Immunomodulation of salivary gland function due to cancer therapy

Abstract

Functional salivary glands (SG) are essential for maintaining oral health, and salivary dysfunction is a persistent major clinical challenge. Several cancer therapies also have off‐target effects leading to SG dysfunction. Recent advances highlight the role of SG immune populations in homeostasis, dysfunction and gland regeneration. Here, we review what is known about SG immune populations during development and postnatal homeostasis. We summarize recent findings of immune cell involvement in SG dysfunction following cancer treatments such as irradiation (IR) for head and neck cancers, immune transplant leading to graft‐versus‐host‐disease (GVHD) and immune checkpoint inhibitor (ICI) treatment. The role of immune cells in SG in both homeostasis and disease, is an emerging field of research that may provide important clues to organ dysfunction and lead to novel therapeutic targets.

1. INTRODUCTION

Receiving a cancer diagnosis is a devastating event for patients and their loved ones worldwide. Recent advances in cancer treatments, such as early detection, improved IR techniques, combinational treatment and immunotherapy has made many cancers manageable, and the numbers of cancer survivors are increasing (Siegel et al., 2022). However, with survivorship comes both acute and chronic challenges due to adverse effects of cancer therapy, and decreased oral health is among these iatrogenic diseases (Epstein et al., 2012).

Oral health is dependent on saliva and the function of SG. Saliva is a complex fluid of water, mucus and enzymes that lubricates, buffers, aids in digestion and has anti‐viral, ‐microbial and ‐fungal properties. Consequently, loss of saliva leads to oral dryness, irritation of the buccal mucosa and soft palate, loss of papillae on the tongue, enamel deterioration, increased infections, difficulty eating and swallowing, and even malnutrition (Epstein et al., 2012). Salivary hypofunction due to adverse effects of cancer treatment is often irreversible, and although some topical treatments are available, it remains a major clinical challenge.

Cancer therapies known to affect SG function include IR and combined chemotherapy treatment of head and neck cancers, hematopoietic stem cell transplant for blood cancers, as well as ICI immunotherapy. Although these treatments are different, they all lead to atrophy of the salivary epithelia with fibrosis and the involvement of immune cells is a common feature of the disease progression. Still, the role of the immune system in responding to cancer therapy, its involvement in the loss of secretory function and its role in the repair or regeneration of the gland are not well understood. In general, the role of the immune system in SG development, homeostasis and regeneration is not well established. Here, we review current knowledge about immune populations during embryonic and postnatal SG development, their presence during homeostasis and their response to cancer therapy. Understanding how the immune system modulates development and homeostasis is vital to clarify immunomodulation of the SG in response to cancer treatment and how this could subsequently be harnessed to resolve insult and regenerate salivary function.

1.1. Immune cells in SG during embryonic development

Most of our understanding of SG development has come from analysis of murine models. Embryonic SG development starts at embryonic day 11 (E11) by invagination of oral epithelia to form a single bud which undergoes extensive branching morphogenesis and cytodifferentiation (Chibly et al., 2022). Studies of signaling pathways and cell interactions during SG development has led to identification of key factors required for epithelial growth and differentiation and understanding of the essential roles for neuronal, mesenchymal and endothelial cells (Chibly et al., 2022). The role of immune cells during developmental processes is not well understood. However, recent work has shed light on some immune populations and their ontogeny, in particular macrophages, which are a predominant immune cell in SG and enhance tissue repair (McKendrick et al., 2023; Sathi et al., 2017).

Macrophages are innate immune cells that play a central role in organ development, homeostasis and repair in addition to protective immunity. Embryonically, they develop from three waves of spatiotemporally distinct progenitors. The first stemming from the yolk‐sac around E7.5, a second transient wave of erythro‐myeloid origin at E8.5, and a third wave from hematopoietic cells at E10.5 (Ahlback & Gentek, 2024). These developmentally distinct lineages either replace each other or coexist in the same tissue niches throughout life (Ahlback & Gentek, 2024). The ontogeny of macrophages varies between tissues and adds a layer of heterogeneity and this interplay together with microenvironmental imprinting could be important for outcomes of disease states (Ahlback & Gentek, 2024). In SG, macrophages are detected as early as E12 and found throughout embryonic development (Hauser et al., 2020; Lu et al., 2022; McKendrick et al., 2023; Sathi et al., 2017) (Figure 1). Flow cytometry analysis of macrophages at pre‐ and post‐natal developmental stages suggests an age‐dependent switch in macrophage subsets based on CD11c expression (Lu et al., 2022). Recent fate‐mapping experiments using a Cdh5^CreERT2/+.Rosa26^{GAG‐LSL‐tdT/+}.Cx3cr1^GFP/+ mouse model confirmed this switch and showed that embryonic glands are initially seeded from the yolk sac progenitors, and while a small portion are still present in the adult glands, they are almost completely replaced by hematopoietic stem cell‐derived macrophages postnatally (McKendrick et al., 2023).

FIGURE 1.

Macrophages in mouse SG during embryonic development. Representative images showing wholemount immunohistochemistry staining of epithelia (E‐cadherin, ECAD, magenta) and macrophages (F4/80, green) from E13 to E18. Inserts showing close‐up of endbud/acinar area. Scale bars: 500 μm.

