The oral cavity comprises a large and heterogeneous group of malignancies with distinctive histological features, among which oral squamous cell carcinoma (OSCC) is the most common (1). Such heterogeneity is also evident within OSCC at different levels: starting from its etiopathogenesis, where associations with lifestyle habits (smoking, alcohol consumption, exposure to wood and leather dust) and viral infections [human papillomavirus (HPV) and Epstein-Barr virus] have been demonstrated, continuing with its various histopathological variants (such as verrucous carcinoma, basaloid carcinoma, papillary squamous carcinoma, spindle cell carcinoma, acantholytic, or lymphoepithelial carcinoma) and ending with the degree of tumor-associated inflammation (2,3). Although tumor-infiltrating immune cells are recognized as important biomarkers in several malignancies, such as breast, colorectal and lung cancers, their prognostic role in OSCC remains controversial (4,5). Currently, based on the distribution of inflammatory cells within the tumor microenvironment (TME), 'hot' tumors (characterized by a significant infiltrate of tumor-associated immune cells) can be distinguished from 'cold' tumors (showing an almost complete absence of immune cells; Fig. 1) (6).

These findings have markedly influenced oncological decision-making in patients with OSCC, particularly in advanced and metastatic disease (4,7). Similar awareness is increasingly emerging across other therapeutic modalities, emphasizing that the cellular constituents of TME exert a substantial effect on patient prognosis and may even serve as therapeutic response predictors (5,7).

The present review provided a comprehensive overview of the principal cellular components of the OSCC TME and discussed their intrinsic involvement in mechanisms of immune evasion, immunosuppression and response to immunotherapy.

The concept of TILs was first described in 1986 to denote immune cells [mainly T, B and natural killer (NK) lymphocytes] within the TME, where they play a central role in anti-tumor immunity (8). TILs are established prognostic biomarkers in several malignancies, however, their prognostic relevance in OSCC remains controversial (4).

According to the International Immuno-Oncology Biomarker Working Group guidelines (5), TIL density should be reported as the percentage of the stromal area occupied by lymphocytes, a method originally developed for breast cancer and later extended to other tumor types (4). Nonetheless, the literature shows considerable variability in the cut-off values used to distinguish high from low TIL infiltration, with 20 and 30% thresholds being the most frequently adopted in the absence of formal consensus (7).

Furthermore, based on TILs spatial distribution within the TME, solid tumors can be classified into three major immune phenotypes: Immune-inflamed, immune-excluded and immune-desert (Fig. 1). Immune-inflamed tumors are characterized by a dense infiltration of immune cells within the tumor parenchyma. By contrast, immune-excluded tumors display immune cells confined to the peritumoral stroma, without effective infiltration of the tumor core. The immune-desert phenotype ('cold' tumor), exhibits an almost complete absence of immune infiltration. This may result from several mechanisms, including defective recruitment of antigen-presenting cells, impaired T-cell activation or migration and altered cytokine production (5,9). Low Human Leukocyte Antigen (HLA) and interferon-γ (IFN-γ) expression levels have also been reported in such tumors (10). Notably, in a study by Troiano et al (9), all OSCC cases with an immune-desert phenotype were classified as stage IV and exhibited markedly poorer overall survival compared with the other immune profiles.

Macrophages are innate immune cells characterized by remarkable functional plasticity. They can exhibit two main phenotypes: M1 (classically activated) and M2 (alternatively activated). M1 macrophages produce type I proinflammatory cytokines such as IL-1β, IL-1α, IL-12 and TNF-α. They contribute to tumor suppression by presenting tumor-specific antigen and eliminating cancer cells through reactive oxygen species production or antibody dependent cellular cytotoxicity (41).

Conversely, M2 macrophages secrete type II cytokines, including IL-4, IL-6 and IL-10, thereby promoting anti-inflammatory responses. They stimulate tumorigenesis, angiogenesis, matrix remodeling and metastasis and can be identified by surface markers CD163, CD204 and CD206. Macrophages are abundant in the TME of solid tumors, where they are termed tumor-associated macrophages (TAMs); most localize in the tumor stroma and are polarized toward the M2 phenotype (42).

Tumor cells drive M2 polarization of TAMs, while M2 TAMs in turn enhance tumor progression, establishing a positive feedback loop that reinforces malignancy. OSCC cells actively recruit peripheral monocytes to the TME via chemokines such as CCR7 (43) and promote M2 polarization through cytokines such as TGF-β (38), PAI-1 and IL-8 (44), as well as through tumor-derived extracellular vesicles (EVs) (43,45).

