Sarcopenia-driven gene model as a clinically actionable prognostic signature for head and neck squamous cell carcinoma

Introduction

Head and neck squamous cell carcinoma (HNSCC) ranks among the most prevalent malignant tumors worldwide, demonstrating particularly high epidemiological burden. Annual diagnoses exceed 600,000 cases globally, and HNSCC is associated with a persistently high mortality rate (1,2). Although considerable progress has been made in therapeutic approaches, including surgery, radiotherapy, chemotherapy, and targeted therapy in recent years, HNSCC maintains a sobering 50–60% 5-year survival, with metastatic cases faring markedly worse (3,4). Therefore, identifying factors influencing the prognosis of HNSCC patients, elucidating the mechanisms behind these crucial factors, and attempting to address related issues are of critical importance for improving survival outcomes and optimizing therapeutic strategies.

In recent years, sarcopenia has received increasing attention as a syndrome closely associated with the prognosis of cancer patients (5-7). Clinically, it presents with declining skeletal muscle quantity and quality, which is a frequent comorbidity in patients with malignant tumors. It is potentially linked to tumor-related systemic inflammation, metabolic dysregulation, and treatment-related side effects (8). Clinically, sarcopenia exerts detrimental effects across multiple dimensions: it increases treatment-related complications (surgical morbidity, chemotoxicity), impairs functional status and quality of life, and reduces tolerance to anticancer therapies (9,10). Most importantly, sarcopenia independently predicts poorer overall survival (OS) in HNSCC patients, with increased mortality risk vs. non-sarcopenic patients (11,12). These findings not only highlight the urgent need to develop effective strategies for managing sarcopenia in HNSCC patients but also emphasize the potential of exploring innovative research directions to identify effective interventions for sarcopenia management in HNSCC patients.

Radiogenomics, an emerging research approach, synthesizes radiomics characteristics with molecular profiling to elucidate the associations between tumor phenotypes and molecular characteristics, offering novel insights into precision diagnosis and treatment of cancer (13,14). This innovative methodology enables a comprehensive analysis of tumor heterogeneity and biological behavior by linking radiological phenotypes with underlying genomic alterations. However, despite its growing application in various cancers, no previous studies have employed radiogenomics to investigate sarcopenia in HNSCC patients, particularly in elucidating its prognostic significance and molecular basis.

Our research endeavors to clarify the prognostic implications of sarcopenia in HNSCC patients via radiogenomic integration, concurrently deciphering its associated genomic signatures. Thus, this study aimed to investigate the clinical impact of sarcopenia on prognosis in HNSCC patients and explore its underlying molecular correlations using a radiogenomic approach. Specifically, the objectives were to: (I) assess the association between sarcopenia, as determined by computed tomography (CT) imaging, and OS; (II) characterize differences in genomic, transcriptomic profiles, and immune tumor microenvironment between sarcopenic and non-sarcopenic HNSCC patients; and (III) develop a novel genetic prognostic signature based on sarcopenia-associated gene expression differences and subsequently validate its performance on predicting survival with an independent external cohort. We present this article in accordance with the TRIPOD reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1310/rc).

Methods

Data collection

Clinical and radiological data for HNSCC patients were obtained from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). Gene expression profiles of HNSCC patients from the GSE65858 dataset were downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) and utilized as the validation cohort. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Patients with available CT scan imaging data in the TCGA-head and neck squamous cell carcinoma (HNSC) database were included in this study. TCGA-HNSC data were publicly released with follow-up through February 28, 2017; GSE65858 data were deposited in 2016. The entire research process was presented in Figure 1.

Figure 1 Flowchart of the study. C3, third cervical vertebra; CT, computed tomography; GEO, Gene Expression Omnibus; GO, Gene Ontology; HNSCC, head and neck squamous cell carcinoma; KEGG, Kyoto Encyclopedia of Genes and Genomes; LASSO, least absolute shrinkage and selection operator; TCGA, The Cancer Genome Atlas.

Assessment of sarcopenia

The presence of sarcopenia was determined by measuring the skeletal muscle area at the level of the third cervical vertebra (C3) on baseline CT images from the TCGA-HNSC cohort (Figure 2). The CT images were sourced from the TCGA database and underwent rigorous quality control and curation by The Cancer Imaging Archive (TCIA). This standardized process ensures the technical consistency and analytical suitability of the dataset (15,16). Measurements were quantified using ITK-SNAP software (version 4.0). Two independent radiation oncologists performed manual segmentation of the sternocleidomastoid and paravertebral muscle groups, with a third expert resolving any discrepancies. The optimal cutoff value of the C3 skeletal muscle area for defining sarcopenia was determined using the surv_cutpoint function in R (version 4.3.2).