Many factors have been identified in various tissues that regulate macrophage differentiation and maintenance, and colony stimulating factor 1 (CSF‐1) is known to be a primary regulator (Keshvari et al., 2021; Sathi et al., 2017). Accordingly, CSF‐1 Receptor‐deficient mice have severe depletion of macrophage populations throughout the body, including the SG (Cecchini et al., 1994; Lu et al., 2022; McKendrick et al., 2023). Interestingly, Sathi and colleagues showed that CSF‐1 orchestrates embryonic SG branching morphogenesis, although they concluded it was acting directly on the epithelium rather than via macrophages (Sathi et al., 2017). Thus, the specific role of CSF‐1 on macrophages during embryonic SG development is not clear.

Another immune cell population that is present during embryonic development in both mice and humans is mast cells, and they share much of their ontogeny with macrophages (Chia et al., 2023). Mast cells have non‐immunological roles during development, such as mediating vascular and nerve branching via VEGF and neurturin in the developing cornea (Chia et al., 2023). In SG, a small population of mast cells marked by their core signature genes (Tpsb2, Cma1, Mrgprb1, Mrgprb2 and Cpa3) were identified by scRNAseq and microarray of embryonic development (Hauser et al., 2020) (http://sgmap.nidcr.nih.gov/sgmap/sgexp.html), although their specific functional role in SG development has not been studied.

Recently, sci‐RNA‐seq of human fetal SG identified an immune population at 12–19 weeks of gestation (Ehnes et al., 2022). However, the specific immune cell type or subpopulations were not investigated and potential similarities to embryonic mouse SG immune cell development is unclear.

1.2. Immune cells in SG during postnatal development and homeostasis

Postnatal development of the SG is also a critical phase of functional secretory specialization and maturation of the ductal, myoepithelial and acinar cell compartments. Similarly, the composition of SG immune populations also changes after birth due to the exposure of the pup to the external environment and then again with the increasing external microbiota challenges around weaning (Zubeidat et al., 2023). Here, we are defining tissue homeostasis as the state where regulated variables are kept within the homeostatic range. In general, organ steady‐state is maintained by supportive cells within tissues and in SGs, this include endothelia, pericytes, fibroblasts, nerves and immune populations. They all aid in epithelial maintenance and tissue homeostasis and are often referred to as the epithelial stem cell microenvironment.

Although the influx of immune cells to SG has previously simply been associated with pathological conditions, it is now appreciated that both innate and adaptive immune cells are present in SG throughout development and during homeostasis. Immune cells in steady‐state tissues are in general thought to play an integral role at all stages of the immune response and maintaining regulated variables within the organ by responding to infectious insults, resolution of inflammation and tissue repair. Though resident immune cells in SGs are not well understood, recent advances have shed light on some of these populations and their potential interactions (Table 1). In this section, we summarize known and suggested roles of both innate and adaptive immune populations in steady‐state SG.

TABLE 1.

Summary of recent descriptions of immune populations in homeostatic postnatal SGs. Cell images from www.biorender.com.

Cell type	Functional roles	Human	Mouse
Innate immunity
Macrophage	Majority immune cell in homeostasis Subpopulations with specific anatomical niches Lining epithelia and crosses the BM barrier for surveillance Interact with ILC1 and T‐cells Cross‐talk with epithelium for maintenance?	Horeth et al. (2023) Huang et al. (2021) Zong et al. (2023)	Hauser et al. (2020) Lu et al. (2022) McKendrick et al. (2023) Rheinheimer et al. (2023) Zhao et al. (2023) Zhao et al. (2020)
Dendritic cell	Migrate from the oral mucosa upon microbial challenge around weaning Two subpopulations are described in mouse Abundantly found within epithelia during homeostasis	Huang et al. (2021) Le et al. (2011)	Lu et al. (2017) Rheinheimer et al. (2023) Zubeidat et al. (2023)
Neutrophil	Highly present in neonatal glands Decrease during postnatal development and make up a low percentage of immune cells in adult homeostatic glands		Zubeidat et al., 2023)
NK cell/ILC1	TGFβ dependent NK‐like ILC1 cells develop alongside SG postnatal maturation Suggested to interact with macrophages via Csf2 and IL‐15		Cortez et al. (2016) Cortez et al. (2014); Hauser et al. (2020) Horeth et al. (2021) Rheinheimer et al. (2023) Zubeidat et al. (2023)
Mast cells	Identified by histology in mouse SG Identified in human MSG through scRNAseq Homeostatic functional role not explored	Huang et al. (2021)	Majeed (1994)
Adaptive Immunity
T‐cells	Detected in human and mouse homeostatic SGs CD4+ and CD8+ T‐cells home to SG upon direct or indirect viral infection CD8+ T‐cells interact with macrophages and migrate along epithelia for homeostatic surveillance	Horeth et al. (2023) Huang et al. (2021) Xiang et al. (2023)	Caldeira‐Dantas et al. (2018) Hauser et al. (2020) Horeth et al. (2021) Stolp et al. (2020) Thom et al. (2015) Woyciechowski et al. (2017) Zubeidat et al. (2023)
B‐cells	Produce immunoglobulins that can be secreted in saliva Dependent on microbial challenge, T‐cells and IL‐17 and readily described in postnatal SGs	Horeth et al., 2023 Huang et al., 2021 Xiang et al., 2023	Horeth et al. (2021) Rheinheimer et al. (2023) Zubeidat et al. (2023)

1.3. Innate immunity

The innate immune system is the first line of defense from tissue insult, and it includes epithelial cells that secrete salivary mucins and antimicrobial peptides. Common innate cells are neutrophils, macrophages and dendritic cells that produce major immune mediators including chemokines, cytokines and lipid mediators, which are important to trigger both tolerogenic/anti‐inflammatory or inflammatory responses.