TAMs promote cancer progression directly and indirectly. Directly, they secrete growth factors (such as EGF, TGF-β and CCL2) that stimulate cancer cell proliferation and invasion, aided by metalloproteinases (MMPs) (46-48). M2 macrophage-derived EVs enhance OSCC malignancy by delivering miR-23a-3p, which promotes tumor progression via phosphatase and tensin homolog (PTEN) targeting (46). Indirectly, TAMs shape an immunosuppressive milieu by expressing PD-1 ligands (PD-L1 and PD-L2) and immunosuppressive cytokines, thereby inhibiting lymphocyte function, promoting FOXP3+ Treg expansion and reducing immunotherapy efficacy (17).

TAM infiltration is also linked to increased microvessel density (MVD), facilitating angiogenesis, tumor growth and metastasis. TAMs secrete vascular endothelial growth factor (VEGF) and other molecules that promote endothelial cell proliferation, basement membrane degradation and migration, resulting in neovascularization (49,50). Furthermore, TAMs are enriched within lymph node metastasis (LNM), where they remodel the TME by activating fibroblasts and inducing T-cell exhaustion through SPP1-CD44 and CD155-CD226 interactions, aiding metastatic colonization (51).

Fusobacterium nucleatum (Fn), an oral commensal bacterium implicated in chronic periodontitis, also influences macrophage-tumor interactions. Nie et al (52) demonstrated that Fn activates the NFκB pathway in OSCC cells and macrophages, inducing C-X-C motif chemokine ligand 2 secretion that enhances OSCC proliferation, migration and macrophage M2 polarization. Weber et al (53) reported increased macrophage infiltration and M2 polarization in transforming oral leukoplakia (OLP) compared with non-transforming OLP and normal mucosa, suggesting that M2 macrophages facilitate malignant progression by secreting growth factors. Smoking appears to promote this process by inducing M2 polarization (54).

TAMs also contribute to therapy resistance. HPV-negative HNSCC exhibits greater M2 macrophage infiltration than HPV-positive tumors. In this context, TAMs confer radioresistance by releasing heparin-binding EGF-like growth factor (HB-EGF), which activates the non-homologous end-joining (NHEJ) DNA repair pathway, promoting post-radiation survival (55). Similarly, EVs from M2 macrophages can decrease tumor sensitivity to fluorouracil and cisplatin by activating AKT/GSK-3β signaling in cancer cells, leading to apoptosis resistance and enhanced proliferation (56).

Studies have shown that TAM density, particularly CD163+ M2 macrophages, is higher in HNSCC than in normal mucosa, associated with LNM and poor survival (20). A meta-analysis by Chohan et al (57) confirmed that elevated CD163+ TAM levels markedly associate with poor prognosis in OSCC, consistent with several other studies (42,58).

Neutrophils are innate immune cells frequently found in solid cancers as tumor associated neutrophils (TANs). Similarly to macrophages, TANs may acquire either antitumor (N1), or pro-tumor (N2) activity depending on the cytokines profile of TME. N1 TANs express high levels of IL-12, tumor necrosis factor-α (TNF-α), TRAIL and chemokines CCL3, CXCL9 and CXCL10. Their antitumor effects occur directly, via phagocytosis, degranulation of antimicrobial peptides, generation of reactive oxygen species), or indirectly, through antibody-dependent cytotoxicity and activation of other immune cells (T cells, NK cells, B cells and dendritic cells) in a pro-inflammatory and immunostimulatory manner (59,60).

By contrast, N2 TANs promote tumor growth and dissemination by secreting extracellular matrix (ECM) remodeling enzymes, MMPs and pro-angiogenic factors such as VEGF and immunosuppressive cytokines (IL-10, TGF-β and IL-2). These mediators recruit Tregs, myeloid-derived suppressor cells (MDSCs) and cancer-associated fibroblasts (CAFs), enhancing tumor immune evasion (17,59,60).

