The Course of Immune Stimulation by Photodynamic Therapy: Bridging fundamentals of photochemically-induced Immunogenic Cell Death to the Enrichment of T Cell Repertoire

Abstract

Photodynamic therapy (PDT) is a potentially immunogenic, and FDA-approved anti-tumor treatment modality that utilizes the spatiotemporal combination of a photosensitizer, light, and oftentimes oxygen, to generate therapeutic cytotoxic molecules. Certain photosensitizers under specific conditions, including ones in clinical practice, have been shown to elicit an immune response following photoillumination. When localized within tumor tissue, photogenerated cytotoxic molecules can lead to immunogenic cell death (ICD) of tumor cells, which release damage-associated molecular patterns and tumor-specific antigens. Subsequently, the T lymphocyte (T cell)-mediated adaptive immune system becomes activated. Activated T cells then disseminate into the systemic circulation and eliminate primary and metastatic tumors. In this review, we will detail the multistage cascade of events following PDT of solid tumors, that ultimately lead to the activation of an anti-tumor immune response. More specifically, we connect the fundamentals of photochemically-induced ICD with a proposition on potential mechanisms for PDT-enhancement of the adaptive antitumor response. We postulate a hypothesis that during the course of the immune stimulation process, PDT also enriches the T cell repertoire with tumor-reactive activated T cells, diversifying their tumor-specific targets and eliciting a more expansive and rigorous antitumor response. The implications of such a process are likely to impact the outcomes of rational combinations with immune checkpoint blockade, warranting investigations into T cell diversity as a previously understudied, and potentially transformative paradigm in anti-tumor photodynamic immunotherapy.

Graphical Abstract

Photodynamic therapy (PDT) is a potentially immunogenic anti-tumor treatment modality that utilizes the spatiotemporal combination of a photosensitizer (PS), light, and oftentimes oxygen, to generate therapeutic cytotoxic molecules. Upon light irradiation (PDT), tumor cells undergo cell-death pathways and release damage-associated molecular patterns (DAMPs) and tumor-specific antigens (TSAs) in the tumor microenvironment. Dendritic cells (DCs) internalize the TSAs and activate T cells, which undergo clonal expansion. As a result, tumor-reactive T cell repertoire is thought to be enriched. Activated T cells reach the primary or metastatic tumor sites via systemic circulation and eliminate the tumors.

INTRODUCTION

Photodynamic Therapy (PDT) is a photochemistry-based, non-thermal treatment modality used to treat numerous cancers, in addition to a range of non-cancer indications, including ophthalmological and dermatological conditions, amongst others. In the US, PDT is clinically-approved for esophageal and non-small cell lung cancers, and is being tested in clinical trials for several other types of cancers, including glioma, pediatric brain cancer, prostate cancer, head and neck cancer, cervical cancer, pancreatic cancer and oral cancer (1–3). PDT utilizes photochemical processes to generate reactive molecular species (RMS) via a light-responsive molecule, a photosensitizer (PS), upon irradiation with a wavelength of light that is specific to the PS. PS molecules are inherently non-toxic to cells unless a threshold concentration of RMS is produced following light activation. Owing to non-overlapping mechanisms of action, PDT has been demonstrated to synergize with a plethora of traditional chemotherapeutics, achieving efficacy against chemo-resistant tumors in pre-clinical studies, while avoiding overlapping modes of toxicity (4–8).

In this review, we will outline the course of multiplexed immune stimulation that can be induced by PDT and will discuss the connection between the fundamentals of photochemistry, molecular and cellular mechanisms of PDT, and the subsequent impact on immunogenic cell death (ICD). As outlined in Figure 1, the multifaceted progression from ICD to the recruitment of the innate immune system will be discussed, in addition to how PDT of tumor tissue leads to an eventual T cell-mediated adaptive response and prolonged anti-tumor immunity. The review will conclude with a speculative hypothesis on the role of PDT in enriching the T cell repertoire as a novel and unique mechanism of enhancing the adaptive antitumor immune response. This unchartered territory promises to expand the current understanding of the diverse roles of PDT on immunological control over tumor progression and provides avenues for exploring the synergy between emerging immuno-oncology approaches with PDT-induced clonal expansions in the diversity of tumor-specific T cells.

Figure 1.

Overview of this review article discussing the fundamentals of PDT and how they relate to the stimulation of the innate and adaptive immune responses that are becoming increasingly critical in achieving control over distant metastases and disease recurrence. Finally, speculation on the effect of PDT on T Cell Repertoire will be discussed.

FUNDAMENTALS OF PHOTOBIOLOGY AND PHOTOCHEMISTRY

A number of PSs are clinically approved for the cancer therapy and many others are in clinical trials and pre-clinical development. A list of photosensitizers that have been approved for use in humans is presented in Table 1. PSs or formulations of PSs, such as excipients and nanocarriers, can be administered topically or systemically through intravenous administration with the ultimate aim of delivering the PS into the malignant tissue. Certain classes of PS formulations preferentially accumulate in tumors due to their impaired vasculature. This phenomenon is known as the Enhanced Permeability and Retention (EPR) effect (Figure 2A) (9). Other strategies to enhance tumor-specific delivery of PSs include the use of tumor-targeting ligands, such as peptides and monoclonal antibodies that are directly conjugated to PSs (‘photoimmunoconjugates’; PIC). A panel of commonly used PS-delivery vehicles and targeted carriers are depicted in Figure 2B.

Table 1:

A list of clinically-approved PS for cancer therapy

PHOTOSENSITIZER (TRADE NAME)	CANCER INDICATIONS	COUNTRY OF APPROVAL
Porfimer Sodium (Photofrin)	Esophageal and non-small cell lung cancer Lung, esophageal, gastric and cervical cancer Esophageal and endobronchial cancer	USA Japan Canada and most European countries
Temoporfin (Foscan®)	Head and neck squamous cell carcinoma	Most European countries
Padeliporfin (Tookad®)	Prostate cancer	Most European countries
Talaporfin (Laserphyrin®)	Lung cancer	Japan
Methyl aminolevulinate(Metvix)	Basal cell carcinoma and actinic keratosis	Most European countries, Australia, New Zealand

Figure 2.

Photosensitizers and photodynamic therapy. (a) A simplified workflow of in vivo PDT that starts with systemic PS administration and distribution followed by tumor accumulation, photoactivation and local and systemic management of the disease. (b) Representations of classical and cutting-edge nano-sized carrier systems often leveraged to improve PDT efficacy. (c) A simplified Jablonski diagram portraying the energetics of PS molecules following non-thermal photoexcitation. The excitation of a sensitizer in its singlet ground state (PS¹) with red light (most common for in vivo PDT), results in the molecule rising to a higher energy level singlet excited state (PS¹*). The molecule then undergoes non-radiative intersystem crossing to the long-lived triplet excited state (PS³*) whereby type I and type II photochemical reactions proceed. These reactions ultimately lead to therapeutic antitumor biological consequences.