Macrophages comprise the majority of immune cells present during tissue homeostasis in adult mouse SG (Lu et al., 2017, 2022; McKendrick et al., 2023). After ontogeny (described above), tissue resident macrophages self‐maintain their population and have tissue‐specific transcriptomic profiles and functional specialization (Guilliams et al., 2020). Macrophages can be simplistically classified into M1 (classically activated) and M2 (alternatively activated) polarized phenotypes although increased plasticity and heterogeneity is observed in vivo. Environmental stimulus plays an important role in macrophage function during homeostasis or disease states. The heterogeneity of macrophages has been demonstrated by recent studies using scRNAseq and flow analysis in both mouse and human SG (Costa‐da‐Silva et al., 2022; Huang et al., 2021; Lu et al., 2022;McKendrick et al., 2023; Zhao et al., 2023). Although there are some discrepancies in subpopulation markers, likely due to different technical strategies, two self‐maintained distinct macrophage subpopulations are present in steady‐state SG (Lu et al., 2022; McKendrick et al., 2023; Zhao et al., 2020, 2023) (Figure 2). Further, they are physically located in two niches, one surrounding blood vessels and nerves and one in contact with the epithelium, suggesting niche‐specific roles (McKendrick et al., 2023; Zhao et al., 2023). Macrophages line the epithelial ducts and acini and frequently cross the basement membrane barrier to localize between epithelial cells (Stolp et al., 2020). Macrophages communicate and facilitate tissue surveillance via other immune populations, described below, making them central players in gland homeostasis. Depleting macrophages during postnatal growth by treating mice with anti‐CSF‐1R, did not alter gland weight or gross morphology (Lu et al., 2022). In adult glands, transient depletion of macrophages did not affect expression of aquaporin 5, an acinar water channel or saliva flow after 90 days (Zhao et al., 2020), whereas permanent depletion led to a general epithelial disruption (McKendrick et al., 2023). Importantly, this suggests macrophage‐epithelial crosstalk during homeostasis as critical although further work is needed to understand the related specific interactions and physiological role of this.

FIGURE 2.

Mouse macrophage phenotype markers. Overview of phenotype markers used in mouse salivary glands. Only recent work focusing on macrophages and subpopulations are included in the overview. C, cluster; iMφ, infiltrating Macrophage, rMφ, resident macrophage. Figure created with www.biorender.com.

Dendritic cells (DC) serve as sentinels against environmental insults, such as infections and tissue damage. They are found in both lymphoid and non‐lymphoid tissues and connect innate sensing and adaptive immunity (Guermonprez et al., 2019). In mice, dendritic cells migrate from the oral mucosa and transiently accumulate in SG around weaning upon exposure to microbial challenges and can attract and activate T‐cells via the chemokine CCL19 (Zubeidat et al., 2023). They are present in both mouse and human adult steady‐state SG (Le et al., 2011; Lu et al., 2017). In mouse SG, two groups of bona fide classical DC types (cDC), cDC1 and cDC2, are recognized by CD103⁺CD11c⁻ and CD103⁻DC11c⁺, respectively (Lu et al., 2017). They are capable of antigen cross presentation and thus are suggested to surveil and maintain salivary immune homeostasis, although specific interactions are not clearly defined. In human SG, DC are abundant within the duct and acinar epithelium although specific cell–cell interactions are also not defined (Le et al., 2011).

Neutrophils are quickly recruited to the sites of inflammation and neonatal mouse SG have high levels of neutrophils, which quickly decline after birth (Zubeidat et al., 2023). They are attracted to the oral mucosa by several external stimuli, such as invading microbiota and activated epithelial cells. Nonetheless, only a small proportion originate from SG (Domnich et al., 2020). Formation of neutrophil extracellular traps by inflammatory stimulus is an essential step for SG stone development (Schapher et al., 2020).

Innate lymphoid cells (ILCs) are a group of immune cells with lymphoid morphology that lack antigen‐specific receptors. These cells contribute to both defense against invading pathogens as well as tissue homeostasis and repair through release of key cytokines or expression of immune checkpoint receptors, such as PD‐1 (Ryu et al., 2023). During postnatal development, expansion of a unique TGFb‐dependent natural killer (NK)‐like ILC1 population occurs in parallel to SG differentiation. These SG ILCs acquire markers of residency and, as other resident cells, may be important for tissue homeostasis as well as control of infection and tissue repair (Cortez et al., 2016). Recently, it was predicted that ILCs interact with Csfr2b ⁺ macrophages via Csf2 and IL‐15 during homeostasis, although the significance of this is not understood (Zhao et al., 2023). Interestingly, in human minor SG, most ILCs are the ILC3 subset (Kawka et al., 2023) and although their role in gut homeostasis is well known, their role in SG has yet to be studied.

Postnatally, mast cells are tissue resident cells found in most vascularized tissues and are potent immune effectors most known to mediate allergy and anaphylaxis (Chia et al., 2023). Although mast cells have been described by histology in homeostatic murine postnatal glands, recent scRNAseq analysis has not identified this population, although this could be a technical issue due to dissociation protocols (Hauser et al., 2020; Horeth et al., 2021; Majeed, 1994). Mast cell populations have been described in healthy human SG (Huang et al., 2021), and are reported to be increased in salivary cancers, with mucoepidermoid carcinomas containing more mast cells, than pleomorphic adenomas (Kamal et al., 2015). However, further work is needed to understand their role in steady‐state SG.

