The human papillomavirus infection is a predominant cause of oropharyngeal SCC, a subset of HNSCC (25). The other smaller subsets of SSC arising from the oral cavity, larynx, nasopharynx, and paranasal sinuses may also test HPV-positive (26). Approximately half of the patients diagnosed with SCC originating from an unknown primary site in the head and neck region exhibit HPV-positivity (27). It is well-established that a strong correlation exists between HPV infection and increased risk of oropharyngeal SCC (28, 29). HPV 16 is the most common type in HNSCC, accounting for 90% of the cases, and HPV 18, 33 and 35 are responsible for the small remaining fraction (30).

Numerous studies have consistently shown that HPV-positivity in oropharyngeal SCC is associated with a significantly lower mortality risk and better prognosis compared to HPV-negative cases (31, 32). This may be attributed to the distinct genetic profiles of HPV-negative and HPV-positive HNSCC, which ultimately influence the initiation and progression of the disease (4).

The presence of HPV-mediated disease can be evaluated via direct methods, e.g., the detection of HPV DNA or E6/E7 mRNA detection by either polymerase chain reaction (PCR) or in situ hybridization (ISH), or via indirect methods, e.g., the identification of p16 expression via immunohistochemistry (IHC) (33). p16 protein, a CDKN2A gene product, inhibits the cell cycle through binding and inhibiting cyclin-dependent kinase- and 6 (CDK4 and CDK6). When the HPV E7 protein binds phosphorylated retinoblastoma (pRb) protein, it disrupts a feedback mechanism that normally limits p16 expression through free pRb, resulting in p16 overexpression (34).

The impact of HPV status on prognosis in oropharyngeal SCC led to the introduction of distinct staging criteria in the 8th edition of the American Joint Committee on Cancer Manual for HPV-positive and HPV-negative oropharyngeal cancer (35). The most accurate technique for evaluating HPV infection remains the detection of mRNA transcripts. However, a combined strategy involving p16 evaluation by IHC and HPV DNA by PCR has shown similar sensitivity and specificity rates (36). Consequently, the National Comprehensive Cancer Network (NCCN) guidelines recommend the use of p16 IHC as a surrogate biomarker for HPV testing in all patients diagnosed with oropharyngeal SCC. This recommendation stems from the strong agreement observed between p16 expression detected by IHC and HPV status determined by HPV E6/E7 mRNA expression (37–39). However, for non-oropharyngeal cancers, routine HPV testing or p16 testing is not recommended owing to the small proportion and lack of consistent evidence in support of prognostic significance (40).

Recently, assays for circulating tumor HPV DNA (ctHPV DNA) have emerged as promising tools for enhancing post-treatment surveillance. Chera et al., found that detecting ctHPV DNA in two consecutive plasma samples during post-treatment monitoring yielded both 100% positive predictive value (PPV) and negative predictive value (NPV) for identifying biopsy-proven recurrence (41). Similarly, another study reported high per test sensitivity (92.5%) and NPV (99.4%) of ctHPV DNA surveillance in oropharyngeal SCC when matched to physician-reported clinical outcome data, highlighting its potential for early disease recurrence detection (42). Tanaka et al., found that ctHPV DNA polymerase chain reaction had similar NPV (89.7% vs 84.0%) and higher PPV (100% vs 50.0%) compared to positron emission tomography and computed tomography (PET-CT). These findings suggest that post-treatment ctHPV DNA testing complements PET-CT and provides valuable insights for managing patients with HPV-related HNSCC post-therapy (43).