Following systemic administration, PSs reach the tumor microenvironment (TME) via the blood vasculature or the lymphatic system and then allowed to cross the lipid bilayer of the tumor cell membrane, unless a predominantly vascular-PDT effect is required. The distinction between vascular photodamage and tumor cell photodamage is regulated by the PS-light intervals. The cellular uptake of PS or PS formulations can occur either by passive diffusion or by active endocytosis (10, 4). The physical and chemical nature of PS or PS formulations and the composition of the tumor microenvironment (TME) play a critical role in this internalization processes. (11–14). Studies by Huygens et al. showed that hypericin, a water-insoluble PS, can accumulate ununiformly within 3D tumor spheroids and in a manner that is inversely proportional to the expression levels of the cell adhesion molecule E-cadherin (11, 14). A subsequent study by Roelants et al. showed that a water-soluble, serum albumin-associated formulation of hypericin had an even greater uniformity in distribution within tumor spheroids, ultimately resulting in improved PDT outcomes (13). Most importantly, the type of PS can also influence the anti-tumor immune response following PDT treatment. For example, in a study by Garg et al., hypericin-PDT was able to induce damage-associated molecular patterns (DAMPs), such as., ecto-calreticulin (surface-exposed calreticulin) on dying cells. However, although specific doses of 5-ALA PDT had been previously reported to induce ecto-calreticulin, at PDT doses that elicited a similar degree of cytotoxicity as hypericin-mediated treatment, PDT with 5-ALA failed to induce a comparable level of ecto-calreticulin (15). This emphasizes the dose-dependence and sensitizer-dependence of ecto-calreticulin induction when using different PDT regimens and underscores the criticality of fine-tuning PDT treatments for optimal immunological responses. In addition to hypericin and ALA, other photosensitizers, such as photofrin (16), foscan (17), Rose Bengal acetate (18), were also reported to trigger surface exposure of DAMPs following photoactivation.

The outcomes of PDT also rely on the selection of an appropriate PS that has a high extinction coefficient and a high efficiency of generating RMS (Figure 2C). PSs are often preferred to have absorption maxima within the optical therapeutic window (ʎ: 600–900 nm), where light has better penetration into biological tissues. PDT exerts intrinsic dual selectivity due to the spatiotemporal control over the light illumination and tumor-specific localization of the PS. Both the light and PS must be present to initiate phototoxicity, and thereby, minimize photodamage in the nearby non-targeted healthy tissues and immune cells. In a recent study by Kercher, tumor-targeted activable PIC (PS conjugated to anti-EGFR antibody) was used to selectively deliver PS into tumor cells in a 3D co-culture of ovarian cancer cells and T lymphocytes (19). Following high-dose light irradiation, the targeted photoimmunotherapy (TaPIT) showed the selective killing of the cancer cells with a higher proportion of sparing of intratumor T lymphocytes, compared to the conventional PDT with unconjugated PS.

PS-generated RMS includes not only reactive oxygen species, such as singlet oxygen and superoxide radical anions, but also reactive nitrogen species, such as peroxynitrite (20). Tumor cell death occurs as a complex interplay between direct cellular assault through the oxidation of membrane lipids, proteins, and organelles, in addition to the indirect tumor tissue damage through an anti-vascular photodynamic effect. PDT-mediated vascular damage depends on multiple treatment-and TME-specific factors, and, like other anti-vascular treatment modalities, is thought to play a role in inflammatory and anti-tumor immune responses (21, 22). PDT-mediated induction of angiogenesis (also known as neo-angiogenesis), which could potentially promote tumor regrowth, is considered as a major reason for PDT failure in oncological applications (23). Thus, inhibition of angiogenetic pathways in combination with PDT holds promises for wider clinical applications of PDT. In a recent study by Bao et al., tumor vasculature-targeted PDT was combined with immune checkpoint blockade, (discussed in a later section) which resulted in the effective control of primary 4T1 tumors, as well as disseminated 4T1 tumor metastases (24). Interestingly, in another study by Muchowicz et al. using verteporfin-based PDT of orthotopic mammary tumors, the sustained inhibition of lymphangiogenesis in fact significantly impaired the outcome of PDT (25). The study also found a decreased number of tumor-infiltrating immune cells, suggesting a role for intact lymphatic vessels in achieving an anti-tumor immune response.

Thus, PDT-mediated tumor ablation could occur in three main ways: (i) direct photodamage of the cell and cellular organelles, (ii) activation of the immune system, and (iii) photodamage of the tumor vasculature. Depending on the intracellular localization of PS-induced RMS generation and the overall degree of cellular photodamage, a PDT-treated cell can undergo different modes of cellular death, as described in the next section.

CELLULAR MECHANISMS OF PDT-MEDIATED CELL DEATH

PDT induces cell death in a manner that is dependent on the mesoscopic and subcellular localization of PSs, in addition to the effective PDT doses delivered to the tumor cell (26, 27). PDT-mediated generation of RMS leads to cellular death in four predominant pathways: autophagy, apoptosis, necrosis, and paraptosis (28–32, 26, 33). Often, PDT-mediated induction of these pathways is not mutually exclusive. For example, autophagy and apoptosis can be initiated concurrently following photodamage to the lysosomes and mitochondria (lyso-PDT and mito-PDT, respectively) (34). A recent study by Kessel and co-authors proposed that a sequential targeting of subcellular organelles could result in more effective PDT outcomes (26, 35). Understanding these processes is critical to appreciating how different cellular death pathways could differentially stimulate the host immune system. Here, we briefly described the necrotic, autophagic and apoptotic pathways that follow PDT treatment in the context of immune activation.

Necrosis

The term ‘necrosis’ refers to uncontrolled cell death due to the loss of cell membrane integrity as a result of severe cellular insult. Necrosis leads to spillage of the cellular contents into surrounding tissues, which is followed by an inflammatory response (36). For many years, necrosis was presumed to be the only mechanism of cell death following PDT, as a result of the observed tissue necrosis that is ultimately seen in response to treatment. However, in recent years, the understanding of cell death has become more comprehensive, and now includes additional modalities including apoptosis, autophagy, and paraptosis. Although complex, the newly understood intricacies and complexities of biological reactions and molecular cascades that pertain to specific death pathways have also opened up a number of opportunities for appropriate combinations, including those with immune checkpoint inhibitors, to potentiate the response to anti-tumor PDT. In general, the poorly controlled necrotic pathway is considered the most pro-inflammatory pathway, releasing a number of DAMPS that activate the innate and adaptive immune system, as discussed in greater detail in the section discussing tumor-specific antigen release and activation of dendritic cells. The role of the immune system following PDT-induced necrosis has been discussed in significant detail in the literature (27, 29, 33). The role of paraptosis in anti-tumor immune responses are not well understood at this stage and offers significant opportunities for a detailed investigation into that specific death modality.