1.4. Adaptive immunity

Effective protection against internal and external pathogens depends on the fine coordination between innate and adaptive immune mechanisms. Adaptive immunity, involving T‐ and B‐cells, responds to specific antigens more sensitive and accurate than innate immune responses. In SG, systemic viral infections result in substantial formation of CD8⁺ and CD4⁺ tissue resident memory cells (TRM) that are largely disconnected from circulatory interchange (Thom et al., 2015). Interestingly, SG can recruit activated T‐cells independent of local inflammation or infection. T‐cell homing to both naïve and infected SG is dependent on endothelial VCAM‐1, a cellular adhesion molecule on endothelial cells, and integrin α₄b₁ (Woyciechowski et al., 2017). It is suggested that the chemokine receptor CXCR3 and its ligands are required for T‐cell migration to non‐inflamed glands, while CXC‐signaling is dispensable during gland infection, although further work is needed to confirm this (Caldeira‐Dantas et al., 2018). This establishes a mechanism for the efficient recruitment of activated T‐cells to SG after a systemic infection even though it may not directly involve the SG. While CD8⁺ TRM formation is independent of both antigen and inflammation, CD4⁺ T‐cells require local antigen for maintenance and TRM formation (Stolp et al., 2020). CD8⁺ TRMs survey the epithelial microenvironment by migrating along macrophages and accumulate at experimentally induced inflammatory hotspots, which allows them to short‐cut epithelial barriers (Stolp et al., 2020). Together, this cooperative effort may be adapted within branched organs to allow for maximal efficiency of barrier surveillance involving immune defense, repair and tissue homeostasis.

Plasma B‐cells are part of the adaptive mucosal immune system and may become disseminated to SG from inductive mucosal barriers (Aase et al., 2016). B‐cells are dependent on microbial challenge, T‐cells and IL‐17, which are all present in neonatal glands and are found at relatively steady frequencies during postnatal development (Zubeidat et al., 2023). In SG, B‐cells produce immunoglobulins that can be secreted into saliva via receptor‐mediated epithelial export and act as protective agents to fend off infections (Brandtzaeg, 2013). Recently, tissue resident B‐cells have been described in non‐lymphoid tissues such as the lung, kidney, liver and urinary bladder (C. M. Lee & Oh, 2022; Suchanek et al., 2023). These tissue resident B‐cells do not go back into circulation and contribute to bacterial defense and orchestrate macrophage polarization (Suchanek et al., 2023). This discovery is a paradigm shift in B‐cell biology, showing that homeostatic seeding of B‐cells is not limited to body cavities but extends to all major organs analogous to macrophages. Both mouse and human scRNAseq datasets have identified B‐cell populations during homeostasis (Chen et al., 2022; Horeth et al., 2023; Huang et al., 2021; Rheinheimer et al., 2023), however, whether these populations include bona fide resident B‐cells has not been determined.

The role of immune cells in SG development and homeostasis is an emerging field, and we are only beginning to investigate the subpopulations, niches and cell–cell interactions. The use of OMICs data provides an overview of cells present, however, more directed experimental models and studies involving enrichment of immune populations are needed to understand the potential of these subpopulations. In the following sections, we will summarize our knowledge on how immune cells contribute to both SG dysfunction and regeneration in response to various cancer therapies. Understanding these interactions may lead to novel targets to aid in prevention or resolution of adverse effects from cancer treatments.

2. IMMUNE CELL INVOLVEMENT IN IR THERAPY FOR HEAD AND NECK CANCERS

Head and neck cancer refers to malignancies of the upper aerodigestive tract, thyroid and SG and accounted for ~900,000 new cases world‐wide in 2020 (Sung et al., 2021). Depending on the cancer position and stage, IR is used alone or combined with surgery and/or chemotherapy, and IR is the most common treatment regimen. IR leads to cancer cell death by triggering DNA damage. In the tumor microenvironment, IR induces a plethora of inflammatory mediators activating resident macrophages and stimulating infiltration of circulating monocytes and DCs (Cytlak et al., 2022). While maturation of DCs promotes effector T‐cell responses which are required for efficacy and anti‐tumor effect, severe IR‐induced lymphopenia is associated with poor outcome for head and neck cancers (Cytlak et al., 2022; Kut et al., 2023; Y. Lee et al., 2009).

In addition to the anti‐tumor effect, IR also damages the surrounding non‐cancerous tissues such as SG, and loss of SG function is a commonly reported adverse event (Han et al., 2019). The degree of salivary IR damage is directly linked to radiation dose, and in patients, a combined dosage to the SG not exceeding 50 Gy is estimated to retain ~40% of salivary function, while doses under 25 Gy allow for complete recovery (Chibly et al., 2022; Jasmer et al., 2020). Despite technical advancements, approximately ~80% of head and neck cancer patients still experience salivary hypofunction post‐treatment (Jensen et al., 2010). Salivary hypofunction can be transient or become permanent and comes with further complications that diminish quality of life (Gupta et al., 2015). While clinical observation has given insight into histopathological features in chronic dysfunction, animal models reveal underlying mechanisms of acute tissue damage. Murine models have shown that SG undergoes an initial regenerative phase within the first days and weeks following IR damage, however, this is time‐limited, and the long‐term outcome is gland dysfunction (Chibly et al., 2022; Jasmer et al., 2020). Recent findings highlight the role of immune cells in these events (Figure 3).