TANs typically display an antitumor N1 phenotype at early stages but progressively shift toward a tumor-promoting N2 phenotype as cancer advances, consistent with the immunoediting theory. TGF-β plays a central role in this transition (61). This polarization model, initially identified in mice, has also been demonstrated in human peripheral blood. In cancer patients, high-density neutrophils (HDNs) and low-density neutrophils (LDNs) reflect N1 and N2 phenotypes, respectively. LDNs comprising immature MDSCs and mature HDN-derived cells in a TGF-β-dependent manner, appear transiently during self-limiting inflammation but accumulate persistently in cancer, exhibiting immunosuppressive functions unlike mature HDNs (62).

In OSCC, tumor cells and the TME regulate TAN recruitment, which reciprocally influences tumor growth and TME composition. Chemerin, a multifunctional adipokine secreted by OSCC cells, promotes neutrophil chemotaxis in a concentration-dependent manner. High chemerin expression correlates with increased TAN density, LNM, higher clinical stage, recurrence and reduced OS and DFS, thus serving as an independent prognostic factor (63).

Yu et al (64) demonstrated that TGF-β1 and IL-17A synergistically induce N2 polarization, characterized by high MMP-9 and low CCL3 expression. The main sources of these cytokines are tumor and immune cells within the TME. N2 TANs facilitate OSCC progression by inducing epithelial-mesenchymal transition (EMT) via NF-κB activation and releasing immunosuppressive cytokines. Circulating neutrophils in OSCC patients exhibit a phenotype analogous to TAMs (MMP-9 high/CCL3 low).

Another mechanism contributing to neutrophil-mediated tumor progression is the formation of neutrophil extracellular traps (NETs), web-like DNA structures decorated with histones and enzymes such as myeloperoxidase (MPO), neutrophil elastase (NE)and MMP-9. While MPO may exert cytotoxic effects and histones induce thrombosis (reducing tumor perfusion), NET proteases degrade ECM components, facilitating invasion and metastasis (65,66).

NETs have been detected in OSCC tissues by Garley et al (65,67) and Zhai et al (68) using multiplex immunohistochemistry (mIHC) co-localization of citrullinated histone H3 (citH3) and MPO. NET-rich tumors were associated with larger size and LNM, suggesting a role in disease progression. Also, in vitro, OSCC cells exposed to NET-containing conditioned media exhibited EMT-related gene upregulation, pyroptosis inhibition and increased proliferation and invasion, effects primarily attributed to NE (68).

The oral microbiota also modulates neutrophil-tumor interactions. Streptococcus mutans, which relocates from the mucosal surface to tumor tissue during OSCC development, stimulates cancer cells to produce the oncometabolite kynurenic acid (KYNA). KYNA reprograms the TME by expanding N2 TANs, which release IL-1β, leading to CD8+ T-cell exhaustion and FOXP3+ Treg induction, thereby fostering tumor progression (69).

Across several solid cancers, a high TAN density in histologic sections and an elevated neutrophil-to-lymphocyte ratio (NLR) in peripheral blood are associated with advanced stage, LNM and poor prognosis, making them potential prognostic biomarkers (59,70). Studies in OSCC confirm that increased TAN infiltration and elevated NLR are markedly associated with worse clinical outcomes and reduced OS and PFS (63,70-72).

Mast cells (MCs) are tissue-resident myeloid cells of the innate immune system whose cytoplasm contains basophilic granules rich in histamine, leukotrienes, tryptase, chymase, heparin and inflammatory mediators. MCs are well known for their roles in allergic and autoimmune diseases but are also present in various solid tumors, where they exert tumor-promoting or tumor-suppressive effects depending on tumor type and activation state (73,74).

In OSCC, their role remains controversial. This inconsistency probably arises from variations in case selection, detection methods (such as toluidine blue, Alcian blue-safranin, or anti-tryptase antibodies) (75), MC subtypes (chymase+ vs. chymase-) and tumor site. Some authors reported a positive association between MC density and histological grade, lymphovascular invasion and depth of invasion (76), although its prognostic significance remains uncertain (77,78).

Cai et al (74) demonstrated in their findings that MCs are involved in oral cancer progression and may serve as a potential diagnostic and prognostic marker. Current evidence suggests that MCs are most active in the early stages of OSCC tumorigenesis. Shrestha et al (79) observed the lowest MC count in normal mucosa and significant increases in OPMDs with dyspasia and OSCC, indicating a role in malignant transformation. IL-1 released by MCs promotes epithelial proliferation in OLP, while histamine increases mucosal permeability, facilitating irritant infiltration and enhancing the risk of dysplastic transformation (79,80). Degranulated MCs also release TNF-α, which activates additional MCs, sustaining chronic inflammation and promoting carcinogenesis (81). By contrast, in a systematic review, Tzorakoleftheraki and Koletsa (78) described that some studies propose that mast cells may exert antitumor effects by enhancing immune responses against the cancer cells.