Autophagy

Autophagy is both an intracellular recycling process that helps in cellular homeostasis and also an auto-destructive cell death pathway. As such, autophagy also has an impact on the immunosurveillance process (30, 37, 38). Autophagy is mediated by the formation of a transient double-membrane structure, known as the autophagosome. Autophagy (ATG)-related genes are involved in the formation of autophagosomes (39). Autophagy is thought to play a crucial role in protecting the cell after sub-cytotoxic PDT, most likely by recycling damaged mitochondria or ER before they can trigger apoptosis (40, 41, 34). A series of studies by Kessel et al. and others showed that autophagy-related genes (ATG5 and ATG7) play an important role in cell death induced by lysosome-targeted PDT (34, 42). When ATG7 was depleted in hepatoma cells, photokilling was significantly impaired following lysosomal photodamage (34). In a study by Garg et al., hypericin-mediated PDT was used to induce autophagy in cancer cells, which resulted in decreased immune-stimulatory effects. As a result of the induction of autophagy, the interaction between dying cells and professional antigen-presenting cells (APCs) was abrogated. This interaction between dying cells and APCs is thought to be required for achieving a robust immune response. Thus, when autophagy was attenuated by depleting ATG5, the cancer cells exhibited enhanced immune-stimulatory effects, including dendritic cell (DC) maturation, IFN production and ecto-calreticulin expression (as described in the section discussing ICD and the release of DAMPs) (38).

The autophagic pathway plays an important role in antigen presentation. Broadly speaking, endogenous antigens from viruses, tumors or self-proteins are processed by cytosolic proteases, then enter the ER lumen, and are presented in the context of major histocompatibility complex (MHC) class I molecules to CD8⁺ T cells (43, 44). In contrast, exogenous antigens are endocytosed and usually processed through the endosomal-lysosomal pathway to present the antigen in the context of MHC class II to CD4⁺ T cells. Multiple studies have implicated the autophagic pathway in having a crucial role in ‘cross-presenting’ exogenous tumor antigens by the DCs in the context of MHC class I molecule to activate CD8+ T cells (45, 46). In an interesting new study by Haug et al., photochemical internalization (PCI) (47) was used as a mechanism to selectively disrupt endosomes in order to increase the cytosolic availability of OVA antigen for MHC class I presentation (48). In this study, a peptide antigen OVA_257–264 (peptide sequence SIINFEKL) was also sequestered in the endosomes following endocytosis by DCs. The endosomes were then photochemically damaged to release the peptide into the cytosol. The selective cytosolic release of the peptide antigen enhanced the presentation of the MHC class I molecule-bound peptide by 30-fold and increased the efficiency of antigen-specific cytotoxic T lymphocytes (CTLs) activation by 30-to 100-fold.

Apoptosis

In contrast to autophagy, apoptosis is a programmed cell death process that requires sequential activation of a number of enzymes, such as caspases and endonucleases. Oleinick et al. first reported that PDT could induce rapid cell death by apoptosis (49). Here, we will briefly discuss the molecular events that lead to apoptosis following PDT (Figure 3).

Figure 3.

Mechanism of PDT-induced apoptosis. Photosensitizers that localize in mitochondria or lysosome can directly damage the cell organelles upon light irradiation. Lysosomal phototoxicity promotes the release of proteases from the lysosome to the cytosol which cleave BID to generate truncated BID (tBID). Next, tBID activates BAX and BAK (BH123 protein) to form mitochondrial outer membrane channels (pores). As a result, cytochrome c and second mitochondrion-derived activator of caspases (SMAC) are released leading to the activation of initiator caspase-8/9 and downstream executioner caspases (caspase 3/7). The release of SMAC leads to inhibition of IAPs (inhibitor of apoptosis proteins) preventing them from binding to and inhibiting activated caspases. Finally, the activated executioner caspases initiate the degradation of cellular components, including proteins and nucleic acids and commence cellular death.

Lyso-PDT disrupts the integrity of the lysosomal membrane and permeabilizes it to stored proteases, such as cathepsins. As a result of lysosomal photodamage, cathepsins enter the cytosol and cleave a proapoptotic protein, Bid, into a truncated active form (tBid) in a caspase-8-independent manner (50). Lyso-PDT is thought to evade the protective shield of the autophagy process (32). Mito-PDT selectively destroys anti-apoptotic proteins of the BCL-2 family, which are trafficked to the mitochondrial outer membrane, while proapoptotic proteins such as Bax and Bak remain intact in the cytosol. With the help of tBid, Bax and Bak are inserted into the mitochondrial outer membrane and oligomerize to form a pore to facilitate the release of cytochrome C into the cytosol. Cytochrome C interacts with apoptotic protease-activating factor 1 (Apaf-1) and procaspase 9 to form the apoptosome. The apoptosome eventually activates downstream executioner caspases (caspase 3/ 7) that carry out the degradation of multiple cellular components, such as proteins and DNA, amongst others.

Sequential targeting of lysosome followed by mitochondria was shown to elicit a strong PDT-mediated direct killing of the cancer cells (35, 51). In a recent study, Visudyne and a lipid-anchored BPD liposomal formulation were used to target mitochondria and lysosomes using a single wavelength laser that resulted into increased cellular death in a 3D model of ovarian cancer (52, 53). However, the specific subcellular target or combination of targets for PDT, which can induce the most potent immune response is yet to be explored. Evidence suggests that low-dose Lyso-PDT results in cytosolic calcium release, activation of calpain, fragmentation of ATG-5, suppression of autophagy and enhanced photokilling when combined with mito-PDT (54). Such an approach to simultaneously suppress autophagy and enhance apoptosis could presumably augment ICD since PDT-induced autophagy is often considered as immunosuppressive (38, 55).

In view of the mechanistic complexity of PDT and the relatively agnostic damage to subcellular compartments by commonly used PSs, a wide variety of avenues exists for exploring the immunomodulatory roles that PDT plays, with respect to the modes of cell death induced at the cellular level, and the pro-inflammatory environments it produces at the macrophysiological level. The immunological impact of photochemical tumor tissue damage will be discussed in the remaining sections of the review.

