After the screening, 31 studies were included in the final analysis. Different definitions of ‘younger’ and ‘older’ patients groups were observed ( Figure 1A ). Twenty-four studies defined younger patients as those aged<40 years (n=12) or <45 years (n=12), five studies defined younger patients as those aged<50 years, and one study defined younger patients as those aged<70 years. Studies defined the older group as follows: patients >45 years old; patients aged 40 (n=5), 50 (n=4), and 60 years (n=4); and patients aged >70 years (n=2).

The sample sizes also varied ( Figure 1B ). Nine studies did not include an older group, and the number of younger and older patients ranged from 4 to 386 and 4 to 294, respectively, with a relatively similar median. Among the 31 studies, 6 included The Cancer Genome Atlas (TCGA) data ( Figure 1C ), 24 matched TNM stages between patient groups ( Figure 1D ), and 10 used healthy or blank controls for comparison with the study groups ( Figure 1E ). Finally, among 20 comparative studies, 2 studies showed that younger groups were positive for tumor aggressiveness compared to older groups, 9 showed negative results, and 9 showed non-significant results. In addition, among 11 correlation studies, 7 presented disease risk and 4 prognosis correlations ( Figure 1F ).

Among the 31 studies included in the final analysis, 20 reported comparisons between patients with early- and late- onset HNSCC, and 11 reported potential correlations between specific factors and early onset. The possible TME based on these studies are shown in Figure 2 . The two positive results showed that young patients presented a less harmful TME than older patients. These include lower expression of enhancer of zeste homolog 2 (EZH₂) (23) and lower neutrophil-to-lymphocyte ratio (NLR) (24) in younger patients, reflecting lower invasion and metastasis of the tumor behavior in this age group. However, the remaining nine studies showed opposing results, indicating that younger tumors have a more aggressive TME. Increased extracellular matrix (ECM) remodeling, epithelial-to-mesenchymal transition (EMT), and a decreased quantity of CD8+T cells lead to a more immunosuppressive TME (20). The enrichment of inflammatory microbes, increased levels of reactive oxygen species (25) and MMP-9 expression (26) result into inflammatory TME. In addition, the increased WGD (27), EGFR level (26, 28, 29), nuclear polymorphism and mitotic index (30), and P16 methylation leads to abnormal cell proliferation and invasion (31).

Correlation studies have identified early-onset factors for both disease and prognosis. The following factors have been associated with HNSCC risk in young patients: major histocompatibility complex class I-related chain A (MICA) A5.1 homozygous genotype (32), germline variants in FANCG, CDKN2A and TPP (33), drive genes ATXN1 and CDC42EP1 (34), rs6942067 GG genotype (35) in non-HPV and non-smokers, TP53 variation (36), HPV16 positive (37) and GSTM1 null genotype (38) were reported to be associated with HNSCC risk in young patients. In addition, copy number variation (39), ERCC1 (40), galectins-7 (41) and estrogen hormone receptor expression (42) are associated with the prognosis to young HNSCC patients.

The evolving landscape of early-onset HNSCC demands a better understanding of host-tumor interactions in the TME to improve the effectiveness of immunotherapy. However, the impact of complex tumor-infiltrating immune cell profiles on responses to immune checkpoint inhibitors is not fully understood; limited evidence has yielded contradictory findings. Although only direct analysis of the TME was reported, the studies showed that early-onset HNSCC is distinct from average-onset terms of tumor behavior and prognosis, indicating different therapeutic demands.

According to Révész et al. (23) and Zhang et al. ‘s (24) studies, TME of younger patients presented more gently compared to older ones. EZH₂ expression and NLR were lower in the young groups than in the older group. Having a well-defined oncogenic role in cancer initiation, progression, metastasis, metabolism, and drug resistance, and in the modulation of antitumor immunity in various cancers, EZH₂ has been defined as an effective marker of the tumor aggressiveness and tumorigenic potential and plays essential roles in driving cancer cell immunoediting and as an immune escape regulator (43). Inhibition of EZH₂ could suppress oral squamous cell carcinoma (OSCC) progression via modulate EZH₂/Wnt/beta-catenin pathway, both in vitro and in vivo (44). Clinically, the use of EZH₂ inhibitors in combination with IO represents a compelling strategy to remodel the TME, potentially overcoming immune evasion and enhancing therapeutic outcomes in breast cancer (16), mesothelioma (45), non-Hodgkin lymphoma (46) and other cancers.

Similarly, a low NLR is associated with reduced cancer invasion and metastasis of the cancer. When this is observed in young patients, a better prognosis is usually expected. In patients with muscle-invasive bladder cancer (MIBC), this is associated with increased CD3⁺ T cells and B cell infiltration, lead to improved response and long-term outcomes (47).