Several studies have reported an increase in MC density in OPMDs compared with OSCC, probably due to TME remodeling and impaired MC migration (79,82). Conversely, Iamaroon et al (83) and Michailidou et al (84) observed persistent or increased MC infiltration in OSCC, particularly within the lamina propria and invasive front, as well as in high MVD areas, supporting their role in invasion and angiogenesis.

MCs facilitate these processes by releasing proteases (tryptase, chymase) that activate MMPs, leading to ECM degradation and metastasis. They also promote angiogenesis through VEGF, tryptase, IL-8 and histamine release (81,85,86), explaining the observed increase in MVD during disease progression (83,87).

Fibroblasts are the most abundant non-immune cell type within the TME of solid neoplasms, where they are referred to as CAFs. In OSCC, as in other solid tumors, CAFs predominantly originate from quiescent oral mucosal fibroblasts activated by a variety of tumor-derived cytokines, notably TGF-β, angiopoietin-like protein 3 (ANGPTL3) (88) and platelet-derived growth factor-BB (PDGF-BB) (89). However, pericytes, adipocytes, endothelial cells and bone marrow-derived mesenchymal stem cells have also been identified as potential CAF progenitors in specific contexts (90,91). Most CAFs display a myofibroblastic phenotype, virtually absent in normal oral mucosa (92), similar to that seen in healing wounds and fibrotic disorders. The most commonly used markers for their identification include α-smooth muscle actin (α-SMA), integrin α6 and fibroblast activation protein (FAP) (20). However, these markers lack complete specificity (being expressed by pericytes and smooth muscle cells) and sensitivity (as not all CAF subsets express them) (91).

Attempts to subclassify fibroblasts according to their phenotype, secretory profile, or gene expression have been made, but characterization of CAF heterogeneity in OSCC remains incomplete. CAFs probably represent a subpopulations spectrum with partially overlapping functions and the ability to interconvert (88). In OSCC, CAFs are generally tumor-promoting, with the following main roles: i) Production of cytokines and growth factors that support cancer growth and stimulate fibroblast activation; ii) provision of metabolic support to cancer cells; iii) promotion of tumor invasion; iv) stimulation of angiogenesis; v) facilitation of immune suppression; and vi) induction of therapy resistance (90-92).

Inflammatory cells represent a principal component of the TME (116) and their interactions with cancer cells constitute a major mechanism enabling carcinomas to evade immune surveillance (5,117). Immunotherapy can be divided into two categories: Active, in which the immune system directly targets tumor cells, and passive, in which costimulatory mechanisms enhance immune activation through cellular receptors, conferring cytotoxic capacity. The discovery of immune checkpoints (ICs), which modulate inflammatory cell activity and consequently tumor progression, has profoundly transformed cancer therapy (118-120).

Among ICs, PD-1 and its ligand PD-L1 are key regulators of T-cell suppression and immune escape (121,122). PD-L1 is a transmembrane protein expressed on tumor and immune cells, including macrophages, dendritic cells and T cells, where it binds to PD-1 on activated lymphocytes to attenuate immune activity. This interaction downregulates TCR signaling, decreases cytokine production and induces T-cell apoptosis or exhaustion, thereby promoting immune tolerance within the TME (123-128). In OSCC, PD-L1 expression has been associated with tumor aggressiveness, advanced clinical stage, LNM and poor prognosis. Hypoxia, oncogenic pathways (such as PI3K/AKT and MAPK) and proinflammatory cytokines like IFN-γ are major inducers of PD-L1 transcription in tumor cells (129).

Importantly, PD-L1 also exhibits non-immune functions that contribute to oncogenesis. Its cytoplasmic domain interacts with downstream signaling molecules that promote proliferation, EMT and resistance to apoptosis. Elevated PD-L1 levels have been associated with increased resistance to radiotherapy and chemotherapy, highlighting its multifaceted role in therapeutic evasion (122). Moreover, PD-L1 expression on exosomes released by tumor cells can suppress immune surveillance at distant sites, enabling systemic immunosuppression and facilitating metastatic dissemination. Thus, PD-L1 serves not only as a predictive biomarker for immune checkpoint blockade but also as a dynamic effector of tumor progression and immune modulation in OSCC (129).