Beyond its role as a prognostic biomarker, HPV status also serves as a predictive biomarker. HPV-positive HNSCC has been observed to exhibit heightened responsiveness to ICI treatment compared to HPV-negative cases because of its distinct intrinsic characteristics and elevated expression of programmed cell death-ligand 1 (PD-L1). This has been supported by the findings of a retrospective study wherein high PD-L1 expression in immune cells (≥5%) provided predictive information in HPV-positive oropharyngeal SCC irrespective of the tumor stage (44). Further, the findings from the HNSCC cohort of the multibasket phase I KEYNOTE-012 trial suggested that HPV-positive patients with HNSCC are more responsive to ICIs, wherein an increased overall response rate (ORR) to pembrolizumab was observed in the former group compared to HPV-negative patients (25% vs 14%) (45). However, the outcomes of phase III KEYNOTE-040 could not corroborate these findings. Further studies investigating other anti-PD-1/PD-L1 agents have yielded mixed results, e.g., increased response rates were observed among HPV-positive patients compared to HPV-negative patients when treated with durvalumab (46), while no differences were seen with atezolizumab (47). These observed disparities could potentially be explained by additional concurrent factors beyond PD-L1 expression and immune infiltration such as smoking (to be discussed further) and tumor mutational burden (TMB), which are believed to influence response to ICIs in HNSCC, although their relevance differs between HPV-positive and HPV-negative disease (48). Therefore, HPV-positivity alone may not serve as a reliable predictive biomarker for response to ICI and should be interpreted in conjunction with other accompanying clinical and molecular biomarkers (49).

Recent studies have revealed the presence of EBV-encoded small RNAs (EBERs) in tumor cells from various subsets of HNSCCs, including those originating in the tonsils, hypopharynx, larynx, and tongue (50, 51). The detection of EBER-RNA within carcinoma cells has been linked to a poorer prognosis in these patients (50). Carpen et al., reported that EBER expression to be present in stromal lymphocytes adjacent to the tumor and correlated this finding with HPV-positivity, highlighting a complex interplay between EBV and HPV in these tumors. Patients with EBER-positive but HPV-negative OPSCC were found to have significantly poorer overall survival and disease-free survival than those with HPV-positive OPSCC and slightly worse prognosis compared with the patients with EBER-negative and HPV-negative OPSCC. These findings indicate that EBV infection status may have a prognostic impact among patients with HPV-negative OPSCC (51).

The testing methods for the detection of EBV in tumor include ISH for EBV-encoded RNA and IHC staining for latent membrane protein 1 (LMP1). Comparatively, ISH for EBV-encoded small RNA is generally considered a more sensitive testing method for carcinoma than LMP1 IHC testing. Sensitivity and specificity of ISH have been reported to be 94% and 69%, respectively; whereas for IHC, a sensitivity of 44% and specificity of 93% were found (52). However, the traditional chromogenic ISH for EBERs lacks robustness and sensitivity due to the presence of diffuse, nonspecific signals in the extracellular region. Newer detection methods like RNAscope, PrimeFlow, and ViewRNA use specific probes and amplifiers to enhance signal amplification and analytical sensitivity. However, these advanced techniques are often expensive and time-consuming (53).

The correlation between elevated levels of tumor-infiltrating lymphocytes (TILs), particularly CD3⁺ and CD8⁺ cells and improved patient outcomes has already been affirmed by various studies (54, 55). Immunoscore, assessing TILs (CD3/CD45RO, CD3/CD8 or CD8/CD45RO) both in the core and invasive margin of tumors, has emerged as a pivotal prognostic marker. It categorizes patients from immunoscore 0 (I0, low densities in both regions) to immunoscore 4 (I4, high densities in both regions) (56). Two real-world investigations assessing the prognostic significance of immunoscore, have shown that patients with an intermediate to high immunoscore (score of 1) generally experience better prognoses compared to those with a low immunoscore (score of 0) (57, 58). Similarly, Wang et al., conducted a study affirming the prognostic significance of immunoscore, reflecting the density of total CD3⁺, CD8⁺, and memory T-cells (CD45RO⁺) within either the tumor or stroma. The findings revealed that patients with a high immunoscore (score of 3-6) exhibited prolonged OS and DFS compared to those with a low immunoscore (score of 0-2) (59). Recently, studies have highlighted that an immunoscore combining the density of CD8⁺, FoxP3⁺, and CD68⁺ cells within stromal and intra-tumoral compartments serves as a significant prognostic marker in HNSCC (60–62). These findings advocate that immunoscore could complement the existing array of biomarkers for predicting patient outcomes in HNSCC.