STIMULATION OF THE INNATE IMMUNE SYSTEM

PDT induces cell death and local inflammation at the site of irradiation, causing a release of multiple cytokines by resident macrophages and stromal cells that further attract inflammatory cells to the site of photodamage (56, 57). These proinflammatory cytokines include IL-1β, IL-6, and TNF-α (58–60, 56). Evans and colleagues were the first to report the release of TNF-α by PDT-treated macrophages which was thought to have a direct cytolytic effect on the tumor cells in addition to its anti-vascular effect (58). IL-1β induction has been implicated in the expression of endothelial adhesion molecules and the recruitment of neutrophils (61). PDT was initially believed to be a local treatment that causes a local cytokine storm. Subsequent studies indicated that PDT could also induce the systemic release of cytokines, complement proteins, and acute-phase proteins (62, 57, 63–66). A study by Cecic and Korbelik reported the release of mediators of peripheral blood neutrophilia following Photofrin-based PDT of mouse EMT6 tumors (57). In addition to the above-mentioned cytokines, this study identified release of IL-10, granulocyte colony-stimulating factor (G-CSF), thromboxane, histamine, prostaglandins, leukotrienes, and coagulation factors as a result of the in vivo PDT. Elevated levels of IL-1β and G-CSF were also observed by de Vree and colleagues following PDT of rat rhabdomyosarcoma tumors (67). The study also suggested that there was also an association between the serum levels of IL-1β and the number of circulating neutrophils. In patients, Nseyo and colleagues examined the presence of urinary cytokines IL-1β and TNF-α in subjects with bladder cancer who underwent PDT (63). Interestingly, these cytokines were only detected in patients treated with the highest light fluences, but not in the control or the lowest light energy treatment groups. These inflammatory mediators play a critical role in the activation and recruitment of immune cells at the irradiated tumor site. However, surgically induced IL-6 is thought to have a negative impact on the efficacy of PDT (68). Here, we discuss how PDT induces immuno-stimulatory cell death that eventually leads to the activation of the adaptive immune system and anti-tumor immunity (Figure 4).

Figure 4.

Innate immune stimulation by PDT. Light-irradiation on photosensitizer-loaded tumor cells leads to necrotic or apoptotic cell death. The dying cells express or secret DAMPs on the membrane or in the TME, respectively. These DAMPs include HMGB1, calreticulin (CRT), Hsp70, extracellular ATP, etc. At the site of light irradiation, various cytokines and chemokines are released by the photodamaged tumor cells and tissue-resident immune cells. These cytokines such as IL-1β, IL-6, TNF-α, etc recruit more innate immune cells such as neutrophils, macrophages, natural killer cells and cause local inflammation. The DCs connects the innate immune system with the adaptive one. DCs engulf the dying cells and undergo maturation. They migrate to the dLN to activate the adaptive immune system as shown in figure 5.

Immunogenic Cell Death (ICD) and release of DAMPs

In addition to eliciting inflammation, PDT also induces immunogenic cell death (ICD) (69–71). ICD refers to the increased immunogenicity of the dying cells that exhibit and secrete danger signals to the immune cells (72). These danger signals are collectively known as Damage Associated Molecular Patterns (DAMPs). DAMPs can be secreted extracellularly or exposed on the plasma membrane of a damaged cell. In a series of reviews by Garg and colleagues, the authors described how PDT-generated DAMPs play a crucial role in the development of anti-tumor immunity (72, 71, 73, 74, 70, 75–78). PDT-generated DAMPs include ecto-calreticulin, heat shock protein 90 (HSP90), HSP70, extracellular ATP, and HMGB1 (70, 69). Calreticulin is one of the most well-studied ICD-markers, which is sequestered in the ER lumen until the cell experiences certain stresses. Upon photodamage of a cell, calreticulin can translocate to the surface of the dying cell and bind to the calreticulin receptor on DCs, (79) enhancing the interaction between a dying cell and a DC. Following this interaction, DCs undergo maturation and activation, which are prerequisites to achieving a strong adaptive immune response (see the section describing tumor-specific antigen release and the activation of dendritic cells).

Infiltration of innate immune cells

PDT induces localized inflammation and the release of cytokines, which promote the infiltration of immune cells including neutrophils, macrophages, NK cells, DCs, and lymphocytes at the irradiated site (80–83) potentially transforming a ‘cold’ non-immunogenic tumor into a ‘hot’ immunogenic one. ‘Cold’ tumors refer to the cancers that, for a number of reasons, haven’t been recognized by the immune system or provoked immune cell infiltration. In contrast, ‘hot’ tumors are characterized by increased infiltration of the immune cells (84). Infiltration of certain immune cells are often corelated with the expression of DAMPs, such as ecto-calreticulin. In a study on non-small cell lung cancer patients by Fucikova et al., intratumoral infiltration of DCs was found to positively correlate with increased ecto-calreticulin expression (85). Ecto-calreticulin serves as an ‘eat me’ signal for the APCs and is thought to be a powerful prognostic biomarker for elevated local antitumor immune response. In a study by Peng et al. on stage III colon cancer patients, ecto-calreticulin was also shown to correlate with T cell infiltration and longer overall survival (86). However, this observation was not in the context of PDT and more studies are needed before concluding that ecto-calreticulin expression and extended patient survival are concurrent.

Neutrophils are the first line of defense that reach the site of inflammation, and their infiltration is enhanced following PDT (87–89). One study by Cecic et al. reported that PDT of solid tumors leads to systemic neutrophilia in a complement-dependent manner (90). PDT-induced neutrophilia was completely abolished when the complement system was inhibited by blocking C3 convertase, an essential enzyme in the complement cascade system. Several studies have demonstrated direct evidence for the active role of neutrophils in anti-tumor immunity. Korbelik and Cecic reported that selective immunodepletion of neutrophils in EMT6 sarcoma-bearing mice resulted in a significant reduction in tumor cures following Photofrin-based PDT(91). A similar observation by de Vree and colleagues also provides insights into the critical roles of neutrophils in the outcomes of in vivo PDT treatment (89). In this study, anti-neutrophil serum was administered before the initiation of Photofrin-based PDT in rhabdomyosarcoma-bearing rats. The results from this study showed that the efficacy of PDT was dependent on the number of circulating neutrophils. In addition to the neutrophils’ migration to the irradiated tumor sites, neutrophils were also reported to accumulate rapidly at the tumor-draining LNs (dNLs) following PDT (92). PDT-induced release of IL-17 by the T helper 17 (Th17) cells has been implicated in this emigration process (88). Activated neutrophils can help DCs to mature by secreting alarmins such as cathelicidins, α-defensin, and HMGB1, amongst others (93). Neutrophils were shown to express TNF-α on the cell surface and migrate to the tumor-draining lymph nodes, where they interact with DCs. These interactions with DCs lead to DC activation, stimulation of CD8⁺ T cells and enhanced anti-tumor immunity (94, 92). A recent study using luminol-generated chemiluminescence as a reporter for neutrophil activation following PDT showed that the early activation of neutrophils within 1 hour following PDT was critical for eliciting a durable PDT response in a mouse model of mesothelioma (95).

Another class of key players within the inflamed tumor following PDT are macrophages, commonly referred to as ‘tumor-associated macrophages’ (TAMs). Korbelik and Hamblin recently described how TAMs play a key role in the enhancement of PDT-mediated immune stimulation (80). In most tumors, the majority of the TAMs adopt an immunosuppressive M2 phenotype and promote tumor angiogenesis and metastasis (96, 97). TAMs are known to entrap drugs, colloids, and other materials that reach the tumor microenvironment due to their highly active phagocytic nature, and thus have been shown to accumulate the highest cellular levels of PSs amongst all the cell types present in tumors (80, 98, 96). Upon light irradiation, these M2 TAMs were eliminated and replaced by a fresh generation of macrophages that have differentiated from tumor-invading monocytes in response to PDT. These new macrophages adopted an immunostimulatory M1 polarization phenotype and secreted pro-inflammatory cytokines, such as IL-1, IL-6, IL-12, TNFα, NO, etc. to elicit exhibit anti-tumor activity. PDT skews the M1/M2 ratio in favor of a pro-inflammatory environment (99, 100, 80) that could therefore potentially support tumor regression and improve overall survival rates, as observed in clinical trials using chemotherapy (101).