Over the years, in order to identify patients who could benefit from this treatment, researchers have investigated possible cut-offs of PD-L1 expression showing a significant association between their levels and response to treatment (128-131). However, such studies revealed different cut-offs of IHC expression in tumor cells and/or in TME (130-132). Considering the different scoring systems used for the quantification of PD-L1 expression, which represents the gold standard in a number of solid tumors, the combined positive score (CPS), that evaluates through a mathematical formula its expression in tumor cells and immune cells (125), currently is recommended for the suitability for immunotherapy of patients with OSCC (125,126).

The PD-1 blockade agents, pembrolizumab (133) and nivolumab (134), have been approved for metastatic OSCC following important clinical trials. The KEYNOTE-048 phase III trial enrolled 882 patients with recurrent or metastatic HNSCC (339 oral primaries) and showed that pembrolizumab, either alone or combined with platinum and 5-fluorouracil, achieved a statistically significant survival advantage compared with standard therapy (EXTREME regimen) in PD-L1-positive cases (CPS ≥1) (133). After a median follow-up of 45 months, OS improved with pembrolizumab monotherapy in the PD-L1 CPS ≥20 [hazard ratio (HR) 0.61; 95% CI 0.46-0.81)] and CPS ≥1 (HR 0.74; 95% CI 0.61-0.89) groups and was noninferior in the overall population (HR 0.81; 95% CI 0.68-0.97). OS also increased with pembrolizumab-chemotherapy in all CPS categories (≥20, ≥1 and total), with HRs ranging from 0.62-0.71 (135). The CheckMate-141 trial enrolled 361 patients (175 oral cavity primaries) with recurrent HNSCC progressing within six months of platinum-based therapy. Nivolumab markedly improved OS compared to standard agents (HR 0.68; 95% CI 0.54-0.86) after 24.4 months of follow-up (136).

Expanding immunotherapy to earlier disease stages is a current focus. In a multicenter phase II study, pembrolizumab administered preoperatively to 36 patients with locally advanced, non-HPV-related HNSCC improve DFS to 17% at 12 months in high-risk patients. An association between partial tumor response (pTR1-2) and PD-L1 expression was observed, suggesting that pTR could predict favorable outcomes even with a single neoadjuvant dose (137). Another phase II study using two preoperative pembrolizumab cycles reported 1-year DFS rates of 97% (95% CI 71-90%) in intermediate-risk and 66% (95% CI 55-84%) in high-risk groups (138). These findings were confirmed in a subsequent phase III trial (139) and align with results from other solid malignancies (140,141).

Ongoing studies continue to explore combination strategies. Cemiplimab, another PD-1 inhibitor, is being tested with platinum-doublet chemotherapy and cetuximab in locally advanced HNSCC (142). A phase II trial evaluated camrelizumab (anti-PD-1) with nab-paclitaxel and cisplatin as neoadjuvant therapy in 48 patients (16 oral cavity primaries), achieving an overall response rate (ORR) of 89.6% (95% CI 80.9-98.2) (143).

Following the success of PD-1/PD-L1 blockade, novel immune checkpoints are under investigation to further enhance anti-tumor immunity (144,145). Tumor necrosis factor receptor superfamily member 4 (OX40), a T-cell costimulatory receptor, has shown promise: in a phase Ib study, Duhen et al (146) reported increased CD4⁺ and CD8⁺ lymphocyte infiltration in peripheral blood and post-surgical specimens following neoadjuvant anti-OX40 (MEDI6469) therapy in locally advanced HNSCC.

Other checkpoint inhibitors under study include antibodies targeting CTLA-4, TIM-3, lymphocyte-activation gene 3 (LAG-3) and IDO, although clinical evidence in OSCC remains limited. CTLA-4 (CD152), expressed on CD4⁺ and CD8⁺ T cells, regulates immune homeostasis by inhibiting T-cell activation; its blockade may enhance recognition and elimination of neoplastic cells (146-148). Combination regimens such as nivolumab plus ipilimumab (anti-CTLA-4) and nivolumab plus relatlimab (anti-LAG-3) are currently being evaluated (NCT04080804) (149).