There is mounting evidence that tumor-infiltrating B cells (TIL-Bs) have a crucial role in tumor regulation; correlating with patient prognosis, immune cell infiltration, and response to immunotherapy (63). Ruffin et al., evaluated patients with HNSCC based on enrichment of B cell signatures in TME and observed that not only higher numbers of TIL-Bs, but also the specific phenotype and localization of TIL-Bs in the TME contribute to overall survival. Additionally, TIL-Bs in HPV-positive and HPV-negative HNSCC exhibit distinct transcriptional profiles. HPV-positive HNSCC patients showed higher levels of TIL-Bs in germinal centers and tertiary lymphoid structures (TLS) with germinal centers, which were found to be associated with better clinical outcomes (64).

In the ever-evolving realm of HNSCC research, continual discoveries of additional prognostic markers under investigation underscore the distinctive nature of the mutational landscape, which is dominated by tumor suppressor genes and activating oncogene mutations. Notably, TP53, CDKN2A, and NOTCH1 emerge as frequently mutated genes in HNSCC, often correlating with poorer OS in patients with HNSCC (65). Elevated levels of telomerase expression, detected in 75-100% of patients, and high telomerase reverse transcriptase (TERT) levels (> 93.8 copies) in cancer tissues have been linked to worse survival compared to cases with low TERT levels (< 93.8 copies; HR, 3.30; 95% CI: 1.98-5.52; p < 0.0001) (66). Tano et al., evaluated the relationship between the expression ratio of mRNAs for the antiapoptotic protein Bcl-2 and the proapoptotic protein Bax (the Bcl-2/Bax ratio) and clinical outcomes in patients with HNSCC; wherein DFS of patients with Bcl-2/Bax ratios ≥ 1.2 was noted to be longer than that of patients with Bcl-2/Bax ratios < 1.2 (67). Eukaryotic translation factor 4E is a protein involved in protein synthesis and its overexpression has been correlated with an increased risk of disease progression and poor prognosis and loco-regional recurrence of HNSCC and hence, it can be an independent prognostic predictor in terms of recurrence and survival in patients with HNSCC (68).

Recently, studies have documented the significance of CXC chemokine receptors (CXCR2, CXCR4, CCR2 and CCR7) in HNSCC tumor progression and organ-specific metastasis. This indicates the role of chemokines and their cognate receptors in HNSCC prognosis (69). Also, the overexpression of cytokeratins, one of the major components of intracellular filament networks found in different tissues, has also been observed to be related to tumor progression and prognosis (70). Melanoma-associated gene (MAGE), with an expression rate of 85.5% to 90% in HNSCC tissue, predicts poor oncologic outcomes like poorer 5-year survival rate, in patients with SCC of the larynx and hypopharynx (71).

Furthermore, the TP53 gene alterations are commonly reported in carcinomas including HNSCC and researchers have found that patients with TP53 wild-type status tend to exhibit higher survival rates compared to those harboring TP53 mutations. Consequently, this genetic marker holds significant potential for both diagnostic and prognostic purposes in HNSCC patients. Siemert et al., determined the potential of pre-therapeutic plasma vascular endothelial growth factor (VEGF) levels as biomarkers for outcomes in HNSCC and observed that patients with plasma VEGF < 26 ng/L had superior nodal, local and loco-regional control leading to significant prolonged progression-free survival (PFS) and event-free survival (72). A study conducted by Stoiber et al., revealed that high expression of β-catenin, a pivotal mediator of white blood cell activation, is linked to better OS, and thus was postulated as a potential marker of better outcomes in patients with HNSCC (73).