Figure 2 CT imaging at the C3 level for skeletal muscle area segmentation and sarcopenia assessment. (A) Sagittal CT image of the cervical spine with the C3 labeled. (B) Axial CT image at the level of C3. (C) Axial CT image at the C3 level: muscle areas marked in red represent delineated regions used for sarcopenia assessment. C3, third cervical vertebra; CT, computed tomography.

Survival analysis

The Kaplan-Meier method was utilized to analyze the survival disparities between sarcopenic and non-sarcopenic patients. Meanwhile, the log-rank test was carried out to assess the statistical significance of the differences in survival rates between the two groups. Multivariable Cox regression was used to evaluate the independent prognostic value of age, gender, clinical stage, tumor grade, and sarcopenia on patient survival.

Analysis of genetic and transcriptomic profiles

Differentially expressed genes (DEGs) between the sarcopenia and non-sarcopenia groups were identified using the DESeq2 method. Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) analyses were conducted on these DEGs to explore their potential biological functions and signaling pathways. Subsequently, least absolute shrinkage and selection operator (LASSO) regression was utilized for the DEGs to develop a Genetic Risk Score (GRS). The optimal cutoff for stratifying patients by GRS was calculated using the surv_cutpoint function from the R package survminer. Kaplan-Meier curves evaluated the prognostic power of GRS for OS. The GRS, along with other clinically significant factors, including age, gender, clinical stage, and histological grade, was included in a multivariate Cox proportional hazards model to evaluate the independent impact of the GRS on patient outcomes. Furthermore, we systematically compared immune cell infiltration between sarcopenic and non-sarcopenic cohorts using CIBERSORT.

Validation of the GRS model

The GSE65858 validation cohort differed from the TCGA development cohort in setting (single-center German vs. multi-institutional USA) and technical platform [microarray vs. RNA sequencing (RNA-seq)], requiring cross-platform normalization. While the development cohort selection mandated multimodal data availability, the validation cohort reflected standard gene-expression-based eligibility. Crucially, both cohorts maintained identical outcomes (OS) and utilized the identical four-gene predictor [fibrinogen beta chain (FGB)/kinesin family member 1A (KIF1A)/hepatic leukemia factor (HLF)/yippee-like 1 (YPEL1)], ensuring direct model applicability. In the GSE validation cohort, univariate and multivariate Cox regression analyses were performed based on the key genes identified through LASSO regression. A Cox Risk Score (CRS) model was subsequently constructed using the results from the multivariate Cox regression. Based on the CRS, we calculated time-dependent risk probabilities at 1/3/5 years and evaluated their predictive performance for OS.

Statistical analyses

The baseline characteristics of participants were stratified based on sarcopenia status. Continuous variables were compared using independent t-tests, and categorical variables were analyzed with Chi-squared tests. The prognostic value of sarcopenia for OS was examined through Cox proportional hazards modeling, presenting effect estimates as hazard ratio (HR) and corresponding 95% confidence interval (CI). The analysis was restricted to cases with complete data. The validation cohort (GSE65858) had complete survival and expression data for all signature genes. All statistical analyses were performed in R (v4.3.2) and Sangerbox (http://www.sangerbox.com). No generative artificial intelligence (AI) tools were employed in the preparation of this manuscript or the analysis of the research data. A two-sided P value of less than 0.05 was considered to indicate statistical significance.

Results

Baseline characteristics of patients stratified by sarcopenia status

We retrieved clinical and radiological data for 176 HNSCC patients from the TCGA database. After excluding patients lacking gene expression data (n=9), a total of 167 patients were included in the analysis. The optimal cutoff value for sarcopenia, based on the C3 skeletal muscle area (in cm²) measured from baseline CT images, was determined to be 3,374.639 cm². Consequently, 67 patients were classified into the sarcopenia group, while 100 patients were assigned to the non-sarcopenia group. The baseline characteristics of the two groups were presented in Table 1. The mean age of sarcopenic patients was 61.42 years, compared to 58.82 years in non-sarcopenic individuals. Among non-sarcopenic subjects, males accounted for 95.5%, whereas females represented only 4.5%. In contrast, the sarcopenic group exhibited a significantly higher proportion of females (41.0%) than the non-sarcopenic group (4.5%), with males still constituting the majority (59.0%).