In the context of precision oncology, further modulation of IC pathways, including LAG-3 and TIM-3 blockade, has demonstrated in murine models the capacity to restore CD8⁺ T-cell activity and suppress tumor growth (146,149). Additional immunostimulatory approaches include cytokine-based therapies using IL-2, IL-1β, IFN-γ and TNF-α, which enhance immune activation and inhibit tumor-induced T-cell apoptosis (150). Monoclonal antibodies targeting Tregs (151) or TAMs (152) also hold potential. These findings are supported by the antitumor activity of IL-2 and IFN-α in enhancing T and NK cell responses within the TME, as demonstrated in other solid tumors (153,154).

Genetic susceptibility may influence immune modulation. For instance, Chien et al (155) found that the IL-23R rs10889677 polymorphism (C allele) conferred a 1.5-fold higher risk of oral cancer and associated with LNM. Furthermore, decreased IFN-γ protein and mRNA expression were shown to negatively are associated with OSCC development (156).

Emerging approaches include adoptive cellular immunotherapy, involving large-scale engineering of tumor-reactive immune cells and therapeutic vaccines. A phase II study evaluating cemiplimab with ISA-101b (targeting HPV16 E6/E7 oncoproteins) in 198 patients demonstrated an ORR of 25.3%, compared with 22.9% in controls. Patients with CPS ≥20 treated with the combination achieved markedly higher response rates and overall survival (157).

The growing understanding of the TME in OSCC has transformed the perception of this malignancy from a purely epithelial disorder to a complex ecosystem of cellular and molecular interactions: CAFs, immune cells and ECM components form a dynamic and interdependent network that governs tumor progression, therapeutic resistance and patient outcomes.

Emerging evidence highlights that immune phenotypes ranging from 'immune-inflamed' to 'immune-desert' microenvironments strongly determine prognosis and response to immunotherapy. Patients with high densities of CD8⁺ T cells, TLSs and NK cells, tend to experience improved OS and favorable responses to PD-1/PD-L1 blockade. Conversely, tumors enriched with M2-polarized macrophages, NETs and CAF-driven fibrosis exhibit immune exclusion, metabolic stress and therapy resistance. Prospective studies on composite biomarkers, should validate integrated scores that combine CPS (PD-L1), immune contexture (CD8⁺/FOXP3⁺ ratios, TLS density), stromal load (α-SMA⁺ CAF index) and systemic inflammation (NLR) to stratify patients for immunotherapy and combination regimens (primary endpoints: OS/DFS; secondary: On-treatment immune remodeling signatures). It may be also possible to pair PD-1/PD-L1 inhibitors with agents that counter macrophage-mediated radioresistance (such as HB-EGF/NHEJ pathway modulators) and with ECM-softening strategies to improve T-cell trafficking. In CAF-directed stroma modulation research, early-phase trials should test inhibitors of TGF-β/IL-6 signaling, LOX-mediated stiffness, or FAP-targeted agents, alone or layered onto PD-1 blockade, to convert immune-excluded tumors into inflamed responders (biopsy-paired spatial readouts recommended). Furthermore, apply single-cell and spatial transcriptomics/proteomics on longitudinal (pre-/post-therapy) specimens to map niches of immune exclusion, CAF heterogeneity (such as TDO2⁺ subsets) and treatment-induced state shifts that predict durable benefit. In addition, mechanistic and interventional studies on microbiome-immune intervention (such as targeted oral microbiome modulation) should test whether reshaping dysbiosis attenuates N2-skewing and T-cell exhaustion in OSCC.

The summarized data presented in Table I provided an integrated overview of these cellular players, their functional mediators and their prognostic associations within the OSCC microenvironment.

Together, these data position the CAF-myeloid-PD-L1-oral microbiome-axis as a unifying framework for prognosis and therapeutic design in OSCC.

Translating TME biology into practice will require composite diagnostics and mechanism-based combinations that are tested in biomarker-anchored, biopsy-paired trials. Aligning multi-omics, digital pathology and immuno-oncology within these designs offers the most direct path to durable benefit for patients with OSCC. In conclusion, the OSCC microenvironment represents both a challenge and an opportunity for precision oncology. Targeting the interplay between tumor cells, stromal components and immune effectors holds the potential to overcome therapeutic resistance and improve long-term outcomes. Future studies integrating multi-omics, digital pathology and immuno-oncology will be essential to translate these biological insights into actionable clinical strategies.