Cancer Immunotherapy

In addition to BCG described above, attenuated, replication-competent viruses and bacteria being clinically developed for cancer immunotherapy also engage PRR signaling. A potential advantage of intratumoral therapy with infectious agents versus that of targeted PRR engagement with PAMPs is that infectious agents generally engage multiple PRRs within the spatiotemporal context of a natural infectious process, possibly recapitulating a more natural T cell priming scenario. However, infectious agents derived from natural animal viruses and bacterium also typically mediate some level of innate or adaptive immune interference (e.g. suppression of antiviral signaling or antigen presentation) that evolved to ensure the successful lifecycle of the pathogen [145,146,147,148]. It remains to be determined which approach (non-infectious agonists vs infectious agents) will be more effective in engaging immune surveillance and controlling tumor growth. Beyond the PRR-engaging attributes of these agents, it also must be noted that oncolytic viruses also mediate killing of cancer cells, adding an additional dimension of anticancer and immunogenic activity.

3.5.3.1 Oncolytic Viruses

While dubbed ‘oncolytic’ due to selective toxicity observed by various attenuated viruses in cancer cell lines [149], the antitumor potential of using viruses for cancer therapy may largely be due to their ability to elicit antitumor CD8+ T cells through PRR activation [114, 150,151,152]. Diverse virus species have been developed for cancer immunotherapy, ranging from large DNA viruses to small RNA viruses, that have distinct tissue tropisms, viral replication strategies, and mechanisms of immune subversion. Thus, as with targeting distinct PRRs for cancer therapy, different virus contexts are likely to mediate antitumor efficacy through different routes, with differing efficiencies. Clinically tested viral cancer immunotherapies are described below, however, numerous virus contexts beyond these agents are being considered for future clinical testing.

  • Herpes Simplex Viruses (HSV). HSV is a dsDNA viruses that engage a number of PRRs including STING, TLR2, TLR3, and TLR9 [153]. Talimogene laherparepvec (T-VEC), an attenuated oncolytic HSV1 (oHSV) expressing GMCSF, is the only FDA approved oncolytic virus to date, which demonstrated a 16.3% durable response rate and 33% 5-year response rate in a randomized phase III clinical trial of melanoma [154]. An abscopal effect was noted, with regression of non-injected lesions occurring in some patients [155]. Other monotherapy clinical trials of oHSVs have shown evidence of efficacy similar to what has been observed for T-VEC in early stage clinical trials [156]. While initial observations in combination with immune checkpoint blockade suggested promise [157], a phase III clinical trial testing T-VEC combined with pembrolizumab was recently discontinued due to futility [158].

    The next generation of herpesvirus-based immunotherapies have been developed with intentions of improving oHSV immunotherapy efficacy, particularly in regards to preventing oHSV mediated disruption of antigen presentation (G47Δ [159]); enhancing oHSV toxicity selectively in cancer cells (rQNestin34.5v2 [160], RP1 [161]); enhancing IFN resistance of oHSV (ONCR-177 [162]); and ‘arming’ oHSV with PD-1 or CTLA4 blocking antibodies and immunostimulatory cytokines, particularly IL-12 (M032 [163], ONCR-177 [164], RP2 [165], MVR-T3011 [166]). Several of these agents are moving into early stage clinical trials in various solid tumor types as intratumorally delivered therapies.

  • Adenovirus. Adenovirus has a dsDNA genome and is recognized by TLR9 on pDCs[167], STING [168, 169], NOD like receptors [170, 171], with evidence for roles of other TLRs in vivo [172]. Immunogenic cell death (e.g., release of HMGB1, calreticulin, ATP, and HSP70) has also been proposed as a key mechanism driving the immunogenicity of oncolytic adenoviruses [173,174,175]. In 2005 China approved the replicating adenovirus H101 (Oncorine) for the treatment of nasopharyngeal carcinoma [176]. DNX-2401 is a modified adenovirus that selectively replicates in cancer cells with defective Retinoblastoma (Rb) and has shown promising phase I results in recurrent glioblastoma wherein 20% of patients surviving > 3 years [177]. A follow-up phase II study of DNX-2401 delivered at the time of biopsy in recurrent glioblastoma patients was conducted in combination with pembrolizumab, wherein 5/42 patients receiving the full DNX-2401 dose had confirmed responses [178]. A randomized phase III study is in planning [178]. Several ‘armed’ adenoviruses are in clinical testing in various indications; armed with GMCSF (CG0070 [179, 180] and ONCOS-102 [181]), immunostimulatory ligands CD40L and 41BBL (LOAd-703 [182]); hyaluronidase (to facilitate viral spread and CD8+ T cell recruitment within the tumor, VCN-01 [183]); IL-12 (AD5-yCD/mutTKSR39rep-hIL-12, [184]); OX40L (DNX-2440 [185]); and CXCL9, CXCL10, and IFN-α (NG-641 [186]). In addition, combination strategies of modified oncolytic adenoviruses with CAR T cell therapy (CAdVec) and chemoradiation (Colo-AD1) are being pursued [187].