FIGURE 3.

Immunomodulation in adverse effects with irradiation of head and neck cancer. Overview of immune cell involvement in irradiation of head and neck cancer. SASP, senescence associated secretory phenotype. Figure created with www.biorender.com.

3. ACUTE IR RESPONSES AND IMMUNE CELL INVOLVEMENT

IR with a single 5 Gy dose to the head and neck area in mouse models leads to immediate detection of double stranded DNA damage (Jasmer et al., 2020). DNA damage responses include cell cycle arrest, which allows for cells to undergo DNA repair and recovery. In SG, IR triggers an increase of cell cycle regulators p53 and p21, however, the number of cells in G2/M phase is not significantly changed, suggesting insufficient DNA repair (Zhang & Xiang, 2023). Consequently, within the first 3 days, there is extensive epithelial and endothelial apoptosis, with decreased progenitor proliferation and acinar and duct cell markers (single doses of 5–15 Gy) (Emmerson et al., 2018; May et al., 2018; Mizrachi et al., 2016; Weng et al., 2018). This is followed by apoptosis‐induced compensatory proliferation, which is a common reparative response in injured tissues (Jasmer et al., 2020). Increased proliferation is detected in acinar cells at day 3 and in duct progenitors at day 7 post‐IR (Emmerson et al., 2018; May et al., 2018; Weng et al., 2018). This leads to lineage restricted duct and acinar recovery within 14–30 days, at which point acinar markers are comparable to control levels (Emmerson et al., 2018; May et al., 2018; McKendrick et al., 2023; Weng et al., 2018).

The role of immune cells in the initial phase post‐IR remains enigmatic. Acute SG IR damage does not increase chemokines and cytokines and there is no measurable recruitment of immune cells, including neutrophils or monocytes (Lombaert et al., 2008; McKendrick et al., 2023; Urek et al., 2005). Whether this is due to specific radiosensitivity of innate and adaptive immune cells in SG is not known, but it could explain the lack of acute inflammatory infiltration. It is noteworthy that an increasing number of studies have demonstrated a heterogenous quantitative and functional response of different components of the immune system to IR (Lumniczky et al., 2017). Low doses of IR, below 0.5 Gy, induce the release of oxygen and nitrogen radicals that can induce apoptosis and stimulate an anti‐inflammatory response whereas high doses, above 1 Gy, cause substantial cell death by necrosis, release of damage‐associated molecular pattern and further activation of the adaptive immune response by DCs (Formenti & Demaria, 2009; Lumniczky et al., 2021). Whether SG DCs are specifically radiosensitive is not known, and it is not clear why there is not an initial immune infiltration, which could potentially aid in the long‐term recovery.

Despite the absence of cellular infiltration, recent advances have highlighted the essential role of macrophages in this acute recovery period. Macrophages are radiosensitive with decrease of cell‐specific markers and almost complete loss of cycling cells within 3 days post IR (McKendrick et al., 2023; Zhao et al., 2020, 2023). However, surviving tissue resident macrophages recover independently of circulating monocytes, and by 30 days post IR, macrophage and ILC levels are comparable to control (10–15 Gy) (McKendrick et al., 2023; Zhao et al., 2020, 2023). In irradiated SG, these macrophages have increased physical association with DNA damaged epithelia, suggesting phagocytic activity aiding in the acute recovery. Indeed, depletion of macrophages prior to IR leads to increased epithelial apoptosis and reduced compensatory proliferation (McKendrick et al., 2023). Depletion of macrophages at day 17 post IR either by using the Mafb ^Cre/+:Cx3cr1 ^LSL‐DTR/+ genetic model or CSF‐1R antibody results in significant loss of acinar markers and saliva secretion by day 28, indicating that continued presence of functional macrophages is essential for acute recovery (McKendrick et al., 2023). Recently, it was shown that a specific subset of macrophages and ILCs are Sonic Hedgehog (Shh) responsive. AAV transfection of Shh or Gli1, which activates Shh pathway in vivo, 3 days after IR resulted in increased macrophage and ILCs markers, suggesting stimulation of these populations (Zhao et al., 2020, 2023). Accordingly, this transfection is protective of salivary flow post IR damage (Zhao et al., 2020). Whether there are additional targets that can specifically stimulate in situ macrophages to improve the post‐IR outcome is not clear, but this is an attractive avenue for further research.

Alternatively, transplanting immune cells or stimulating infiltration have also been shown to have positive outcomes. Isolated effectively conditioned peripheral blood mononuclear cells (E‐MNCs) are enriched in CD11b⁺ M2 macrophage‐like cells (Honma et al., 2023). Transplanting the macrophage‐like subset into SG 3 days after IR, promoted endothelial recovery, epithelial proliferation and improved saliva flow rates (Honma et al., 2023). In line with this, treatment with FMS‐like tyrosine kinase‐3 ligand (Flt‐3 L), stem cell factor (SCF) and granulocyte colony‐stimulating factor (G‐CSF) in IR damaged SG, promote immune infiltration and endothelial proliferation resulting in increased acinar cell number and saliva output (Lombaert et al., 2008). Taken together, these studies show how immune cells, particularly macrophages, play an essential role in initial SG damage response and recovery.