Genetic factors were reported to be associated with the occurrence of HNSCC or prognosis. The membrane protein MICA generate response to various cellular stresses such as infection and oncogenic transformation, its mechanism of A5.1 allele association with disease risk of young OSCC remained unclear. Current evidence shows it’s essential in etiology and immune response of cancer, both positive or negative. Li et al. reported MICA expression positively related to the CD8⁺ T cell infiltration in hepatocellular carcinoma (48), however Wu et al. found the releasing of MICA progressed tumor immune escape (49).

Cury et al. reported germline variants CDKN2A and RECQL4 are associated with young HNSCC (33). CDKN2A variant closely associated with weak expression of immune-inflammatory pathways in the TME, potentially leading to reduced immune cell activity and weakened immune responses. RECQL4 variant may play a critical role in tumorigenesis and progression by regulating immune responses. Also, these two variants played roles in immune infiltration and the interactions of chemokines and their receptors under immune cells in melanoma (50), and existed in pancreatic ductal adenocarcinoma but are not currently actionable targeted (51). Similarly, ATXN1 and CDC42EP1 have also been reported as driver genes in HNSCC; however, their relevance to young patients cannot be conclusively determined (34). On SNP level, rs6942067 GG genotype is significantly higher in young and in HPV negative non-smoking HNSCC than in other HNSCC, which associated with DCBLD1 expression (35, 52). In addition, TP53 variation, HPV16 positive and GSTM1 null genotype were also mentioned. Different studies also showed CNV, ERCC1 expression, galectin-7 and Estrogen related to prognosis in young HNSCC. Overall, distinct genotypes may either promote or inhibit tumor progression by influencing various components within the TME. In young HNSCC patients, the high expression of specific genotypes has been associated with immune microenvironment activation. In addition, CD8⁺ T cells are crucial effector cells in anti-tumor immune responses the decreased amount in TME suggested a heightened state of immune suppression. The diminished immune surveillance allows tumor cells to evade detection and destruction by the immune system, thereby promoting tumor growth and metastasis.

Furthermore, the marked acceleration in cell proliferation of younger patient indicated that tumor cells may exhibit dysregulation in the signaling pathways controlling proliferation. Such abnormal proliferation is often associated with imbalances in cell cycle regulation and disruptions in apoptotic mechanisms, further driving rapid tumor growth and dissemination (53, 54).

The crosstalk among cells in the TME plays essential roles in the development and progression of HNSCC. Although there is no direct evidence showing exact cellular crosstalk, it can be inferred that the early-onset group exhibits altered communication between immune cells (lymphocytes and CD8+ T cells) and tumor cells compared to the older group. Additionally, the different crosstalk induced by changes in cellular molecules (e.g., EZH₂ protein) also plays an indirect role.

The included studies were highly heterogenous. The definition of younger groups had varying age ranges depending on different guidelines. The samples sizes of the young and older groups were relatively similar; however, large cohort data is lacking. Some studies involved supplementary data from TCGA, which increased the number of patients available for comparison; however, considering the characteristics of different ethnicities and nationalities, bias may exist. Most studies matched TNM stages between groups, while less involved healthy patients as blank controls. These variations contribute to the unreliability of the studies, complicating efforts to combine and compare results. To reduce heterogeneity and achieve more definitive conclusions, subgroup analyses and meta-analyses are recommended when sufficient data points are available.

Personalized immunotherapy emphasizes treatment to each patient’s unique tumor profile and immune response, significantly enhancing effectiveness and reducing side effects compared to standard approaches (55–57). To develop personalized immunotherapy for specific cohorts, understanding the TME is essential (58, 59). Although current evidence is limited in scale and fragmented, it can be concluded that immunotherapy for early-onset HNSCC patients should focus on targeting EGFR, inhibiting ECM remodeling and EMT, and paying attention to high P16 methylation and specific coexisting microbial infections.

Additionally, defined genetic risk factors, including variations in MICA A5.1, FANCG, CDKN2A, and TPP, as well as alterations in ATXN1, CDC42EP1, and TP53, offer potential therapeutic and preventive pharmacological targets. Furthermore, CNV, ERCC1, galectins-7, and ER expression are promising candidate predictive biomarkers. It can be learned that multi-dimensional approaches including blood test, immunohistochemistry, PCR, RNA-sequencing, whole exosome sequencing, microbiota and whole genome have been used from comparative and correlation studies.

For young patients, the focus should be on strategies that restore CD8+ T cell function, regulate associated genetic factors, and target immune escape mechanisms (such as MICA shedding). EZH₂ inhibitors have shown potential in remodeling the TME and enhancing immune responses. Therefore, in young patients, if the immune suppression in the TME is low, the combination of EZH₂ inhibitors with immunotherapy therapies may be particularly effective. For older patients, this combination therapy may also be effective, but the treatment regimen should be adjusted based on the specific TME characteristics.

Overall, based on the current findings, larger-scale clinical studies are necessary in the future to verify these results. Further investigations into TME cellular characteristics, such as EZH₂, MICA expression, and NLR, would also provide valuable evidence. Additionally, developing new therapeutic targets and predictive biomarkers based on these findings and translating them into real clinical practice is anticipated.