Association between sarcopenia and survival outcomes

Kaplan-Meier survival analysis revealed that the OS of patients in the sarcopenia group was significantly shorter than that of patients in the non-sarcopenia group (Figure 3A; HR =2.36; 95% CI: 1.33–4.19; P=0.003). After comprehensive adjustment for age, gender, clinical stage, and histological grade, sarcopenia retained independent prognostic significance for OS (Figure 3B; HR =1.98; 95% CI: 1.08–3.64; P=0.03). Kaplan-Meier survival analysis stratified by treatment modality (pharmaceutical therapy or radiation therapy) and sarcopenia status revealed significant differences in OS among subgroups, which demonstrates the consistent prognostic value of sarcopenia across different treatment modalities (Figure S1A, log-rank P=0.02). Furthermore, subgroup analyses were further performed based on age, gender, clinical stage, histological grade, and treatment modality. A consistent trend towards poorer OS was observed in association with sarcopenia across all predefined clinical subgroups. Although the association did not reach statistical significance in certain smaller subgroups, sarcopenia achieved consistent prognostic stratification in the majority of subgroups. This effect was particularly evident in larger cohorts, including the overall population, all age strata, male patients, those with stage III–IV tumors, those with grade G2–3 tumors, and those who received radiotherapy (Figure S1B).

Figure 3 Association between sarcopenia and patient survival. (A) Kaplan-Meier plot depicting the association between sarcopenia and OS. (B) Univariate and multivariate Cox proportional hazards models assessing the association of sarcopenia with OS. C3, third cervical vertebra; G, grade; OS, overall survival.

Identification of DEGs and functional pathway association analysis

DEGs analysis using DESeq2 revealed 76 significantly upregulated genes and 90 significantly downregulated genes in the sarcopenia group, with the top 10 upregulated and downregulated genes illustrated in Figure 4A. Association of selected DEGs with OS was shown in Figure S2A. Functional enrichment analysis identified distinct pathway alterations between upregulated and downregulated genes in HNSCC-associated sarcopenia. KEGG analysis of upregulated genes demonstrated significant enrichment in cytokine-cytokine receptor interaction, JAK-STAT signaling, PPAR signaling, complement and coagulation cascades, and extracellular matrix (ECM)-receptor interaction pathways (Figure 4B). These findings highlight the potential involvement of chronic inflammatory responses, metabolic dysregulation, and ECM reorganization in the pathogenesis of muscle wasting. GO analysis of upregulated genes revealed enrichment in biological processes related to secretory granule function, hormone metabolism, platelet alpha granule components, fatty acid binding, and junctional sarcoplasmic reticulum membrane proteins (Figure 4C). These observations suggest concurrent disturbances in secretory mechanisms, lipid homeostasis, and calcium handling that may collectively impair muscle contractility and metabolic function. In contrast, KEGG analysis of downregulated genes showed significant reductions in Rap1 signaling, amyotrophic lateral sclerosis-related pathways, PPAR signaling, NF-κB signaling, and glycosphingolipid biosynthesis (Figure 4D). This pattern indicates compromised cell-matrix interactions, diminished neuroprotective responses, and altered lipid metabolism that may contribute to progressive muscle degeneration. GO analysis of downregulated genes further identified impairments in chromosome organization, meiotic chromosome segregation, cell cycle regulation, nuclear chromosome condensation, and synaptonemal complex assembly (Figure 4E). These findings suggest fundamental defects in genomic maintenance and cellular proliferation that may underlie the observed muscle deterioration in this condition.

Figure 4 Genetic and pathway analyses related to sarcopenia. (A) Identification of DEGs between the sarcopenia and non-sarcopenia groups. (B) KEGG pathway analysis of upregulated genes in sarcopenia. (C) GO analysis of upregulated genes in sarcopenia. (D) KEGG pathway analysis of downregulated genes in sarcopenia. (E) GO analysis of downregulated genes in sarcopenia. DEG, differentially expressed gene; FDR, false discovery rate; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes.

