1. Difference between Carcinogen-Associated and Virus-Associated HNSCC
Using multiplex immunohistochemistry, recent studies explored the head and neck squamous cell carcinoma (HNSCC) tumor immune microenvironment (TIME) and found that a myeloid-inflamed profile was associated with a poor prognosis and that high numbers of CD8
+ T cells at the invasive margin of human papillomavirus (HPV)-negative HNSCC were associated with prolonged overall survival, respectively
[1][2].
Saloura et al. observed that HPV-positive tumors are enriched with CD8
+ T cells and Tregs, and HPV-negative tumors show a lower abundance of CD8
+ T cells but a high infiltration of M2 macrophages compared to HPV-positive tumors
[3].
The difference in survival between HPV-positive and HPV-negative HNSCC patients could be due to an adaptive immune response directed against the viral antigens expressed by tumor cells, which determines a higher presence of tumor-infiltrating lymphocytes (TILs) and an inflamed gene expression profile
[4]. A minority of HPV-positive oropharyngeal squamous cell carcinoma (OPSCC) patients respond poorly to treatment and have a dismal prognosis
[5]. Smoking has been shown to reduce the survival benefit of individuals with HPV-positive oropharyngeal squamous cell carcinoma (OPSCC)
[6], and therefore, it might be that these patients are also smokers.
The disease-specific survival of OPSCC patients, stratified according to HPV status and tumor-infiltrating lymphocyte (TIL) levels, is lower in HPV-positive/low-TIL patients, and it is similar to that of HPV-negative patients
[5]. This is consistent with the observation that not all HPV-positive OPSCCs are the same.
Epstein–Barr virus (EBV)-positive nasopharyngeal carcinoma (NPC) displays an inflamed phenotype, according to Chen and Mellman
[7]: immune cells are in close proximity to and in contact with NPC cells, instead of being embedded in the surrounding area away from the tumor core. Since the nasopharynx is one of the first defensive organs against viral and bacterial entry and infection, its microenvironment is physiologically highly reactive and immunogenic. In the normal nasopharyngeal stroma, two different major cell lineages are present: CD45+ immune cells, including T cells, B cells, NK cells, and MDSCs, as well as CD45- non-immune stromal cells, including fibroblasts and endothelial cells. The non-cancer-associated inflamed nasopharyngeal microenvironment differs from the TIME of NPC: the former shows an abundance of B cells, whereas T cells, NK cells, myeloid-derived cells, and fibroblasts are more likely to infiltrate the NPC TIME
[8][9].
Finally, even if arising in similar anatomical sites, carcinogen-associated HNSCC and virus-associated HNSCC are characterized by distinct immune landscapes that strongly influence patients’ responses to immunotherapy and their outcomes, as described in
Table 1. It is noteworthy that carcinogen-associated HPV-negative HNSCC displays a higher mutational load but low immune infiltration
[10] compared to HPV-positive tumors; these factors influence the different clinical behaviors as well as the sensitivity to treatment and the prognosis. Exploring the distinct TIME features can help HNSCC investigators to rationally identify new immune targets and consequently plan new strategies for TIME-oriented clinical trials.
Table 1. Factors influencing TIME: difference between carcinogen-associated and virus-associated HNSCC.
2. How Smoking Affects HNSCC TIME and Its Influence on Escape Mechanisms to Immunotherapy
Carcinogen-associated HNSCC is usually diagnosed in men in their 50s or 60s, and it is strongly associated with smoking and alcohol; it is slowly declining globally, in part because of the decreased use of tobacco
[11]. Even if treated with the best multimodal treatment (surgery, radiotherapy, and chemotherapy), locally advanced carcinogen-associated HNSCC still presents a dismal prognosis with a 40–50% 5-year OS
[12].
Yet, in 2013, Hernandez C.P. et al.