Podoplanin is a mucoprotein specifically expressed in the lymphatic endothelium, that influences regulating cell proliferation and lymphatic vascular development. Its overexpression correlates with the histological grade of the tumor and lymph node metastasis, offering prognostic insights (74). Thereafter, a metanalysis involving nine studies delineated that the expression of fibroblast growth factor receptor 1 (FGFR-1) also predicted poor OS (HR, 1.97; 95% CI: 1.49-2.61; p < 0.001) in patients with HNSCC (75).

Circulating tumor DNA is released from apoptotic and necrotic tumor cells. Its tumor-specific characteristics make it a valuable component in liquid biopsy, enabling the detection of minimal residual disease and positioning it as a potential biomarker for HNSCC prognosis and disease monitoring (76). Kogo et al. demonstrated that HNSCC patients who tested negative for ctDNA during follow-up after initial curative treatment of HNSCC had a significantly better prognosis compared to those who tested positive for ctDNA (76). Additionally, another study reported that post-treatment ctDNA was detected in 71.4% of patients with recurrent disease, whereas it was not observed in any of the patients without recurrence or metastasis, indicating its potential to prognosticate early detection of recurrence (77). Additional research is still needed to fully understand the utility of the above-mentioned biomarkers for determining the prognosis of HNSCC.

Copy number alterations (CNAs) have been frequently observed in HNSCCs, for instance, deletion of the tumor suppressor gene CDKN2A in 9p21.3, which is the most prevalent gene‐specific CNA, has been found in 30.9% of HNSCCs. Other common cancer gene‐specific CNAs in HNSCC are amplifications of CCND1 and FGF3/4 in 11q13.3 (23.2%), PIC3CA in 3q26.32 (15.7%), TP63 in 3q28 (16.1%), EGFR in 7p11.2 (10.4%), and MYC in 8q24.21 (9.3%). CNAs have been reported as an independent prognosticate of worse prognosis in patients with recurrent/metastatic HNSCC receiving immunotherapies (123).

Ayers et al., demonstrated that an IFN-γ–related gene profile obtained from baseline tumor tissue was predictive of the best overall response and PFS in cohorts of patients with melanoma, HNSCC, and gastric cancer who were treated with pembrolizumab (124). Further, a retrospective study identified five most significantly overexpressed immune-related genes in HNSCC: TNFRSF9/4-1BB (77%), IDO1 (75%), TNFSF4/OX40L (74%) and TNFRSF18/GITR (74%), and FOXP3 (62%). The majority of the analyzed tumors overexpressed actionable immunity genes, including PD-1/PD-L1, TIGIT, OX40/OX40L and/or cytotoxic T-lymphocyte associated protein-4 (CTLA-4). The overexpression of these genes was found to be directly related to clinical outcomes. Haddad et al., found 18-gene T-cell-inflamed gene expression profile (GEP) to be associated with PD-L1, indicating that GEP may prove helpful in predicting response to anti-PD-1/PD-L1 agents in HNSCC (90).

The functional significance of distinct subsets of head and neck cancer-associated fibroblasts (HNCAFs) in modulating the immunoregulatory environment of HNSCC has been highlighted by a few researchers. Specifically, HNCAF-0/3 was identified as predictive of nivolumab response, whereas HNCAF-1 was associated with immunosuppression (125).

Epidermal growth factor receptor is overexpressed in more than 90% of HNSCC cases and its overexpression corresponds to tumor growth and progression, resistance to therapy and poor outcomes for patients with HNSCC (126). Cetuximab, a monoclonal antibody targeting the extracellular part of EGFR, has shown improvement in survival rates when combined with radiation compared with radiation alone among patients with HNSCC.

Recent research indicates that measuring total exosome levels, the ratio of tumor-derived exosomes (TEX) and total exosome and the specific characteristics of exosomes originating from TEX or T-cells could effectively differentiate between patients with HNSCC who positively responded to tumor treatment and those who did not (127).

A systematic review assessed 40 clinical studies to elucidate the role of indoleamine 2,3‐dioxygenase (IDO), a key intracellular enzyme involved in immune regulation, in HNSCC and found that its immunohistochemical expression was correlated with worse survival. Moreover, IDO expression was found to be correlated with positive PD‐L1 and HPV status and therefore it was proposed as an emerging predictive biomarker for response to PD‐L1 therapy (128).