  • Poxviruses. Poxviruses are large dsDNA viruses (130–300 Kb) that have sophisticated replication strategies and mechanisms to evade viral elimination by the host immune system [188]; attenuated vaccina viruses, are the most extensively tested oncolytic poxviruses. Poxviruses are recognized by several PRRs, including TLR2, TLR6, MDA5, and the NALP3 inflammasome [189, 190]. Interestingly, UV and heat inactivated Vaccina virus was shown to mediate stronger innate inflammation through STING signaling compared to replicating Vaccina, possibly reflecting strategies by which Vaccina interferes with innate signaling [191]. Pexa-Vec (JX-594), a Vaccinia virus, was tested as an intratumor therapy in hepatocellular carcinoma in a phase II clinical trial where evidence of disease control was reported [192]. However, a follow-up study revealed lack of overall survival benefit in this patient population [193]. A trial of Pexa-Vec in colorectal cancer was pursued with immune checkpoint inhibitor combination, but failed to show a significant improval in response [194]. As with other DNA viruses used for virotherapy, a current emphasis on arming poxviruses is driving ongoing clinical efforts, including with GM-CSF, chemokines, IL-15 and PD-1 blocking antibodies [195]. GL-ONC1 and vvDD are vaccina viruses that were delivered intravenously [196, 197], other studies in solid tumors are ongoing and evaluation of antitumor effects have not yet been reported. Myxoma viruses are also being developed for virotherapy in preclinical settings [195].

  • PVSRIPO. PVSRIPO, a (+)stranded RNA picornavirus, is the live attenuated type I Sabin strain of Polio with exchange of the Sabin Internal Ribosomal Entry Site (IRES) with that of human rhinovirus type II [198, 199]. This substitution neuroattenuates the virus, but does not impair its ability to kill malignant cells [198]. PVSRIPO requires poliovirus receptor (PVR) expression for viral entry, which is highly expressed on both antigen presenting cells and malignant cells [114, 198, 200]. PVSRIPO infection activates MDA5, leading to a sustained type I/III IFN dominant IFN signature in tumor-associated macrophages and dendritic cells that culminates in antitumor CD8+ T cell immunity [114, 115, 201]. Importantly, in preclinical models the antitumor efficacy of PVSRIPO was primarily dependent upon viral infection of TME constituents as opposed to malignant cells, indicating that PVSRIPO may function primarily as an engager of MDA5 within the TME [114]. A phase I clinical trial in recurrent GBM demonstrated a 21% survival rate at 36 months, relative to 4% survival in an eligibility criteria-matched historical control cohort of patients [202]. A small phase I trial in anti-PD-1 refractory melanoma demonstrated antitumor responses in both injected and non-injected lesions in 4/12 patients, with 6/12 patients resuming immune checkpoint blockade after PVSRIPO having durable disease control at 18 months of follow-up [203]. Ongoing clinical studies are focused on combining PVSRIPO with PD-1 blockade in melanoma, GBM, and other solid tumors [203,204,205].

  • Reovirus. Reoviruses are segmented dsRNA viruses with a long history of preclinical investigation backing its utility as an immunovirotherapy agent [206, 207]. Reovirus is recognized primarily by RIG-I and MDA5 [208]. The antitumor efficacy of Reovirus is independent of viral replication in preclinical models [209], implying that PRR recognition occurs upon viral entry, leading to antitumor CD8+ T cell priming [210]. Reolysin (aka pelareorep) has been delivered both intravenously and intratumorally in clinical trials. An initial phase I study observed local tumor responses in 7/19 patients, with one complete response in advanced solid tumors [211]. Phase II trials in combination with chemotherapy for malignant melanoma [212], breast cancer [213], non-small cell lung cancer [214], head and neck cancer [215], metastatic pancreatic cancer [216] have been conducted with some indication of efficacy in subsets of patients. Recent work has demonstrated that intravenously delivered Reovirus reaches brain tumors in patients and induces PD-1/PD-L1, possibly indicating its potential use as a systemic agent in combination with immune checkpoint blockade [217]. As with other oncolytic viruses, the ongoing focus of current Reovirus virotherapy is focused on combining with other immunomodulatory agents [218].