NK cells are another type of cytotoxic effector cells that utilize predetermined germline-encoded receptors to sense pathogens and transformed tissue. Therefore, NK cells are considered as a part of the innate immune system. NK cells respond to local inflammation following PDT. Unlike CTLs, NK cells are not tumor-antigen specific, rather they search for missing self-receptor, i.e., MHC class I, on the surface of tumor cells. The cytolytic activity of NK cells is regulated by the relative balance of signals received from cell surface receptors that deliver either activating or inhibitory signals. Under normal physiological conditions, MHC class I molecules on the surface of a healthy cell interact with inhibitory receptors on NK cells, while at the same time, fewer activating ligands interact with activating receptors on the NK cells. However, the inhibitory signal outweighs the activation signal resulting in self-tolerance (102). Tumor cells tend to downregulate MHC class I molecules to escape T cell-mediated killing (103). However, they do become prone to NK-mediated killing. Another mechanism by which PDT may induce NK cell-mediated tumor killing was reported by Belicha-Villanueva et al. (104). In that study, 2-[1-hexyloxyethyl]-2-devinyl pyropheophor-bide-a (HPPH)-PDT was shown to induce the expression of MHC class I-related molecules (MICA) on murine CT26 colon cancer cells and the ligands for NK cell receptors (NKG2DL) on human Colo205 colon cancer cells. The PDT treatment resulted in increased sensitivity of the tumor cells to NK cell-mediated killing. In a similar study by Park et al., hematoporphyrin-PDT was shown to induce NKG2DL expression on human cancer cells which corresponded to increased NK cell-mediated cytotoxicity (105). In the latter study, the use of NKG2D-targeted monoclonal antibodies abrogated the susceptibility of cancer cells to NK cells. All these studies suggest that the induction of NKG2DL on tumor cells following PDT plays an important role in achieving anti-tumor immunity. The activity of NK cells is often modulated by IFNγ, a cytokine that was shown to increase following BPD-based PDT in antigen-bearing tumors (106). Korbelik et al. used genetically altered NK cells that produced IL-2 to enhance the PDT-mediated killing of tumor cells in mouse xenograft models of human colorectal and cervical carcinoma (107). The study demonstrated that NK-based adoptive immunotherapy in combination with PDT could be used to improve the control of the solid tumor. In another study, the PDT-induced anti-tumor activity of CTLs was shown to depend upon the presence of NK cells (81). All these studies indicate that NK cells also contribute to the favorable outcomes of PDT and NK cell-based immunotherapy is a feasible way to control solid tumors.

ACTIVATION OF THE ADAPTIVE IMMUNE SYSTEM

Professional antigen-presenting cells (APCs), such as DCs, are responsible for connecting the innate immune system with the adaptive immune system. Following PDT, DCs respond to the PDT-induced inflammatory stimuli and, following a multi-step process, they educate cells of the adaptive immune system, including CD4⁺ (helper), CD8⁺ (cytotoxic) and regulatory T lymphocytes, to specifically attack the treated tumor(108, 109). In addition to aiding the destruction of the primary tumor, this process also results in the development of memory T cells that are essential for long-lasting anti-tumor immunity, which controls metastatic disease and serves to prevent tumor recurrence after initial PDT-induced regression. Importantly, PDT-induced anti-tumor immunity has also been observed in patients. Most notably in 2007, patients with multi-focal angiosarcoma of the head and neck that were treated with PDT exhibited significant regression of unirradiated lesions within 4 months of treatment (110). In another clinical study, it was shown that PDT was able to induce specific lymphocyte reactivity towards a tumor-specific antigen that is associated with basal cell carcinoma following treatment (111). A separate study also found that patients with vulvar intraepithelial neoplasia, which expressed MHC I (associated with CD8+ T cell activation) responded better to PDT and had a higher tumor infiltration of T cells than patients with MHC I downregulation (112). In a study looking at Bowen’s disease, it was found that immunocompetent and immunosuppressed transplant patients both initially responded equally well to PDT, yet the immunosuppressed patients were ultimately more likely to develop new lesions (113).

Several landmark reviews by Castano et al. (2006), Mroz et al. (2011) and Meading et al. (2016) have described in depth how PDT is capable of eliciting tumor-specific adaptive immunity (108, 109, 114). In this review, we will focus only on the major steps involved in achieving an adaptive anti-tumor response following photodynamic insult to tumors to bridge the fundamentals of photochemistry, ICD and adaptive immunity with the potential for enriching the T cell repertoire (Figure 5).

Figure 5.

Steps for activation of the adaptive immune system by PDT. (1) Photosensitizers accumulate in the malignant tissue. (2) Upon light irradiation (PDT), tumor cells undergo apoptosis and necrosis. DAMPs are expressed on the surface of dying cells. TSAs are released in the TME. (3) DCs internalize the TSAs and become activated and mature. (4) DCs migrate to the dLN and present the processed TSAs to the naïve T cells. (5) Activated T cells undergo clonal expansion and are disseminated into the systemic circulation. (6) Activated T cells reach the primary or metastatic tumor sites. (7) T cell receptor recognizes TSA presented by the tumor cells and form the immunological synapse. (8) Toxic granules containing granzyme B and perforin are secreted at the synapse to specifically kill the tumor cell.

Tumor-specific antigen release and activation of dendritic cells

PDT is thought to enhance the availability of tumor-specific antigens (TSAs) in the TME that are previously not exposed to the immune system. Studies by Gollnick et al. showed that PDT-generated tumor cell lysates were immunogenic and able to elicit a strong CTL-mediated anti-tumor response (115). This study showed that both PDT and UV-generated tumor cell lysates induced phenotypic DC-maturation, however, only PDT-treatment could lead to DC activation with a significant increase in expression of IL-12, a cytokine critical for a cellular immune response. Studies have shown that PDT-induced ecto-calreticulin expression enhanced the interaction between DCs and dying cells (69).

Cellular debris and TSAs released following PDT s are engulfed by immature DCs. TSAs are then processed and presented on the cell surface in the context of MHC class I and II molecules. MHC class I molecules are present on all nucleated cells and present processed antigens to CD8⁺ CTLs. In contrast, MHC class II molecules are seen on all APCs and their expression is further increased upon activation. MHC class II molecules present processed antigens to CD4+ T cells (also known as helper T cells). DCs then mature, express higher levels of MHC molecules and costimulatory molecules, such as CD80 and CD86 on their surface, and secrete proinflammatory cytokines that include IL-12, IL-6, and IL-1β. IL-1β helps in the recruitment of neutrophils at the PDT site and was found to be a critical mediator in PDT outcomes (67, 87, 88). Mature DCs migrate to the regional (draining) lymph nodes (dNLs) where they are thought to present processed TSAs released by PDT to ‘naïve’ T lymphocytes, including CD4⁺, CD8^+, and regulatory T cells.