Patients with HNSCC have shown lower expression of co-stimulatory receptors, resulting in T-cell anergy/apoptosis and diminished immune response (129). GSK609, an agonist of inducible T-cell co-stimulatory, has been investigated in combination with pembrolizumab, in patients with HNSCC who have failed platinum-based therapy. The drug combination has shown encouraging outcomes (ORR, 26%; 95% CI: 12.9-44.4); disease control rate, (68%; 95% CI: 49.5-82.6) suggesting that this combination therapy can effectively activate T-cells and overcome the immune suppression observed in HNSCC (130).

B7-H3 (also known as CD276) is a member of the B7 ligand family, overexpressed in HNSCC, and provides an alternative immune checkpoint to therapeutically target alone or in combination with PD-1-targeted therapies. Recently, in a phase I/II trial, enoblituzumab (anti-B7-H3 antibody) combined with pembrolizumab in post-platinum PD-L1-naive recurrent/metastatic HNSCC showed 33% of patients to experience ORR (131).

The exhausted T-cells are characterized by the expression of inhibitory molecules, such as PD-1, CTLA-4, T-cell immunoglobulin, and immunoreceptor tyrosine-based inhibitory motif: lymphocyte activation gene 3 (LAG-3), T-cell immunoglobulin and mucin domain (TIM-3) and others, that negatively regulate their response, as well as reduced secretion of cytokines and cytolytic molecules (132). The expression of TIM-3, and LAG-3 have been observed to be overexpressed in HNSCC and are reported to be associated with poor prognosis (133). A phase II neoadjuvant study to determine the safety and tolerability of nivolumab alone or in combination with anti-LAG-3 (relatlimab) or anti-CTLA-4 (ipilimumab) in resectable HNSCC is under clinical trial (134).

The defective transforming growth factor (TGF) signaling in HNSCC results in cancer invasion and metastasis and helps tumor cells escape from immune surveillance. A notable increase in TGF levels has been observed in patients with HNSCC patients (135), suggesting its potential as both a marker for malignant transformation and a target for preventive therapy.

The efficacy of enoblituzumab in combination with retifanlimab (anti-PD-1) or tebotelimab (engineered against PD-1) for the treatment of patients with recurrent/metastatic HNSCC is currently under investigation.

Apart from identifying a reliable biomarker associated with response to a given ICI, the forthcoming wave of novel immunotherapeutic approaches combining different ICI with co-stimulatory agonists, therapeutic vaccines and cytotoxic drugs, has recently led to the concept of targeting ‘hot’ tumors. Hot tumors are characterized by an overall activated immune microenvironment with high infiltration of immune cells and are more likely to respond to immunomodulatory therapies. Using gene expression profiles of 421 HNSCC, Foy and colleagues developed and validated a score to identify immunologically active tumors. The 27-gene expression-based ‘HOT’ score was observed to be correlated with PD-L1 and IDO expression CD8 infiltrate and activation of the IFN-γ pathway. Additionally, among patients with HNSCC treated with ICIs, the HOT score was linked with an improved OS and PFS, indicating that this simple and robust approach can be utilized to identify real-world patients with immunologically active HNSCC who may benefit from ICIs (136). Additionally, leveraging specialized tissue models that closely mimic the native TME, encompassing immune signatures and intratumor heterogeneity, enhances the precision of testing targeted therapies. One such innovative tissue model with transformative potential in cancer research and treatment is the organoid model. Organoids are 3D, multicellular in vitro tissue constructs that mimic specific in vivo organs and can be utilized to explore physiological and pathological aspects associated with that organ. In the case of HNSCC, organoids offer the potential to uncover novel subsets of immune markers for targeted therapy, given that drug screenings have demonstrated sensitivity to targeted drugs not traditionally employed in HNSCC treatment (11), thereby paving the way for expanded therapeutic avenues.