[13] demonstrated that cigarette smoke extract (CSE) is able to induce the inhibition of T-cell proliferation and activate T-cell apoptosis in vitro in a dose-dependent manner. Apoptosis enhanced by CSE was independent of caspase activation and endogenously mediated through reactive oxygen species (ROS) and reactive nitrogen species (RNS). This explanation is compelling and well supported by others: in fact, early studies have already shown that a chronically inflamed microenvironment (inflammatory disease or cancer-related) inhibits cytotoxic T cells and strengthens their hypofunction
[14], for example, via NF-κB phosphorylation inhibition
[15], a dimeric transcription factor involved in the expression of proteins necessary for innate immunity, apoptosis, and cell proliferation
[16]. Moreover, the exposure of T cells to CSE induces the phosphorylation of eukaryotic translation initiation factor 2 alpha (eIF2), a factor involved in the expression of proteins promoting cellular apoptosis
[13].
Besides in vitro studies, recently, de la Iglesia et al.
[17] found that in an HPV-negative HNSCC population, active smoking led to an immunosuppressive signature, presenting as a decrease in cytotoxic T-cell tumor infiltration and the reduced expression of genes in the IFNα and IFNγ response pathways compared with former and never smokers. The smoking mutational signature, as found by TCGA, is correlated with tumor mutational burden (TMB)
[10][18]; surprisingly, the authors analyzed the smoking status (using self-reporting) and TMB in a study subpopulation and did not find any correlations between these two parameters. Although the mutagenic effects of tobacco exposure are similar in HNSCC and squamous lung cancer (SLC), Desrichard A. et al. demonstrated an inverse correlation between the mutational smoking signature and the IFNγ signature in HNSCC patients and a positive correlation in SLC patients. Indeed, in HNSCC patients, the mutational smoking signature is associated with poorer survival, fewer tumor-infiltrating lymphocytes (TILs), and strong immunosuppressive effects. Conversely, in the SLC population, smoking is associated with a more inflamed tumor microenvironment, a higher TIL level, and a better response to immunotherapy
[19]. In particular, HNSCC patients with a high mutational smoking signature show both low CD8
+ T-cell infiltration and low IFNγ expression, suggesting that CD8
+ T cells are not only less represented but also less capable of producing IFNγ
[10]. In other words, smokers seem to have fewer infiltrating and also less functional CD8+ T cells.
Chemokine profile analyses performed in current smokers showed the decreased expression of the CXCL9,10,11/CXCR3 axis compared to current non-smokers. These chemokines are known to regulate immune cell migration, differentiation, and activation through the recruitment of cytotoxic lymphocytes and natural killer (NK) cells in response to IFNγ expression
[20]. Recent data suggest that tumors associated with the IFNγ signature and inflamed phenotype have the highest probability of response and survival benefits when treated with anti-PD-1 (programmed cell death protein 1 (PD-1)) checkpoint inhibitors
[21][22]. An in vivo preclinical study showed the inhibition of anti-PD-1 inhibitor effects in CXCR3 knockout mice, indicating that the homing of T cells to the tumor through the CXCL9,10,11/CXCR3 axis may be critical for anti-PD-1 inhibitor efficacy
[23].
Current smokers had significantly lower numbers of PD-L1 (programmed cell death protein 1 and its ligand (PD-L1))-positive cells in the tumor core and tumor margins compared with never and former smokers
[17], which is consistent with data that patients with smoking-high HNSCC have a lower response rate when given anti-PD-1 checkpoint inhibitors compared with smoking-low HNSCC
[19]. Carcinogen-associated HNSCCs are characterized by enriched M2-phenotype macrophages, contributing to the creation of the immune-excluded TIME
[24]. Monocytes recruited by specific cytokines released by the tumor (mainly CCL2
[25]) differentiate into M2-phenotype macrophages (M2s) in the hypoxic environment under VEGF pressure, losing their ability to migrate. Once they become residents in the TIME, M2 cells start to produce VEGF, which enhances the “feed-forward” loop, attracting new macrophages to the TIME, and TGF-β, one of the most potent immunosuppressive cytokines, which transforms normal fibroblasts and probably other stromal cells into cancer-associated fibroblasts (CAFs)
[26]. CAFs, in turn, are known to release immunosuppressive cytokines (such as high levels of TGF-β, IL-10, and IL-6) in the TIME and to produce a stiff extracellular matrix that forms a sort of impenetrable barrier for immune cells and impairs oxygen and drug distribution
[27]. In brief, the HNSCC TIME, ruled by high levels of TGF-β, leads to a self-renewing hypoxic environment. Brooks JM et al.