Cross-priming and activation of naïve T cells

The process of presenting cognate antigens (TSAs) to the naïve T cells is known as ‘priming’. ‘Cross-priming’ refers to the mechanism by which certain APCs take up, process, and present extracellular antigens in the context of MHC class I molecule to naïve CD8+ T cells. Primed T cells exhibit an altered gene expression compared to naïve T cells, and produce elevated levels of cytotoxic agents, namely perforins and granzymes, that are necessary to kill target cells. This process generates CD8⁺ CTLs, which circulate throughout the body to seek out and kill tumor cells expressing antigens specific to those released by PDT (108). CTLs execute this tumor-killing process efficiently by focusing the cytotoxic agents towards the cognate tumor cell avoiding any bystander killing of healthy cells (116). The evidence that CTLs are critical for eliciting an anti-tumor immune response following PDT comes from multiples studies. In a study by Gollnick et al., PDT-generated tumor cell lysates, a heterogeneous soup of TSAs, were reported to stimulate anti-tumor immunity following administration in untreated naive animals (115). In this study, the degree of DC maturation following exposure to PDT-generated tumor cell lysates containing TSAs was compared with exposure to cell lysates generated by a freeze-thaw process, UV radiation or ionizing radiation. DC maturation was characterized by an increase in surface expression of MHC class II molecules, and co-stimulatory molecules CD80 and CD86. Of the lysate-generating modalities tested, including radiation therapy, the released antigens following PDT exhibited the most potent anti-tumor activity in a mouse metastatic rechallenge model. Findings by other groups have also supported the fact that PDT induces an adaptive immune response in a CTL-dependent manner (94, 115, 117–121). In an elegant study by the Gollnick group, local PDT treatments on subcutaneous EMT6 tumors were able to inhibit the growth of distant tumors in the lung (81). However, when CTLs were depleted by administering anti-CD8 antibodies, it led to faster tumor growth, suggesting a crucial role for CTLs in PDT outcomes (81). The CTL-mediated immune response was also shown to depend on the presence of NK cells, but not CD4⁺ helper T cells. Moreover, these immune responses were tumor-specific, as PDT treatment of Colon 26 tumors did not have any anti-tumor effect on EMT6 tumors. This phenomenon implies that PDT treatment of a certain type of tumor activates a specific pool of T cell clones (‘T cell repertoire’, discussed in a later section) that recognizes only specific TSAs-bearing tumor cells. Additional evidence that CTLs are involved in PDT-mediated immune responses comes from a recent study by Garg et al. who used hypericin-PDT to develop a DC-based vaccine that induced a strong immune response and improved overall survival in a mouse model of glioma (15).

An essential aspect of T cell-mediated anti-tumor immunity following PDT is the capacity for CTLs to efficiently infiltrate into the tumor; this has been found to correlate with favorable prognoses in many cancers (122–125). The importance of T cell infiltration in tumors has also been demonstrated in the clinic, with studies on human subjects showing that T cell infiltration into solid tumors correlates with improved clinical outcomes (126, 124, 127, 128). A meta-analysis on ovarian cancer patients revealed that the lack of tumor-infiltrating lymphocytes (TILs) was positively correlated with worse patient survival (128). In another study on colorectal cancer patients, the type, density, and location of TILs were implicated as better predictors of overall survival, as compared to histopathological methods to classify cancer grade (124). Of note, the patient outcomes in the above-mentioned studies were not PDT-specific. However, PDT has been shown to increase the T cell infiltration into the tumor, suggesting that PDT plays a multifaceted role in improving anti-tumor treatment outcomes (82, 81, 83). Gollnick and colleagues showed that a local PDT treatment on EMT6 tumors was able to increase CD8⁺ T cell infiltration into untreated distant EMT6 tumors (81). More recently the Lin group showed that PDT enhances CD3⁺ CD8⁺ T cell infiltration in a primary and distant tumor in mouse models of breast and colon cancers (83, 129). Interestingly, these studies used immune checkpoint blockade (see section discussion immune checkpoint therapy) combined with PDT which showed long-lasting anti-tumor immunity (83, 129).

T CELL REPERTOIRE ENRICHMENT BY PDT – A FORWARD-LOOKING HYPOTHESIS

The immune system has evolved in such a way that T cells are able to recognize a wide variety of antigens through a very specific interaction between the T cell receptor (TCR) and the antigen. This process requires an enormous diversity of TCRs which is estimated to be 1.1 million in human peripheral blood (130). This collective population of T cells that express different TCRs is referred to as the ‘T cell repertoire’ (131, 132). The variety of TSAs that T cells can recognize and elicit an attack on tumor cells is mostly limited by the diversity of the T cell repertoire. Thus, a highly diverse T cell repertoire is desired to combat cancers or infectious diseases, and often used as a predictive marker for favorable outcomes of a cancer immunotherapy treatment (132, 133). Studies have shown that a low or restricted TCR diversity (‘divpenia’) is associated with shorter overall survival in many cancer indications (134, 135).

The TCR diversity is achieved via a process of DNA rearrangement that is known as ‘V-D-J recombination’, which is common in both T cells and B cells for the diversification of their respective receptors. However, unlike the B cell receptor, the TCR specificity is determined before a T cell even encounters any antigens. TCR development does not undergo any antigen-dependent changes, which is seen during the genesis of high-affinity BCRs (known as ‘somatic hypermutation’) (136). Thus, an alternative method by which the responding T cell repertoire of tumor-reactive activated T cells can be enriched is by increasing the availability and diversity of TSAs. Here, we hypothesize that PDT is uniquely capable of presenting a diverse pool of TSAs to the immune system and enriching the responding T cell repertoire with tumor-reactive activated T cells. The diverse PDT-induced pool of TSAs increases the probability of activating different clonotypes of CTLs and thereby kill the tumor cell more efficiently. Here, we will first discuss the molecular mechanism that determines the diversity of antigenic specificity of a TCR, and then we will elaborate further the implications of our hypothesis.

V-D-J recombination and T cell repertoire

T cells originate from the bone marrow and undergo development in the thymus. Bone marrow-resident hematopoietic stem cells give rise to T cell progenitors, which then migrate to the thymus, a primary lymphoid organ, where V-D-J recombination takes place in an antigen-independent manner during the early stages of T cell development. This somatic rearrangement of a variable (V), diversity (D), and joining (J) gene segments occurs at the DNA level to generate the variable region exon that is subsequently linked to a constant region (C) exon by splicing of the primary RNA transcripts in the case of the β-chain of the TCR. Similarly, V-J recombination occurs in the case of the α-chain of TCRs. The α-and β-chains constitute a functional TCRαβ heterodimer of a T cell (Figure 6A). A region of the TCR β-chain, known as Complementarity Determining Region 3 (CDR3) accounts for the highest diversity due to gene recombination, and its amino acid sequence is often used to quantify the diversity of the T cell repertoire. Following V-D-J recombination, T cells undergo rigorous positive and negative selections, resulting in a T cell repertoire of naïve T cells that migrates away from the thymus to circulate throughout the entire body searching for antigens presented by APCs.