LASSO regression and GRS construction

Four key genes were identified through LASSO regression analysis, including FGB, KIF1A, HLF, and YPEL1 (Figure 5A,5B). Univariate Cox regression analysis demonstrated that FGB, KIF1A, and YPEL1 were significantly associated with OS in patients (Figure S2B). Furthermore, expression analysis of the four genes in tumor and normal cells from HNSCC patients revealed differences in FGB, KIF1A, and HLF (Figure S3). A risk score model was developed based on the regression coefficients of the four genes: GRS = (2.604×10⁻⁵ × FGB) + (7.063×10⁻⁵ × KIF1A) − (1.903×10⁻⁴ × YPEL1) − (2.248×10⁻⁶ × HLF). Application of survival-optimized cutoff values stratified the cohort into distinct high-risk (n=76) and low-risk (n=91) prognostic categories. Time-dependent receiver operating characteristic (ROC) analysis validated the GRS’s prognostic performance, with area under the curve (AUC) values reaching 0.70 for 1-year, 0.73 for 3-year, and 0.67 for 5-year survival prediction (Figure 5C). Survival curves showed markedly worse OS in the high-risk group (Figure 5D; HR =2.86; 95% CI: 1.73–4.73; P<0.001), demonstrating the prognostic utility of the GRS model. After adjusting for age, gender, clinical stage, and histological grade, the GRS remained an independent prognostic factor for OS (Figure 5E,5F, P<0.001). Furthermore, survival and multivariate Cox regression analyses using the GRS integrated with clinical features (including age, gender, clinical stage, and histological grade) demonstrated significant prediction of OS (Figure S4A; HR =5.87; 95% CI: 3.57–9.64; P<0.001). The model demonstrated excellent temporal discrimination, with time-dependent AUC values reaching 0.74 for 1-year, 0.79 for 3-year, and 0.77 for 5-year survival prediction (Figure S4B,S4C).

Figure 5 LASSO regression model development and validation for sarcopenia in HNSCC. (A) LASSO regression coefficient profiles of sarcopenia-related features. Each curve represents a coefficient trajectory against the log(lambda) value. (B) Variable selection via LASSO regression. The vertical lines represent optimal lambda values determined by minimum criteria (left dashed line) and 1-SE criteria (right dashed line). (C) ROC curve analysis for the LASSO regression model. (D) Kaplan-Meier survival curves comparing high-risk and low-risk groups stratified by the LASSO-derived sarcopenia signature. (E) Distribution of GRS and corresponding survival status across the patient cohort, arranged from low to high risk. (F) Nomogram integrating the GRS with clinicopathological factors for individualized survival prediction in HNSCC patients. Points are assigned for each variable based on its prognostic weight, with total points corresponding to survival probability estimates. ****, P<0.0001; –, no statistical significance (P≥0.05). 1-SE, one standard error; AUC, area under the curve; CI, confidence interval; FGB, fibrinogen beta chain; G, grade; GRS, Genetic Risk Score; H, high risk; HLF, hepatic leukemia factor; HNSCC, head and neck squamous cell carcinoma; HR, hazard ratio; KIF1A, kinesin family member 1A; L, low risk; LASSO, least absolute shrinkage and selection operator; ROC, receiver operating characteristic; YPEL1, yippee-like 1.

Validation of the GRS in the GSE validation cohort

In the GSE validation cohort, HLF exhibited independent prognostic value in multivariate Cox regression analysis (HR =0.32; 95% CI: 0.17–0.61; P<0.001). A multivariable Cox proportional hazards model incorporating FGB, KIF1A, HLF, and YPEL1 was constructed to derive the CRS for predicting OS probabilities at 1, 3, and 5 years (Figure 6A). The CRS demonstrated significant predictive ability across all time intervals (Figure 6B-6D; 1-year: HR =1.80, 95% CI: 1.18–2.75, P=0.006; 3-year: HR =1.86, 95% CI: 1.21–2.86, P=0.005; 5-year: HR =1.79, 95% CI: 1.16–2.75, P=0.008).

Figure 6 Validation of a multivariable Cox model derived from the four genes via LASSO regression in the GEO cohort. (A) Combined forest plot of 1-, 3-, and 5-year risk subgroups. (B) One-year survival analysis based on the CRS from the multivariate Cox model. (C) Three-year survival analysis based on the CRS from the multivariate Cox model. (D) Five-year survival analysis based on the CRS from the multivariate Cox model. CI, confidence interval; CRS, Cox Risk Score; GEO, Gene Expression Omnibus; LASSO, least absolute shrinkage and selection operator.

Comparison of immune microenvironment infiltration levels

Sarcopenic patients exhibited a significantly attenuated T cell profile, which is characterized by diminished CD8⁺ cytotoxic T cells and activated CD4⁺ memory T cells (Figure S5)—a pattern that suggests impaired tumor immunosurveillance. Building on these observations of compromised immune function in sarcopenic individuals, we further explored how GRS relates to immune-microenvironment remodeling. Spearman correlation analysis demonstrated that a higher GRS was significantly and negatively correlated with the infiltration levels of M1 macrophages (Figure S6A; Spearman r=−0.25, P=0.001), activated CD4⁺ memory T cells (Figure S6B; Spearman r=−0.19, P=0.01), resting CD4⁺ memory T cells (Figure S6C; Spearman r=−0.26, P<0.001), and CD8⁺ T cells (Figure S6D; Spearman r=−0.28, P<0.001). Collectively, these results suggest that a higher GRS may be associated with an immunosuppressive tumor microenvironment, supporting a logical chain linking “sarcopenia-related genetic characteristics” to “impaired immune status”.