[28] validated a combined hypoxia and immune prognostic classifier in HNSCC, finding three different categories: high hypoxia associated with low immune infiltration, low hypoxia associated with high immune infiltration and a mixed category. The first category is composed almost completely of carcinogen-associated HNSCCs with the worst overall survival in comparison to the other two.
p53 is the most frequently mutated gene in carcinogen-associated HNSCC
[29][30][31]. The loss of p53 function promotes the recruitment and instruction of suppressive myeloid CD11b
+ cells, in part through the increased expression of CXCR3/CCR2-associated chemokines and macrophage-colony-stimulating factor (M-CSF), and attenuates CD4
+ T-helper 1 (Th1) and CD8
+ T-cell responses in vivo; additionally, p53-null tumors also show an accumulation of suppressive regulatory T (Treg) cells
[32][33].
3. How HPV Affects HNSCC TIME and Influences Escape Mechanisms
HPV-associated OPSCC is increasing, mainly in the US and Western Europe, with a 10–30-year latency after oral sex exposure
[12][34]. The outcome of non-metastatic HPV-positive oropharyngeal cancer is more favorable than the HPV-negative form, because it tends to have a better response to radiotherapy and chemotherapy, and patients are generally younger with better performance status
[11].
Persistent infection with high-risk HPV (especially type 16) has been demonstrated to be the cause of virus-associated OPSCC. The virus exclusively infects basal keratinocytes and replicates only in fully mature epithelial cells, which are intrinsically programmed for death, and therefore, their death does not alert the immune system
[35]. Hence, viral antigens are detectable only in superficial epithelial cells destined for desquamation and remote from immunological surveillance
[36], enabling the virus to be undetected for long periods
[37]. HPV-associated oncogenesis is controlled by the E6 and E7 oncoproteins
[38]. The former promotes p53 degradation, upregulates telomerase activity, and maintains telomere integrity during repeated cell divisions, while E7 binds to retinoblastoma protein (pRb), allowing uncontrolled cell division. E7 can bind and degrade proteins that control cell-cycle entry in the basal and upper epithelial layers and thus is able to stimulate host genome instability through the deregulation of the centrosome cycle
[39].
In previous clinical studies, the E6 and E7 long peptides showed their immunogenicity in being able to induce HPV-specific CD4
+ and CD8
+ T-cell proliferation and activity. From these findings, vaccines for the immunotherapy of HPV16-induced progressive infections, lesions, and malignancies have been developed
[40].
Viral E6 and E7 oncoproteins are also known to be the main drivers of the immune escape mechanism in HPV-associated HNSCC
[41], deregulating multiple immunity-related pathways to avoid recognition and clearance by the host immune system. E6 and E7 are able to downregulate activating chemokines such as CCL20, CCL2, and IL-8, leading to a reduction in dendritic cell (DC), monocyte, and neutrophil recruitment
[42]; moreover, E6 and E7 enhance the release of inhibitory cytokines, such as: (a) IL-10, one of the most important immunosuppressive cytokines, which is able, among other functions, to initiate CD8
+ cell exhaustion via activation of the transcriptional factor TOX, which contributes to the upregulation of PD1, TIGIT, TIM3, and LAG3
[43]; (b) TGF-β, which induces CD8
+ T-cell and NK cell inhibition and the switch of CD4
+ cells to CD4
+ FOXP3
+ (Tregs), eventually resulting in tumor tolerance and immune evasion
[44]; (c) CXCL12, which contributes to Treg and Th2 cell recruitment
[42].
The secretion of extracellular vesicles (EVs) is another important mechanism of immune escape
[45] in HPV-positive cancers, based on cell–cell and cell–environment interactions between cancer and immune cells. It has been reported that HPV-positive cells release EVs that modify the microenvironment, enhancing tumor development and chemoresistance
[41][46]. The role of EVs in the immune response was first described in 1996
[47]. In the last three decades, several studies have been conducted to explore EVs’ mechanism and influence on immune escape
[48]. EVs harbor immunosuppressive molecules such as Fas-Ligand or tumor-necrosis-factor-related apoptosis-inducing ligand (TRAIL), checkpoint receptor ligands (PD-L1), or inhibitory cytokines (IL-10, TGF-β, and prostaglandin E2)
[49].