Figure 6.

T cell repertoire. (a) T cell receptors (TCR) undergo VDJ recombination in a primary lymphoid organ (thymus) which results in T cell specificity. This somatic rearrangement of a variable (V), joining (J), and diversity (D) gene segments to a constant region (C) in the case of the β-chain of the TCR, and V-J-C recombination (not shown in the diagram) in the case of the α-chain of TCRs, constitute a functional TCRαβ heterodimer of a T cell. A region of the TCRαβ heterodimer, known as Complementarity Determining Region 3 (CDR3) accounts for the highest diversity due to the gene recombination. (b) VDJ recombination takes place before T cells encounter an antigen. Thus a pool of initial repertoire of naïve T cells exists in the body. TSAs released following PDT are carried by the dendritic cells (DCs) to the draining lymph nodes where the DCs activate naïve T cells. TSA-reactive T cells probably undergo clonal expansion and enrich a pool of activated T cells that eliminate the tumors.

During an inflammatory response to foreign antigens, including TSAs or neoantigens, the circulating naïve T cells migrate to the lymph nodes and encounter these antigens for the first time (‘priming’). As described in a previous section, DCs present the processed antigens to the naïve T cells in the context of MHC molecules. Primed T cells become activated and disseminate once again in the entire body through systemic blood circulation to find tumor cells that possess the cognate antigen, for which the T cell is ‘licensed to kill’.

Enrichment of T cell repertoire by PDT-a hypothesis

Enrichment of the T cell repertoire here explicitly refers to the activation and clonal expansion of a greater proportion of pre-existing TSA-reactive T cell clonotypes following PDT, as opposed to diversification (i.e., increase in diversity) of TSA-reactive T cells. As PDT does not generate neo-antigens, its impact on the T cell repertoire is expected to promote the enrichment of a wider variety of TSA-reactive T cells for a more comprehensive immune clearance of tumor cells. Based on the accumulating evidence in the literature, it is conceivable that PDT enriches the T cell repertoire with tumor-reactive activated T cells due to its ability to photochemically damage multiple tumor cell targets, potentially resulting in the exposure of diverse TSAs in the TME. The cytotoxic mechanisms of photochemical damage induced by PDT differ from the mechanisms of cytotoxicity induced by radiotherapy, chemotherapy, biologics or small molecular inhibitors. PDT can simultaneously damage cellular components, organelles and proteins. Thus PDT-induced cytotoxicity is oftentimes agnostic to prior treatments that have failed or have induced resistance.

It is assumed that PDT does not create neoantigens because the targets of conventional PDT do not include nuclei or DNA, whereas, neoantigens exposed following PDT are thought to arise from the intrinsic genomic instability in cancer cells. Liu et al. proposed that PDT may also introduce ‘neoantigens’ by a PDT-crosslinking process (137) but such short-lived altered proteins may only elicit a local inflammation and not an adaptive immune response, which requires genetically encoded neoantigens. These cross-linked proteins are quickly cleared from the TME and are not continuously expressed by the tumor cells following a PDT treatment. As a result, even though some T cells might become activated, they will possibly fail to find cognate antigen-bearing tumor cells within a week after PDT. In contrast, genetically encoded neoantigens derived from mutated genes will continuously be expressed on tumor cells and can be recognized by activated T cells. However, the PDT-mediated adaptive immune response is not restricted to neoantigens only and can also be achieved against TSAs.

PDT conceivably disrupts the TME in a manner that releases an array of TSAs from dying cancer cells. In addition, PDT-generated DMAPs enhance the interaction between DCs and dying tumor cells, which eventually activates the adaptive immune system. Thus, these TSAs released following PDT would otherwise not be accessible to the immune cells when the tumor is treated with conventional treatment modalities. PDT-generated tumor lysates were reported to be more immunogenic than the lysates generated by conventional methods, suggesting that PDT is uniquely advantageous in generating TSAs (115). Determining whether these TSAs are assorted is technically challenging, yet possible due to recent advancements in the field of proteomics and genomics (138). It would be important to identify the most immunogenic TSAs released following PDT treatment, which could potentially help in the development of effective anti-tumor vaccines. This avenue opens up a possibility for PDT-guided personalized immunotherapy in the future for the patients who are ineligible for PDT or other treatment modalities. In such cases, biopsy tissues from the tumor can be treated with PDT ex vivo to identify and expand patient-specific dominant tumor-reactive T cell clonotypes from the patient’s own pool of T cells. These tumor-reactive T cell clonotypes can be infused back for adoptive T cell therapy.

We speculate that a combination of PDT-induced inflammation, a release of DAMPs, immunogenic cell death, and presentation of an array of TSAs is most likely to contribute to the enrichment of the T cell repertoire (Figure 6B). Although this idea of expanding the T cell repertoire has been previously mentioned broadly in a review by Wachowska et al., experimental evidence is yet to come to verify to what extent this hypothesis may hold true (139). Determining the absolute T cell repertoire diversity in blood and tumors has remained a challenge for many decades (131). However, the advent of Next-Generation Sequencing has made it possible to examine the TCR CDR3 diversity in cancer patients in a more rapid, facile and financially viable manner than it was a decade ago (140). Nevertheless, the process does demand a juxtaposition of a multidisciplinary team of scientists with deep expertise in bioinformatics, immunology, cancer biology, and PDT.

Given that PDT results in a plethora of responses that augment the immune-mediated treatment of solid tumors, such as ICD, diverse TSA release and potentially enriching the T cell diversity, it is reasonable to suggest that the combination of immune checkpoint blockade would further potentiate the efficacy of PDT treatment and overall survival. In the next section, we will extend our T cell repertoire hypothesis to discuss its impact on immune checkpoint therapy.

Implications of T Cell Repertoire Diversity in Immune Checkpoint Therapy

Immune checkpoints, in general, refer to a class of surface-exposed inhibitory receptors on both tissue and T cells that halt T cell activation or T cell-mediated killing. Immune checkpoint molecules are components of the normal immune response and occur in response to every immune response. Some checkpoint proteins, such as PD-L1 and PD-L2 are expressed by healthy cells in order to maintain immunological homeostasis and self-tolerance, thereby preventing autoimmunity (141). However, in the context of cancer, immune checkpoints can be leveraged by tumors cells in order to block T cell activation or T cell-mediated killing of cancer cells, and thus sustain their survival by immune evasion. The most studied immune checkpoints include PD1, PD-L1, CTLA-4, IDO, and LAG3, among others (142–144). PD-1 is expressed on the surface of activated T cells and interacts with PD-L1 on the tumor cells. The inhibitory effects of PD-1 are accomplished through dual mechanisms of promoting apoptosis of antigen-specific T cells in the lymph nodes, while simultaneously preventing apoptosis in regulatory T cells, also known as suppressor T cells due to their capacity to suppress the activity of killer T cells (145, 146). Immune checkpoint blockade therapy utilizes small molecules or antibodies to block this inhibitory signaling to potentiate T cell-mediated killing of cancer cells.