4. IMMUNOMODULATION AND LONG‐TERM IR DAMAGE

Despite the regenerative response during the acute phase, late effects of IR damage eventually result in SG dysfunction. Long‐term effects are considered driven by microenvironmental changes including accumulation of cellular senescence and progenitor disruption, vascular and neuronal depletion, chronic impairments of cell metabolism, immune infiltration and fibrosis (Chibly et al., 2022; Jasmer et al., 2020).

Accumulation of senescent cells due to aberrant DNA repair leads to secretion of a variety of soluble mediators collectively known as senescence associated secretory phenotype (SASP), which is reported in both mice and humans post‐IR (Peng et al., 2020; Zhang & Xiang, 2023). Secretion of different forms of SASP components, including cytokines, chemokines, nucleotides and others, can regulate the cellular and local tissue microenvironment, activating immune responses that can range from pro‐ to anti‐inflammatory (Elder & Emmerson, 2020). Indeed, in contrast to the short‐term response, chronic IR damage is linked to increases in infiltrating immune cells in both mice and humans.

Bulk and scRNAseq identified transcriptional changes related to innate and adaptive immune responses 1 year after fractionated IR (5 × 6 Gy) in the CH3 mouse model (Lombaert et al., 2020). In control glands, scRNAseq identified several immune populations including macrophages, DCs, NK cells, B‐cells and T‐cells that were clustered into five subpopulations: CD4⁺, CD4⁺CD8⁺, CD8⁺, FoxP3⁺ and CXCR6⁺ (Rheinheimer et al., 2023). IR led to a long‐term overall increase in immune cells, and specifically CD4⁺CD8⁺ T‐cells had the highest number of differentially expressed genes compared to control. Bioinformatic analysis suggested active involvement of immune cells, particularly by T‐cells, in mediating cellular responses in long‐term IR damage (Rheinheimer et al., 2023). Further work is needed to understand the functional relevance of these interactions and if they are clinically relevant. Another long‐term study identified increases in macrophages (CD68⁺) and T‐cells (CD3⁺) by histology using the C57BL/6 inbred strain, suggesting immune changes as a common feature between mouse strains. Long‐term immune changes are also a feature in IR‐damaged human glands. A recent study analyzing bulk RNAseq of IR‐damaged SG identified immunomodulation along with fibrosis and neurotrophic signaling as the top dysregulated pathways (Chibly et al., 2023). In line with this, histopathological findings in human SG treated with doses between (60–70 Gy), include acinar atrophy and periacinar inflammatory infiltrates of CD4⁺, CD8⁺ and Granzyme B⁺ T‐cells in addition to B‐cells and macrophages (Teymoortash et al., 2005). Taken together, increased infiltration of immune cells over time in IR‐damaged SG is evident in both mouse models and human patients; however, there is not enough data to identify similarities between model systems and key cell–cell interactions that are potential common responses to IR. It is important to consider that if long‐term immune‐related salivary dysfunction is due to senescence, the composition of SASP is an important variable, and can also differ between mouse and human outcomes due to cross‐species discrepancies in soluble products and receptors in certain cells (Walle et al., 2022).

A major hallmark of chronic IR damage is the development of fibrosis and acinar atrophy (Chibly et al., 2022; Jasmer et al., 2020). Fibrosis, defined as excess deposition of collagen and other extracellular matrix molecules, produces a dysregulated microenvironment which can alter molecular and mechanical cues and is closely related to and driven by immunomodulation. The presence of fibrosis as a consequence of chronic IR injury is also reported in other organs (Yu et al., 2023). Many studies show evidence that several SASP components are implicated in fibrosis such as IL‐6, IL‐1b, TGFβ, EGF, VEGF, metalloproteinases and others (Borrelli et al., 2019; Citrin et al., 2013; Su et al., 2021). Furthermore, macrophage senescence induced by IR also plays an important role in the fibrotic niche secreting SASP molecules promoting a profibrogenic phenotype of fibroblasts (Borrelli et al., 2019; Su et al., 2021). The fibrotic process also includes other immune populations such as CD8⁺ and CD4⁺ T‐cells, B‐cells, DCs, mast cells and ILCs (Distler et al., 2019). The contribution of different populations may vary between tissues; however, a common feature is polarization towards Th2 cell‐mediated responses such as secretion of IL‐4 and IL‐13 (Distler et al., 2019). Both CD4⁺ and CD8⁺ T‐cells can secrete IL‐13, which can directly activate fibroblasts as well as induce senescence and increase SASP factors IL‐6 and IL‐1b (Distler et al., 2019; Zhu et al., 2022).

Notably, preclinical mouse models do not consistently progress to fibrosis despite exhibiting decreased salivary function. The cause of this variation is not clear; however, the genetic background of inbred mouse models affects the immune system and, for example, strains differ in regulatory T‐cell phenotypes (Tanner & Lorenz, 2022). It is important to consider that this may affect long‐term outcomes, including progression to fibrosis. Further, it is also clear that human and porcine glands develop fibrosis and epithelial atrophy with higher degree of severity than murine models (Jasmer et al., 2020; Lombaert et al., 2020). There is no clear consensus on the underlying cause for this, however, an important difference between model systems and the clinical setting is the presence of head and neck cancer. Whether such tumors lead to alterations in base‐line composition of immune populations within SG is not clear, but cancer patients have elevated salivary levels of the pro‐inflammatory cytokines IL‐6 and IL‐8 compared to healthy controls (Chiamulera et al., 2021). Further, both salivary and circulating cytokines increase following treatment in patients undergoing IR, suggesting immune responses in cancer patients may differ from healthy model systems (Astradsson et al., 2019; Russo et al., 2016). This suggests a deregulated immune response in cancer patients that may have profound effects on the immune reaction to IR damage and long‐term outcomes, and further work is needed to investigate this.