Discussion

Sarcopenia represents a multifactorial condition marked by the gradual deterioration of both muscle quantity and functional capacity (17). Growing evidence has established sarcopenia as an important predictor of clinical outcomes across multiple cancer types, particularly gastrointestinal and thoracic malignancies (18-21). Due to the frequent occurrence of dysphagia, malnutrition, and weight loss in HNSCC patients, the prevalence of sarcopenia is notably high in this population. Furthermore, sarcopenia may exacerbate physical functional decline and treatment-related complications in these patients (22). Given these clinical implications, developing reliable methods for early sarcopenia detection is critical for implementing timely interventions and improving outcomes. While existing studies have well-established the clinical correlations of sarcopenia, its distinct genomic profile remains poorly characterized (23-26). Therefore, this study employed an integrated radiogenomics approach to comprehensively analyze sarcopenia in HNSCC, revealing distinct genomic characteristics of sarcopenic patients and establishing a novel sarcopenia-associated gene expression signature for prognostic prediction in HNSCC.

Our study demonstrated that sarcopenia was significantly associated with reduced OS and served as an independent adverse prognostic factor in patients with HNSCC. Sarcopenia may influence patient prognosis through multiple mechanisms, including systemic inflammation, metabolic dysregulation, and reduced tolerance to treatment (27,28). The release of systemic inflammatory factors [e.g., interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-α)] and metabolic abnormalities may lead to decreased protein synthesis and increased catabolism, thereby accelerating muscle atrophy (28). Additionally, tumor patients often experience insufficient nutritional intake due to dysphagia, appetite loss, and treatment-related side effects, which may further exacerbate muscle wasting and reduce treatment tolerance. Given the significant clinical impact of sarcopenia, we conducted an integrated multi-omics analysis to characterize its molecular underpinnings, including differential gene expression patterns, transcriptomic profiles, and immune microenvironment alterations in sarcopenic HNSCC patients, with the goal of identifying potential biomarkers and therapeutic targets.

Differential gene expression analysis using DESeq2 identified significant upregulation of carboxylesterase 1 (CES1), serpin family A member 3 (SERPINA3), and albumin (ALB), along with downregulation of paired box 1 (PAX1), RAR-related orphan receptor B (RORB), and nitric oxide synthase 2 (NOS2) in sarcopenic patients. Among these genes, CES1, a critical hydrolase, has been shown to promote proliferation, invasion, and migration of HNSCC cells (29,30), and portends worse survival in HNSCC patients (31). In contrast, PAX1, a well-known tumor suppressor gene, has been demonstrated to contribute to cervical cancer progression when its function is lost (32). Through LASSO regression analysis of DEGs, we identified four key genes: FGB, KIF1A, YPEL1, and HLF. Previous studies have shown that FGB expression is associated with tumor angiogenesis and metastasis (33-35); KIF1A influences the proliferation, migration, and invasion capabilities of tumor cells, thereby affecting the malignancy of neuroblastoma cells (36); diminished YPEL1 levels are strongly linked to survival outcomes in glioma patients and potentially mediate anti-neoplastic activity (37); and reduced HLF expression strongly predicts rapid disease progression and metastatic spread in non-small cell lung cancer (38). The findings offer a starting point for further functional characterization of these genes. Survival analysis demonstrated that the GRS model, constructed based on these genes, effectively predicted patient survival outcomes and was an independent prognostic factor for OS.

In the GSE validation cohort, comprehensive analysis identified four key sarcopenia-associated genes (FGB, KIF1A, YPEL1, and HLF), of which three (FGB, KIF1A, and YPEL1) showed significant associations with OS in HNSCC patients. The CRS derived from these four genes demonstrated consistent prognostic power for OS at 1-, 3-, and 5-year follow-ups. These findings establish the risk score as a robust molecular biomarker for HNSCC prognosis across different time frames.

KEGG pathway and GO analyses uncovered substantial pathway enrichment among genes differentially expressed in sarcopenia, particularly in muscle atrophy-related processes. Among the upregulated genes, the JAK-STAT signaling pathway was a representative pathway identified in the KEGG analysis. The JAK-STAT pathway plays a critical role in cell growth, survival, differentiation, and immune responses, and its activation may influence muscle cell proliferation and apoptosis, thereby affecting muscle mass and function. In the GO analysis of upregulated genes, the fatty acid binding pathway emerged as a representative pathway, primarily involving fatty acid-binding proteins. These proteins participate in the transport and metabolism of intracellular fatty acids, thereby influencing lipid metabolism and energy balance (39). For downregulated genes, the amyotrophic lateral sclerosis pathway was highlighted in the KEGG analysis. This pathway is associated with motor neuron function and degeneration, and motor neuron damage can impair muscle innervation, leading to muscle atrophy and functional decline (40). In the GO analysis of downregulated genes, the condensed nuclear chromosome pathway was identified as a representative pathway, which is involved in chromosome condensation and structural organization during cell division (41). In sarcopenia, the decline in cell division and regenerative capacity may be linked to dysregulation of this pathway.