  • Coxsackievirus A21 is a (+)stranded RNA picornavirus primarily sensed by MDA5 [219], that has been clinically tested in both intravenous and intratumoral contexts as V937 (aka CAVATAK). Indications tested include non-muscle-invasive bladder cancer [220], in which tumor associated inflammation was observed, and melanoma [221], wherein 43.2% of patients had progression free survival at 1 year post treatment. Ongoing studies combining V937 and immune checkpoint blockade are being conducted in unresectable melanoma, and early results indicate the combination of V937 and pembrolizumab has a 47% ORR [222].

  • Vesicular Stomatitis Virus (VSV) is a (−) stranded RNA virus that was developed as an oncolytic virus due to the lack of type I IFN mediated suppression of attenuated VSV replication in human cancer cells [223]. VSV is recognized by RIG-I [224] and TLR7 [225]. However, VSV vectors are capable of causing neurological disease in non-human primates [226]; thus an interferon-β expressing VSV (VSV-IFNβ) was developed to restrict VSV replication beyond normal cells and was found to not cause neurotoxicity in non-human primates [227,228,229]. VSV-IFNβ was further modified with expression of a sodium iodide symporter (NIS) to enable imaging. Ongoing clinical studies in various indications include IV infusion in multiple myeloma, T cell lymphoma, and acute myeloid leukemia [230].

  • Measles Virus is a (−) stranded RNA virus that was originally proposed as an oncolytic virus candidate due to case reports of measles infection being linked to tumor regression [231]. Measles is recognized by MDA5 and RIG-I [232], however it is reported to intercept RLR recognition in antigen presenting cells [233]. Indeed natural (wildtype) measles infection also suppresses adaptive immunity [234]. The live attenuated vaccine strain of Measles (Edmonston-Zagreb strain) has been developed for cancer immunotherapy and tested in early stage clinical trials of T cell lymphoma [235], ovarian cancer [236], glioblastoma, breast cancer, head and neck squamous cell carcinoma, malignant peripheral nerve sheath tumors, bladder cancer, and multiple myeloma [237]. A NIS-expressing version of measles was also generated and tested in patients after intravenous administration [238].

  • Newcastle’s Disease Virus (NDV) is a (−) stranded RNA virus recognized primarily by RIG-I [239, 240] that naturally infects chickens. NDV has been shown to mediate both oncolysis and type I IFN-dependent priming of antitumor T cells in preclinical models [150, 241]. Several clinical trials used NDV-treated oncolysate, or lysed cancer cells, for vaccination in cancer patients [242], most of which were in melanoma where improved overall survival was demonstrated relative to historical controls. A phase III clinical trial demonstrated longer survival after NDV-pulsed autologous vaccine compared to surgical resection alone in colorectal cancer patients [243]. The MTH-68 strain of NDV was tested in various advanced cancer types in small cohorts of patients, including glioblastoma, with potential evidence of efficacy after intravenous administration [244, 245]. The PV701 strain was likewise tested intravenously in small cohorts patients with advanced cancers, documenting some objective responses [246, 247]. Extension of these studies have been complicated by changing regulatory guidelines restricting the use of NDV strains [241]. Ongoing clinical efforts to test NDV include a GMCSF expressing NDV variant (MEDI5395) being tested in various advanced cancers in combination with durvalumab (anti-PD-L1 antibody) [248].

3.5.3.2 Intracellular Bacterium

  • Bacille Calmette-Guerin (BCG) is standard of care therapy for non-muscle invasive bladder cancer, and was the first FDA approved cancer immunotherapy. See Sect. 3.5.1 for a description of the use of BCG in cancer therapy.