Regardless of the severity of damage, immune alterations is a central feature in both animal models and human glands. Deconvoluting the identity of infiltrating cells and their role in parenchymal atrophy and fibrosis will be instrumental to understand the damage progression and to uncover targets for novel therapies.

5. HEMATOLOGIC CANCER TREATMENT AND SALIVARY DYSFUNCTION

For patients diagnosed with hematologic cancers such as lymphoma, myeloma and leukemia, allogeneic hematopoietic stem cell transplant (HSCT) is a standard treatment. Total body irradiation alone or in combination chemotherapy are used as a preparative regimen for allogeneic hematopoietic stem cell transplant (HSCT) to eliminate residual tumor cells and to provide immunosuppression to prevent graft rejection (Gyurkocza & Sandmaier, 2014). Graft‐versus‐tumor (GVT) response is the desired outcome and derives from effective interaction between donor‐derived leukocytes and host tumor cells. However, various complications such as disease recurrence and development of GVHD can occur after transplant. GVHD, with pathogenesis that can be classified as acute or chronic, is a multiorgan syndrome that occurs when donor allogeneic T‐cells attack recipient tissues and organs beyond the cancer. The equilibrium between effector (e.g. Th1/Tc1) and regulatory (e.g. Tregs) mechanisms during HSCT is essential to establish a mixed chimerism where both stem cells from donor and host cells can coexist. This balance is essential since both GVT and GVHD activity are modulated by host antigen‐presenting cells (APCs) and donor T‐cells (Hamers et al., 2019).

Acute GVHD, that affects approximately half of patients receiving matched HSCT, is an inflammatory response to the damage caused by the conditioning regimen, and it affects three organs – skin, liver and gastrointestinal tract. Tissue damage and loss of epithelial barrier function leads to translocation of bacterial components and release of endogenous danger signals (Heidegger et al., 2014). These signals induce proinflammatory cytokines, activation of APCs and subsequently differentiation of alloreactive donor T‐cells (Heidegger et al., 2014). Chronic GVHD, whose incidence ranges from 30%–50% of patients receiving allogeneic HSCT, is more heterogeneous and resembles an autoimmune disease. Although chronic GVHD pathophysiology is not completely understood, it involves impaired mechanisms of tolerance, both central and peripheral, with the presence of effector auto‐ and allo‐reactive donor T‐ and B‐cells (Cooke et al., 2017). Chronic GVHD affects multiple organs, including the oral cavity and SG, referred to as oral GVHD (Cooke et al., 2017) (Figure 4).

FIGURE 4.

Immunomodulation in adverse effects with hematologic cancer therapy. Overview of immune cell involvement in Graft‐versus‐ tumor and Graft‐versus‐host disease. Figure created with www.biorender.com.

Oral GVHD is reported in approximately 70% of patients with chronic GVHD and disease manifestations include lichenoid lesions and SG dysfunction which leads to decreased production of saliva (Fall‐Dickson et al., 2019; Nagler & Nagler, 2004). Besides reduced oral lubrication, hyposalivation alters salivary proteins with increased mucoid viscous saliva, increased adherence of bacterial plaques and debris, and elevated risk of dental caries (Mays et al., 2013; Tollemar et al., 2023). Little is known about disease progression in SG specifically, and the literature is mostly based on histopathological assessment of human minor SG, which show duct dilation and acinar atrophy, apoptosis, diffuse lymphocytic infiltrate and fibrosis (Soares et al., 2005). The immune infiltration is mainly composed by CD4⁺ and CD8⁺ T‐cells and CD68⁺ macrophages while CD20⁺ B‐cells are not increased compared to controls (Soares et al., 2005). Both Th1 and Th2 cytokines increase, and there is a close association between strong infiltration and Th2 cytokines, macrophage‐derived chemokine and the CCR4 (Hayashida et al., 2013). Preclinical studies provide some insight into pathophysiological mechanisms involved in oral GVHD and they confirm what is observed in patients, where periductal lymphocytic infiltration, parenchymal destruction and fibrosis are followed by a significant decrease of saliva secretion (Chu & Gress, 2008; Nagler & Nagler, 2004). Currently there is no permanent treatment for oral GVHD and available options are focused on reduction of immune‐damage and symptom management (Fall‐Dickson et al., 2019).

Although major developments have been made in the understanding of GVHD pathogenesis in general, several key questions remain unanswered. For example, it is not clear why some patients develop only the chronic form of the disease and what key factors limit disease to a single organ as opposed to extending to several organ systems. Although T‐cell dysfunction is a well‐known factor involved in GVHD onset, the molecular requirements for generating specialized subsets of effector cells able to maintain the disease have yet to be identified. In this scenario, multi‐OMICs studies, ranging from single‐cell analyses to spatial surveys, as well as improved pre‐clinical models are critical steps to gain insights into the complex mechanisms underlying chronic GVHD. Answering these key questions will facilitate the development of more accurate, non‐invasive diagnostic tools and new therapeutic strategies.