In the comparison of the immune microenvironment, our study identified characteristic alterations in the tumor immune microenvironment of HNSCC patients with sarcopenia, primarily manifested by significantly reduced numbers of CD8⁺ T cells and activated memory CD4⁺ T cells. Compared to the immune features previously reported in general HNSCC patients (42), these immunological changes in sarcopenic patients appear more pronounced, suggesting that muscle tissue loss may exacerbate immunosuppression in the tumor microenvironment through systemic immune modulation (43). Specifically, the concurrent reduction of both effector T cells and memory T cells not only extends beyond the typical tumor-induced immunosuppression, but may also impair anti-tumor immunological memory function (44). These alterations show close correlation with patients’ metabolic abnormalities. These significant findings provide new clinical evidence supporting the “muscle-immune axis” theory (45) and elucidate the immunological mechanisms underlying the poorer prognosis in sarcopenic HNSCC patients.

This investigation represents the inaugural radiogenomic analysis elucidating the prognostic significance of sarcopenia in HNSCC patients, revealing novel imaging-genomic correlations. While our results provide novel insights into sarcopenia’s prognostic role in HNSCC, as well as its multi-omics characteristics, several limitations should be acknowledged. First, due to the lack of height and weight data in the TCGA database, we used the C3 skeletal muscle area as the criterion for defining sarcopenia, which may introduce potential inaccuracies. Second, this study is primarily based on retrospective data; future prospective investigations are necessary to further confirm the clinical value of the GRS model. Third, because the patient cohort was rigorously selected, the sample size for the multi-omics analysis and model construction was relatively small (n=167), which may limit the statistical power and generalizability of our findings. Future studies with larger cohorts are needed to validate our model.

In summary, this study, through a radiogenomics approach, revealed a significant association between sarcopenia and prognosis in HNSCC patients and constructed a GRS model based on key genes. This model demonstrated robust predictive performance in the GSE validation cohort, offering a novel tool for individualized prognostic assessment in HNSCC patients.

Conclusions

Sarcopenia serves as an independent prognostic factor in HNSCC patients, and the GRS model constructed based on the expression of FGB, KIF1A, YPEL1, and HLF genes holds significant clinical value for prognostic assessment in sarcopenic patients with HNSCC.

Acknowledgments

We would like to express our deepest gratitude to all those who have contributed to this research. The authors did not use any generative AI tools in the creation of this work.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1310/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1310/prf