E6 and E7 oncoproteins are also involved, in a dose-dependent manner, in interfering with the transcriptional activity of NF-KB
[50], with a crucial negative switch on the inflammation triad, composed of IL-1
[51], TNF-α
[52], and IL-6
[53].
Toll-like receptor 9 (TLR9) and stimulator of interferon genes (STING) are specific sensor proteins that are able to recognize DNA from viruses or bacteria within the cell cytosol or endosomal compartments and activate a type I IFN response
[54]. Recently, Wang S et al.
[55] showed that TLR9 was more often underexpressed in HPV-positive HNSCC tumors compared to their HPV-negative counterparts, and it is associated with a relatively poor prognosis. The HPV E7 oncoprotein can antagonize the STING pathway via NLRX1, which is a critical intermediary partner for STING turnover. In a preclinical model, the depletion of NLRX1 resulted in significantly improved type I IFN–dependent T-cell infiltration profiles and tumor control
[56].
The TIME of HPV-positive HNSCC is considered “inflamed” by definition
[24], enriched with CD8
+ cells, CD4
+ cells, Tregs, B cells, NKs, and M1-phenotype macrophages, with high expression of PD-L1
[32].
A recent study established a spectrum of differences between immune lineages in carcinogen-related versus HPV-positive HNSCC: besides CD8+ cells and Treg cells, similar lineages in both types of HNSCC, CD4
+ T cells, B cells, and myeloid cells display different immune lineages, so it may require more tailored therapies
[57]. These differences between HPV-positive and HPV-negative HNSCC TIMEs might be due to the presence of viral antigens (episomal or integrated components)
[38], which may prime HPV-positive patients for enhanced antitumor immunity
[57].
Using an RNA-seq analysis of 84 HPV-positive HNSCC tumors, Koneva et al.
[38] explored the presence of HPV integration sites in cancer transcriptomes. They showed that integration-negative tumors (defined by the absence of the expression of viral–host fusion RNA transcripts) have better OS and higher levels of immune-related genes than those with integration-positive tumors
[38]. Moreover, they found that the OS of integration-positive patients was similar to that of HPV-negative patients. Integration-negative tumors were characterized by strongly heightened signatures for immune cells, including CD4
+, CD3
+, regulatory, CD8
+ T cells, NK cells, and B cells, compared with integration-positive tumors
[38].
4. How EBV Affects Nasopharyngeal Cancer (NPC) TIME and Its Influence on Escape Mechanisms
Non-keratinized undifferentiated NPC is closely related to EBV infection
[58]. Although this tumor originates from squamous cells of the nasopharyngeal mucosa, due to its outcome, it may be considered separately. NPC has a low incidence rate worldwide (just 0.7% of all cancers globally in 2018
[59]), but it is endemic in Southeast Asia with a high mortality rate
[60]. Advanced disease contributes to high mortality rates in these endemic regions
[60]. EBV infection in the epithelium of the nasopharynx can progress from lytic to latent infection, which is strongly associated with the carcinogenesis of NPC
[61]. EBV is able to maintain the expression of various viral proteins, such as EBV nuclear antigen 1 (EBNA1), latent membrane protein 1 (LMP1), LMP2A, and LMP2B, during latent infection inside NPC cells
[61]; all of these proteins are important in balancing viral replication and protein expression in order to prevent the presentation of viral antigens to the immune system
[62], resulting in an oncogenic but weakly immunogenic nature.
The previously mentioned ethnic differences in NPC incidence suggest a major influence of genetic susceptibility, which is strongly linked to the immune escape mechanism underlying this disease. Epidemiological studies and recent large-scale genome-wide association studies have strongly demonstrated the association between HLA class I genes and NPC risk. Since HLA class I genes encode proteins that identify and present foreign antigens for the initiation of the host immune response against infected or malignant cells, it is hypothesized that high-risk populations with specific HLA haplotypes may be less efficient in mounting immune responses against latent EBV infection in the nasopharyngeal epithelium
[63].