  • Listeria monocytogenes is a gram positive, intracellular bacterium that causes listeriosis, a foodborne illness. Listeria is recognized by TLR2, TLR5, NOD-like receptors, and STING [249]. Strains of Listeria have been developed for use as cancer vaccine vectors, with its intracellular lifecycle being an asset to deliver tumor associated antigens and engage antitumor T cell responses [250]. Listeria vaccine clinical trials have been conducted via intravenous delivery in pancreatic cancer, against mesothelin (CRS-207 [251]); in cervical cancer against HPV antigens [252, 253], and in mesothelioma against mesothelin (CRS-207 [254]). Encouraging objective responses have been observed in early stage clinical trials of mesothelioma and cervical cancer; however a phase III trial of Listeria E7 vaccine (AIM2CERV) in cervical cancer was closed by the sponsor [250], and CRS-207 development was recently discontinued after a failed lacking activity in combination with pembrolizumab [255]. Several Listeria-based approaches are in development with ongoing clinical trials.

3.6 The Role of Type I IFN in Mediating the Antitumor Efficacy of PRR Agonists

Type I IFNs are critical toward engaging DC priming of antitumor T cells [256]. Indeed, the efficacy of several PRR activators (both non-infectious agonists and infectious agents) has been shown to be dependent upon eliciting type I IFN signaling in tumors, including: Poly IC/Poly IC-LC [49, 114, 257, 258], RLR agonists [50, 259], TLR7/8 agonists [260], TLR9 agonists [261], STING agonists [262,263,264], PVSRIPO [114], and NDV [150]. IFNAR signaling both primes DC differentiation and expression of costimulatory ligands [47], while also boosting cytolytic function of antitumor CD8+ T cells locally [114, 265, 266]. However, it is critical to note that out-of-context type I IFN does not recapitulate the antitumor efficacy of broader signals delivered by type I IFN during PRR signaling [114, 267], which encompasses a myriad of other pro-inflammatory signals coinciding with type I IFN (Fig. 3.1). Indeed, while exogenous type I IFN treatment in cancer patients has shown some activity in the clinic, the efficacy of type I IFN treatment of tumors/cancer patients was limited [268]. PRR agonists, either infectious or non-infectious, offer potential to contextualize type I IFN signaling, and its T cell engaging capacity, within an inflammatory milieu supporting the production of chemokines, other DC/T cell modulating cytokines, and induction of pro-inflammatory signals within the TME that support CD8+ T cell effector functions. Whether this potential is fully realized clinically remains to be determined.

Yet, IFN signaling also mediates cancer cell chemoradiation resistance and induction of immune checkpoints that prevent antitumor T cell function [269,270,271]. Moreover, type I IFN contributes to T cell exhaustion and dysfunction during chronic viral infection [272], and type I IFN signaling in CD4+ T cells has been shown to negatively associate with immunotherapy response [273]. The context of IFN signaling may well determine whether it promotes or desensitizes antitumor immunity [274]: tumors with active IFN signaling at baseline may be resistant or non-responsive to PRR agonist therapy; and due to the role of IFNs in inducing immune checkpoint ligands [275], combination strategies to mitigate such negative feedback may be necessary to empower the antitumor effects of PRR engaging therapies. Indeed, most clinically pursued PRR engagers have been shown to induce PD-L1 and other immune checkpoint ligands, and are potentiated by immune checkpoint blockade in preclinical models [49, 114, 115, 150, 157, 267, 276].

3.7 Comparison of PRR Activators to Other Immunotherapies and Their Utility in Combination

PRR engaging immunotherapies intended to mediate in situ vaccination differ mechanistically from other anticancer modalities in several complementary ways. Complementary and distinct attributes of PRR activating immunotherapies compared to other established immunotherapy approaches are presented below:

  • Immune checkpoint blockade (ICB): Blockade of PD-L1, PD-1, CTLA4, and other immune inhibitory receptors function to resuscitate antitumor T cell function, and has achieved unprecedented clinical responses in immunogenic tumors with high mutation loads and/or oncogenic viral gene expression [277]. Generally, these modalities rely upon the presence of pre-existing antitumor T cells and are more efficacious in tumors that have higher baseline inflammation [278]. Inclusive of resistance mechanisms to ICB is that of innate immunosuppression, which limits co-stimulation, infiltration, and effector function of antitumor T cells [279]. In contrast, intratumoral therapy with PRR activators mediates innate inflammation within tumors, that enhances expression of co-stimulatory signals on antigen presenting cells, causes chemokine induction that enables trafficking of T cells to the site of the tumor, and directly bolsters the function of antitumor T cells, e.g., via type I IFNs [114, 280]. Numerous pre-clinical studies have shown that PRR activation within tumors leads to priming of antitumor T cells, which may broaden the potential of ICB therapy by bolstering antitumor T cell populations, while supporting their recruitment and function within the tumor microenvironment. Indeed, several studies have demonstrated synergy between PRR activators and ICB [281].