Funding: The work was supported by Sichuan Science and Technology Program (Nos. 2022YFS0217 and 2024ZYD0054), Guizhou Maotai Hospital Research and Talent Development Fund (No. MTyk-2022-03), and the Noncommunicable Chronic Diseases-National Science and Technology Major Project (No. 2023ZD0503004).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1310/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Johnson DE, Burtness B, Leemans CR, et al. Head and neck squamous cell carcinoma. Nat Rev Dis Primers 2020;6:92. [Crossref] [PubMed]
2. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
3. Siu LL, Even C, Mesía R, et al. Safety and Efficacy of Durvalumab With or Without Tremelimumab in Patients With PD-L1-Low/Negative Recurrent or Metastatic HNSCC: The Phase 2 CONDOR Randomized Clinical Trial. JAMA Oncol 2019;5:195-203. [Crossref] [PubMed]
4. Tolstonog G, Simon C. Trends in Surgical Research in Head and Neck Cancer. Curr Treat Options Oncol 2017;18:38. [Crossref] [PubMed]
5. Li S, Wang T, Tong G, et al. Prognostic Impact of Sarcopenia on Clinical Outcomes in Malignancies Treated With Immune Checkpoint Inhibitors: A Systematic Review and Meta-Analysis. Front Oncol 2021;11:726257. [Crossref] [PubMed]
6. Shachar SS, Williams GR, Muss HB, et al. Prognostic value of sarcopenia in adults with solid tumours: A meta-analysis and systematic review. Eur J Cancer 2016;57:58-67. [Crossref] [PubMed]
7. Martin L, Birdsell L, Macdonald N, et al. Cancer cachexia in the age of obesity: skeletal muscle depletion is a powerful prognostic factor, independent of body mass index. J Clin Oncol 2013;31:1539-47. [Crossref] [PubMed]
8. Blum D, Stene GB, Solheim TS, et al. Validation of the Consensus-Definition for Cancer Cachexia and evaluation of a classification model--a study based on data from an international multicentre project (EPCRC-CSA). Ann Oncol 2014;25:1635-42. [Crossref] [PubMed]
9. van Vugt JL, Levolger S, Coelen RJ, et al. The impact of sarcopenia on survival and complications in surgical oncology: A review of the current literature. J Surg Oncol 2015;112:681-2. [Crossref] [PubMed]
10. Kudou K, Saeki H, Nakashima Y, et al. Prognostic Significance of Sarcopenia in Patients with Esophagogastric Junction Cancer or Upper Gastric Cancer. Ann Surg Oncol 2017;24:1804-10. [Crossref] [PubMed]
11. Chargi N, Bril SI, Emmelot-Vonk MH, et al. Sarcopenia is a prognostic factor for overall survival in elderly patients with head-and-neck cancer. Eur Arch Otorhinolaryngol 2019;276:1475-86. [Crossref] [PubMed]
12. Findlay M, Brown C, De Abreu Lourenço R, et al. Sarcopenia and myosteatosis in patients undergoing curative radiotherapy for head and neck cancer: Impact on survival, treatment completion, hospital admission and cost. J Hum Nutr Diet 2020;33:811-21. [Crossref] [PubMed]
13. Gevaert O, Xu J, Hoang CD, et al. Non-small cell lung cancer: identifying prognostic imaging biomarkers by leveraging public gene expression microarray data--methods and preliminary results. Radiology 2012;264:387-96. [Crossref] [PubMed]
14. Kuo MD, Jamshidi N. Behind the numbers: Decoding molecular phenotypes with radiogenomics--guiding principles and technical considerations. Radiology 2014;270:320-5. [Crossref] [PubMed]
15. Clark K, Vendt B, Smith K, et al. The Cancer Imaging Archive (TCIA): maintaining and operating a public information repository. J Digit Imaging 2013;26:1045-57. [Crossref] [PubMed]
16. Prior F, Smith K, Sharma A, et al. The public cancer radiology imaging collections of The Cancer Imaging Archive. Sci Data 2017;4:170124. [Crossref] [PubMed]
17. Escriche-Escuder A, Fuentes-Abolafio IJ, Roldán-Jiménez C, et al. Effects of exercise on muscle mass, strength, and physical performance in older adults with sarcopenia: A systematic review and meta-analysis according to the EWGSOP criteria. Exp Gerontol 2021;151:111420. [Crossref] [PubMed]
18. Fukuda Y, Yamamoto K, Hirao M, et al. Sarcopenia is associated with severe postoperative complications in elderly gastric cancer patients undergoing gastrectomy. Gastric Cancer 2016;19:986-93. [Crossref] [PubMed]
19. van Vledder MG, Levolger S, Ayez N, et al. Body composition and outcome in patients undergoing resection of colorectal liver metastases. Br J Surg 2012;99:550-7. [Crossref] [PubMed]
20. Joglekar S, Asghar A, Mott SL, et al. Sarcopenia is an independent predictor of complications following pancreatectomy for adenocarcinoma. J Surg Oncol 2015;111:771-5. [Crossref] [PubMed]