6. IMMUNE CHECKPOINT INHIBITORS (ICI) AND SICCA SYNDROME

Tumors have a complex microenvironment, which includes fibroblasts, endothelial cells, pericytes, various tissue‐resident cells and diverse immune cell types (de Visser & Joyce, 2023). The tumor microenvironment plays a critical role in cancer pathogenesis by forming an ecosystem supporting cancer cell growth and metastasis. Both myeloid immune cells in the microenvironment and cancer cells frequently overexpress the immune checkpoint ligand PD‐L1 (programmed cell death protein ligand), CD80 which interacts with the receptor PD1 (programmed cell death protein 1) and CTLA‐4 (cytotoxic T‐lymphocyte associate protein 4). Physiologically, this interaction maintains immune homeostasis and prevent autoimmunity. However, overexpression of immune checkpoints in cancers results in immune surveillance suppression and avoidance of immune attack allowing for undisturbed tumor growth. By using ICI, which are monoclonal antibodies, the immune suppression is blocked, which increases T‐cell antitumor activity. Therapeutic use of checkpoint inhibitors has revolutionized cancer treatment and is now standard care for many cancers and has also been used to treat head and neck and salivary cancers (Sato et al., 2022; Shiravand et al., 2022). Even advanced cancers can effectively be treated using ICI therapy and to date, PD‐1, PD‐L1 and CTLA‐4, are all FDA approved ICI cancer treatments (Nurieva et al., 2021). However, ICI therapy is also associated with immune‐based attack on normal tissues, known as immune‐related adverse effects (irAE), due to excessive activation of T‐cells and downregulation of regulatory T‐cells (Khan & Gerber, 2020). Altered interactions between T‐ and B‐cells leading to production of autoantibodies has also been associated with irAEs (Khan & Gerber, 2020). These adverse effects can occur in any organ including SG that result in oral manifestations, referred to as immune checkpoint inhibitor sicca (ICI sicca) (Warner & Baer, 2021) (Figure 5).

FIGURE 5.

Immunomodulation in adverse effects associated with immune checkpoint inhibitors. Overview of immune cell involvement in immune cell inhibitor (ICI) Sicca. Figure created with www.biorender.com.

ICI sicca is characterized as an abrupt and severe onset of dry mouth and may also include lichenoid lesions, ulcers and erythema multiforme (Shazib et al., 2020; Warner et al., 2019). The degree of hyposalivation varies depending on the combinational treatments but can be total loss of function (Pringle, van der Vegt, et al., 2020; Warner et al., 2019). The condition is reported in patients with or without systemic symptoms or in the setting of preexisting autoimmune disease (Warner & Baer, 2021). Biopsies from ICI sicca patients show immune infiltrates ranging from mild to chronic inflammation. The infiltrate is dominated by CD4⁺ and CD8⁺ T‐cells with little to no B‐cells, which distinguishes ICI sicca from the autoimmune disease Sjögrens (Pringle, van der Vegt, et al., 2020; Warner et al., 2019). Severe inflammation is accompanied by acinar atrophy, duct injury, nuclear enlargement, apoptosis and fibrosis (Ortiz Brugués et al., 2020; Warner et al., 2019). Epithelial cells also significantly upregulate expression of PD‐L1, which is suggested to represent an attempt at protection from the PD‐1 expressing immune cells (Pringle, Wang, et al., 2020; Warner et al., 2019). Still, there is no clear explanation for SG involvement, and further work is needed to understand immune‐epithelial interactions.

To date, there is no clear and effective clinical management of ICI sicca. Ending or pausing ICI regimen and treating with corticosteroids may ease the symptoms, however, SG damage or sustained autoimmunity may continue long after completion of treatment leading to persistent dry mouth (Ortiz Brugués et al., 2020; Warner et al., 2019). The etiology and underlying mechanisms of ICI sicca are not understood, and further work is needed to identify key signaling factors and causative events to pinpoint potential effective management. Identification of early inflammatory biomarkers or predictive signatures at risk could also reduce the chances of developing this comorbidity and possibly prevent it.

7. CONCLUDING REMARKS

Patient quality of life is severely diminished by side effects from many life‐saving cancer therapies. In this review, we have summarized the current knowledge on immune cell involvement in some treatments leading to adverse events targeting SG. The role of immune cell populations in SG development and regeneration is an emerging field and gaining further knowledge will aid in understanding homeostatic, damage and repair responses. Although involvement and infiltration of immune cells are central for the three major types of cancer treatments reviewed here that lead to adverse SG outcomes, more experimental data are required to directly compare or identify potential common damage responses. Enhancing our knowledge of immune cells as part of SG homeostasis and how cancer treatments affect the SG immune compartment are important steps towards understanding immunomodulation of damage, which will ultimately provide targets to help repair or regenerate SG function and improve patients' quality of life.

AUTHOR CONTRIBUTIONS

Ana C. Costa‐da‐Silva: Conceptualization; investigation; writing – original draft; writing – review and editing; visualization. Carlos U. Villapudua: Visualization; writing – review and editing. Matthew P. Hoffman: Funding acquisition; writing – review and editing. Marit H. Aure: Conceptualization; visualization; supervision; writing – original draft; writing – review and editing.

CONFLICT OF INTEREST STATEMENT

The authors are unaware of any affiliations, memberships, funding or financial holdings that might affect the objectivity of this review.

Costa‐da‐Silva, A. C. , Villapudua, C. U. , Hoffman, M. P. , & Aure, M. H. (2025). Immunomodulation of salivary gland function due to cancer therapy. Oral Diseases, 31, 2680–2694. 10.1111/odi.14972

DATA AVAILABILITY STATEMENT

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.