Pioneering work by the Lin group showed that PDT synergizes with immune checkpoint blockade in controlling primary and metastatic tumors (129, 83). The group used nanoscale coordination polymer core-shell nanoparticles carrying oxaliplatin in the core and a pyropheophorbide-lipid conjugate (pyrolipid) in the shell for effective chemotherapy and photodynamic therapy (PDT) in a murine model of colon cancer. When PDT was combined with administration of an anti-PD-L1 antibody, potent anti-tumor immunity was elicited, resulting in not only complete tumor eradication at the primary sites but also a systemic anti-tumor immune response that was capable of rejecting and eradicating distant untreated tumors. This process is more widely known as the ‘abscopal effect’. The combination therapy was also found to enhance ecto-calreticulin on the cancer cell surface, which is a hallmark for ICD as discussed earlier, thereby augmenting the innate immune response (129). In the last three years, several other studies were published with the same ethos to combine PDT with immune checkpoint therapy, as were summarized in Table 2.

Table 2:

List of recent developments in PDT and immune checkpoint blockade combination therapy.

CHECKPOINT INHIBITOR	PHOTOSENSITIZER AND CONSTRUCT	CANCER MODEL	REFERENCE
PD-L1	Zn-pyrophosphate nanoparticles loaded with the photosensitizer pyrolipid	4T1 and TUBO bilateral syngeneic mouse models	83
PD-L1	Core-shell nanoparticles pyropheophorbide-lipid conjugate	Primary tumors and distant tumors on syngeneic MC38 and CT26 mouse models	129
PD-1	Integrin αvβ6-specific NIRF probe IRDye700-streptavidin-biotin-HK peptide	Subcutaneous tumors in a 4T1 mouse breast cancer 4T1 (4T1-fLuc) lung metastatic tumors	158
IDO	Chlorin-based nanoscale metal–organic framework	Immunocompetent mouse models using bilateral tumor models of colorectal cancers CT26 and MC38	159
PD-L1	Micelleplexes-PS-lipid conjugate	B16-F10 melanoma xenograft tumor model Lung metastatic tumor model	160
PD-L1	Chlorin e6 (Ce6) manganese dioxide (H-MnO2) nano-platform	4T1 tumor-bearing mice bilateral tumor models	161
PD-L1	Verteporfin	4T1 tumor-bearing mice	25
PD-L1	IR700DX photoimmunoconjugate	H441 tumor-bearing mice	162
PD-1/PD-L1	IRDye700-conjugatedCD276 Fab	Subcutaneous and lung metastatic tumor models	24
CTLA-4	Upconversion nanoparticles with chlorin e6	Subcutaneous CT26 colon cancer	163
CTLA-4	Radachlorin	MC38 and CT26 models in double tumor-bearing mice	164
CTLA-4	Bremachlorin	MC38 and CT26	121
PD-1/PD-L1	Tookad Soluble	Renal cell carcinoma that develops lung metastases	165
PD-L1	EGFR-targeted liposomal formulation of IRDye800CW and Gd-DOTA	Subcutaneous CT26 colon cancer	166

Accumulating evidence suggests that the immune checkpoint therapy works the best against tumors that have propensities to produce neoantigens (147–149) due to DNA-mismatch repair deficiency (150, 151, 149) suggesting that neoantigen-mediated enrichment of T cell repertoire (152) is a prerequisite and favorable prognostic factor to achieve a successful outcome to immunotherapy. Immunotherapies with therapeutic antibodies against CTLA-4 or PD-1 were used successfully to increase and stabilize the peripheral T cell repertoire diversity which resulted in improved overall survival in patients with melanoma, lung, and prostate cancers (149, 153). While DNA-mismatch repair deficiency is often associated with chemoresistance (154), the loss of mismatch repairability was reported not to contribute to resistance to PDT (155), suggesting a basis for the rational combination of PDT and immune checkpoint therapy for such types of cancer indications. PDT was never documented to create any neoantigen, but it probably increases the accessibility of TSAs or neoantigens to the immune system, thereby enriching the tumor-reactive T cell repertoire. Administration of anti-PD1 or anti-CTLA4 antibodies presumably helps to stabilize the PDT-induced, highly diverse, and clonally expanded, activated T cells, resulting in a more robust anti-tumor immunity as reported in several PDT studies listed in Table 2. The concept of T cell repertoire enrichment has been reported for radiotherapy (156, 157), and as such we propose that PDT can also play a similar role in expanding the number of activated TSA-specific clonotypes for a more substantive and comprehensive attack on PDT-treated tumors. Examining the T cell repertoire diversity status before and after PDT treatment could, therefore, be a valuable parameter to evaluate outcomes of PDT combination therapies.

CONCLUSIONS: CHALLENGES AND FUTURE DIRECTIONS

Numerous studies have established the capacity for PDT to eradicate primary tumors and elicit a durable immune memory to combat metastatic disease. PDT induces ICD and releases DAMPs, which activate the innate and adaptive immune systems. PDT also recruits inflammatory cells at the site of irradiation and increases the infiltration of killer T cells into solid tumors. A unique feature of the treatment modality is its ability to convert less immunogenic ‘cold’ tumors into ‘hot’ immunogenic ones. Though PDT is traditionally considered to be a focal treatment, it is increasingly being regarded as a powerful modality with multifaceted global systemic anti-tumor effects. Combinations of PDT with immune checkpoint blockade holds significant promise in advancing this anti-tumor treatment modality to the clinic for many problematic solid cancer indications. In this review, we proposed a hypothesis that PDT enriches T cell repertoire and speculated how it could impact the outcomes of immune checkpoint combination therapies. However, the direct experimental evidence is yet to come to support this hypothesis. Many in vitro studies have revealed that the type of photosensitizer, light doses, tumor heterogeneity, and tumor model used play a critical role in eliciting effective anti-tumor immunity. Thus, optimization of all these parameters and patient-specific personalization is crucial in order to achieve favorable clinical outcomes. Examining the T cell repertoire status of patients may guide, to some extent, the design of such PDT-based precision medicine in efforts to revolutionize photodynamic immunotherapy.

ABBREVIATIONS

APC

Antigen-presenting cells

CTL

Cytotoxic T lymphocyte

DAMP

Damage associated molecular pattern

DC

Dendritic cell

dLN

Draining lymph node

EPR

Enhanced permiability and retention

ER

Endoplasmid reticulum

GNGB

Granzyme B

HSP70

Heat shock protein 70

ICD

Immunogenic cell death

IL-2

Interleukin 2

PDT

Photodynamic therapy

TNF

Tumor necrotic factor

TSA

Tumor specific antigen