21. Go SI, Park MJ, Song HN, et al. Sarcopenia and inflammation are independent predictors of survival in male patients newly diagnosed with small cell lung cancer. Support Care Cancer 2016;24:2075-84. [Crossref] [PubMed]
22. Jovanovic N, Chinnery T, Mattonen SA, et al. Sarcopenia in head and neck cancer: A scoping review. PLoS One 2022;17:e0278135. [Crossref] [PubMed]
23. Jimenez-Gutierrez GE, Martínez-Gómez LE, Martínez-Armenta C, et al. Molecular Mechanisms of Inflammation in Sarcopenia: Diagnosis and Therapeutic Update. Cells 2022;11:2359. [Crossref] [PubMed]
24. Vogele D, Otto S, Sollmann N, et al. Sarcopenia - Definition, Radiological Diagnosis, Clinical Significance. Rofo 2023;195:393-405. [Crossref] [PubMed]
25. Ibebunjo C, Chick JM, Kendall T, et al. Genomic and proteomic profiling reveals reduced mitochondrial function and disruption of the neuromuscular junction driving rat sarcopenia. Mol Cell Biol 2013;33:194-212. [Crossref] [PubMed]
26. Deng H, Wang Y, Dai Y, et al. Unraveling the genetic mysteries of sarcopenia: A bioinformatics approach. Technol Health Care 2025;33:1140-53. [Crossref] [PubMed]
27. Baracos VE, Martin L, Korc M, et al. Cancer-associated cachexia. Nat Rev Dis Primers 2018;4:17105. [Crossref] [PubMed]
28. Argilés JM, Busquets S, Stemmler B, et al. Cancer cachexia: understanding the molecular basis. Nat Rev Cancer 2014;14:754-62. [Crossref] [PubMed]
29. Feng C, Mao W, Yuan C, et al. Nicotine-induced CHRNA5 activation modulates CES1 expression, impacting head and neck squamous cell carcinoma recurrence and metastasis via MEK/ERK pathway. Cell Death Dis 2024;15:785. [Crossref] [PubMed]
30. Argilés JM, Busquets S, Stemmler B, et al. Cachexia and sarcopenia: mechanisms and potential targets for intervention. Curr Opin Pharmacol 2015;22:100-6. [Crossref] [PubMed]
31. Jiang C, Liu C, Yao XI, et al. CES1 is associated with cisplatin resistance and poor prognosis of head and neck squamous cell carcinoma. Oncol Res 2024;32:1935-48. [Crossref] [PubMed]
32. Li X, Liu H, Zhou X, et al. PAX1 hypomethylation as a prognostic biomarker for radioresistance of cervical cancer. Clin Epigenetics 2023;15:123. [Crossref] [PubMed]
33. Schnoor B, Morris K, Kottana RK, et al. Fibrinolytic Platelet Decoys Reduce Cancer Metastasis by Dissociating Circulating Tumor Cell Clusters. Adv Healthc Mater 2024;13:e2304374. [Crossref] [PubMed]
34. Liu H, Xiang L, Mei Y. miR-877-5p Inhibits Epithelial Mesenchymal Transformation of Breast Cancer Cells by Targeting FGB. Dis Markers 2022;2022:4882375. [Crossref] [PubMed]
35. Repetto O, Maiero S, Magris R, et al. Quantitative Proteomic Approach Targeted to Fibrinogen β Chain in Tissue Gastric Carcinoma. Int J Mol Sci 2018;19:759. [Crossref] [PubMed]
36. Zhou Q, Yang M, Fu J, et al. KIF1A promotes neuroendocrine differentiation in prostate cancer by regulating the OGT-mediated O-GlcNAcylation. Cell Death Dis 2024;15:796. [Crossref] [PubMed]
37. Li W, Huang W, Wu K, et al. Yippee Like 1 Suppresses Glioma Progression and Serves as a Novel Prognostic Factor. Tohoku J Exp Med 2022;256:141-50. [Crossref] [PubMed]
38. Chen J, Liu A, Lin Z, et al. Downregulation of the circadian rhythm regulator HLF promotes multiple-organ distant metastases in non-small cell lung cancer through PPAR/NF-κb signaling. Cancer Lett 2020;482:56-71. [Crossref] [PubMed]
39. Hotamisligil GS, Bernlohr DA. Metabolic functions of FABPs--mechanisms and therapeutic implications. Nat Rev Endocrinol 2015;11:592-605. [Crossref] [PubMed]
40. Peters OM, Ghasemi M, Brown RH Jr. Emerging mechanisms of molecular pathology in ALS. J Clin Invest 2015;125:1767-79. [Crossref] [PubMed]
41. Sutani T, Sakata T, Nakato R, et al. Condensin targets and reduces unwound DNA structures associated with transcription in mitotic chromosome condensation. Nat Commun 2015;6:7815. [Crossref] [PubMed]
42. Zhang Y, Li L, Zheng W, et al. CD8(+) T-cell exhaustion in the tumor microenvironment of head and neck squamous cell carcinoma determines poor prognosis. Ann Transl Med 2022;10:273. [Crossref] [PubMed]
43. Deng Y, Zhao L, Huang X, et al. Contribution of skeletal muscle to cancer immunotherapy: A focus on muscle function, inflammation, and microbiota. Nutrition 2023;105:111829. [Crossref] [PubMed]
44. Reiser J, Banerjee A. Effector, Memory, and Dysfunctional CD8(+) T Cell Fates in the Antitumor Immune Response. J Immunol Res 2016;2016:8941260. [Crossref] [PubMed]
45. Afzali AM, Müntefering T, Wiendl H, et al. Skeletal muscle cells actively shape (auto)immune responses. Autoimmun Rev 2018;17:518-29. [Crossref] [PubMed]