  • Cancer vaccines: Various cancer vaccine modalities have been developed including peptide vaccines, autologous dendritic cell vaccines, and mRNA vaccines. Indicating their clinical potential, prophylactic vaccination against HPV antigens has been remarkably successful in preventing cervical cancer [282]. Traditionally, cancer vaccines have been restricted to ‘shared’ tumor associated antigens common across numerous patients, e.g., HER2, EGFRviii, and MART1. With more recent feasibility of whole exome sequencing of biopsy tissue, personalized vaccines based upon patient-specific neoantigens are in development [57]. However, as with immune checkpoint blockade, cancer vaccines require that primed and expanded antitumor T cell populations induced by the vaccine traffic and function within the tumor. Moreover, these strategies require knowledge and accurate prediction of effectively presented, homogenously expressed, and targetable neoantigens. Notably, PRR activating adjuvants are used in cancer vaccines to enable priming and expansion of antitumor T cells in the periphery. In contrast, intratumoral delivery of PRR engaging therapies function to mediate vaccination using the tumor site in an antigen agnostic manner, by activating innate immunity and antigen presentation to prime T cells against antigens present within the tumor. Moreover, PRR agonists induce inflammation that enable trafficking and potentiation of antitumor T cell function. Intratumoral PRR agonist therapy is anticipated to complement cancer vaccines by enabling the recruitment, further tumor/tumor draining lymph node localized expansion of tumor antigen-specific T cells, and by providing inflammation in the tumor that supports antitumor T cell function.

  • Adoptive T cell transfer/CAR T cells: A direct route to bolster antitumor T cell populations in cancer patients is to deliver either expanded autologous antitumor T cells (or tumor infiltrating T cells) or autologous chimeric antigen receptor (CAR) T cells against specific tumor antigens. These approaches have shown promising antitumor efficacy in some cancer types [283, 284]. Distinct from a T cell-based approach to induce antitumor immunity in patients, intratumoral PRR activation leads to priming of T cells in the tumor bed and tumor draining lymph node while providing a supportive innate inflammatory framework for antitumor T cells to function. Intratumoral PRR activation has been shown to potentiate adoptive T cell therapy and CAR T cell therapy in pre-clinical models, primarily by enhancing recruitment of the ex-vivo expanded or engineered autologous T cells to the site of the tumor [285, 286].

3.8 The Future of PRR-Targeted Cancer Immunotherapies: Hurdles and Limitations

Beyond logistical regulatory and manufacturing issues, several hurdles remain for the success of PRR engaging immunotherapy to be realized. First, the optimal dosing of PRR engagers remains unclear, and is likely to be specific to each agent. For example, administration of higher doses of a STING agonist in mice impaired systemic antitumor immunity [287], implying that exaggerated activation of intratumor STING signaling may mediate a deleterious effect on antitumor T cell function. Whether this is true for other PRR engagers remains unknown. It also remains unclear as to which tumor types may benefit most from PRR engaging therapy: should immunologically quiescent tumors be targeted to enhanced intratumor inflammation and engage T cells? Or are immunologically active tumors more responsive to PRR-induced inflammation? As presented in this chapter, PRR agonists have been tested in both notoriously immunosuppressed tumors (e.g., glioblastoma) as well as immunogenic tumors (e.g., melanoma).

In addition, PRR-induced inflammation plays both anti- and pro-tumor roles [288]. For example, TLR3 signaling in the tumor microenvironment has been shown to enhance cancer metastases [288]; VEGF, matrix metalloproteinases, and other inflammatory features induced by PRR signaling may facilitate tumor vascularization; interferon responses induce APOBEC which can add to the evolutionary potential of cancer cells by increasing mutation rates [289]; PRR signaling promotes NFkB signaling, which can enable cancer cell survival and resistance to T cell mediated killing [290, 291]; and PRR signaling may exacerbate T cell exhaustion and dysfunction. In some respects, combination therapies like immune checkpoint blockade, anti-VEGF therapies, and other mechanisms may complement PRR engaging therapies to mitigate these effects. Overcoming and defining these limitations will be critical to optimize PRR activation for future cancer therapy.