The knowledge of the role of EBV latent genes in immune evasion by NPC is yet to be completely achieved. Nevertheless, growing evidence shows several mechanisms that protect NPC cells from the host immune system. EBV-positive NPC cells are able to secrete cytokines and exosomes that drive the TIME toward immune suppression
[64]: data from whole-exome sequencing and single-cell sequencing studies have progressively shown tumor infiltration by dysfunctional and exhausted CD8+ T cells and effector T (Teff) cells that overexpress inhibitory immune checkpoints, such as PD-L1, LAG3, galectin 9–TIM3, TIGIT, and CTLA4; moreover, other immunosuppressive cells, such as Tregs, TAMs-M2, and MDSCs, and various inhibitory cytokines have been identified to contribute to immunosuppression
[64].
T-cell exhaustion represents one of the most important ways to block antitumor immune responses; unfortunately, the underlying mechanisms of this process are still largely unknown
[65]. Recently, two different single-cell sequence analyses of CD8
+ T cells from the TIME and the peripheral blood of EBV-positive NPC identified high numbers of exhausted CD8+ T cells, together with a significantly more restricted T-cell receptor (TCR) repertoire in both compartments, which explains the reduction in cytotoxic activity
[9][66]. EBV-positive tumors are able to induce a highly variable pattern of TIME with increased numbers of different immune cell subsets, in particular, high frequencies of effector T cells, Tregs, and TAM-M2 cells. Of special interest, NPC cells can induce Tregs, suppress effector T cells, and regulate HLA class I expression, producing the so-called EBV-associated BamHI-C fragment rightward reading frame 1 (BCRF1) protein, which can encode viral IL-10 (vIL-10). vIL-10 has a very high sequence similarity to its human counterpart, IL-10
[67], exerting immunosuppressive effects on T cells but lacking the immunostimulatory effect of IL-10, which may contribute to the progression of tumors
[68]. Furthermore, EBNA1 expression, for example, can upregulate CCL20, which recruits Treg cells that inhibit cytotoxic T-cell activities
[69]. LMP1 can induce PD-L1- and galectin-9-containing exosomes, which enhance T-cell apoptosis and inhibit the functions of immune cells
[70]. LMP2A and LMP2B are able to downregulate the antiviral response to interferon, inducing an increase in the turnover of interferon receptors
[71]. Moreover, the abundant presence of a particular group of EBV-encoded microRNAs, named miR-BARTs, encoded by specific intronic regions of NPC, can activate the evasion of cell-surface major histocompatibility complex class I–related chain B for immune cell recognition, reducing the transcriptional activation of IFNγ and inhibiting NLRP3 inflammasome activation
[72][73][74].
In NPC, the persistent activation of NF-kB pathways by somatic gene alterations or viral oncoproteins has been shown to play a crucial role in NPC tumorigenesis
[75]. NF-kB is a family of five transcription factors: NF-kB1 (p105/p50, encoded by Nfkb1), NF-kB2 (p100/p52, Nfkb2), RelA (Rela), RelB (Relb), and c-Rel (Rel); in the resting state, NF-kB subunits are retained in the cytosol by IkB proteins and by unprocessed p105 and p100, which function as inhibitors. The activation of diverse receptors leads to the nuclear translocation of homo- or heterodimers of NF-kB subunits, which can then activate or repress gene transcription
[76]. NF-kB signaling is key for immune function, and it is likely necessary for antitumor immunity
[75]. Li YY et al.
[77] showed that the majority of NPCs display the activation of the NF-kB signaling pathway as a result of somatic inactivating mutations in negative regulators of NF-kB. Previous studies have suggested a role for the non-canonical NF-kB pathway in Treg development and maintenance. Additionally, chronic inflammation recruits myeloid-derived suppressor cells (MDSCs), which promote NF-KB-controlled Treg cells to stimulate tumor angiogenesis and immune evasion
[78]. In EBV-positive NPC cells, activated NF-kB regulates a number of chemokines (CXCL9, CXCL10, CX3CL1, and CCL20), which recruit tumor-infiltrating T lymphocytes and modulate the NPC tumor environment
[63].
This entry is adapted from the peer-reviewed paper 10.3390/biomedicines10102498