Skip to main content
NCBI home page
Journal of Clinical Laboratory Analysis logoLink to Journal of Clinical Laboratory Analysis
. 2022 Sep 30;36(11):e24722. doi: 10.1002/jcla.24722

Exploration of a predictive model based on genes associated with fatty acid metabolism and clinical treatment for head and neck squamous cell carcinoma

Zhechen Yuan 1,2, Juntao Huang 1,2, Bing Mei Teh 3,4,5, Shiyu Hu 1,2, Yi Hu 1,2, Yi Shen 1,2,✉
PMCID: PMC9701875  PMID: 36181275

Abstract

Background

Head and neck squamous cell carcinoma (HNSCC) is one of the most prevalent malignant tumors of the head and neck and presents high risks of recurrence and poor prognosis postoperatively. The aim of this study was to establish a predictive model based on fatty acid metabolism (FAM) genes to forecast the prognosis of HNSCC patients and the subsequent treatment strategies.

Methods

We accessed the TCGA and GEO databases for HNSCC genes and clinical data. The FAM risk score model was created and validated using a combination of univariate Cox analysis and least absolute shrinkage and selection operator (LASSO) regression analysis. Combining risk scores and clinical characteristics, a nomogram was established and assessed. Subsequently, the function, gene mutation, immune difference, and chemotherapeutic drug sensitivity of the groups with high‐ and low‐risk scores were analyzed. Consequently, the mode's validity was evaluated comprehensively by combining single gene analysis.

Results

The FAM risk score model for predicting HNSCC prognosis had certain validity. Patients in the high‐ and low‐risk groups had genetic mutations, and the prognosis was the poorest for the high‐risk groups with high genetic mutations. The patients with low‐risk scores were suitable for immunotherapy since they had increased infiltration of immune cells. In contrast, the patients in the other groups were more suitable for chemotherapy.

Conclusion

The results of this study demonstrated that the FAM risk score model may predict the prognosis of HSNCC and has a certain therapeutic guidance value.

Keywords: chemotherapy, fatty acid metabolism, head and neck squamous cell carcinoma, immunotherapy, prognosis

Head and neck squamous cell carcinoma (HNSCC) is one of the most prevalent malignant tumors of the head and neck, with high risk of recurrence and a poor prognosis after surgery. It is reported that fatty acid metabolism‐related genes (FAMGs) have potential mechanistic and prognostic roles in HNSCC. However, there has been no investigation on the relation between FAMGs and HNSCC. In this study, we constructed a risk score model related to fatty acid metabolism, which was evaluated to have independent prognostic predictive value and could provide guidance for the clinical treatment of HNSCC.

graphic file with name JCLA-36-e24722-g001.jpg

1. INTRODUCTION

As the most common pathological type of head and neck tumor, head and neck squamous cell carcinoma (HNSCC) has a high incidence. 1 , 2 In the literature, the global incidence of HNSCC is high worldwide, accounting for 89,000 newly diagnosed cases and 450,000 deaths in 2018. 3 , 4 HNSCC usually occurs in the nasopharynx, sinuses, oropharynx, hypopharynx, larynx, and oral cavity, 5 and its etiologies include smoking, alcohol consumption, and genetic and viral infections. 6 , 7 At present, the combination of surgery, radiotherapy, and chemotherapy is a routine clinical treatment strategy. 8 , 9 Unfortunately, the patients' prognosis is unsatisfied. The postoperative recurrence rate of patients is 40%–60%, 9 and the five‐year survival rate is only approximately 50%. 10 Moreover, few studies have reported effective indicators to predict the prognosis of HNSCC worldwide. Therefore, it is essential to find reliable strategies to predict the prognosis of HNSCC.

Due to the fast growth and absence of blood arteries, tumor cells usually exist in the acidic, hypoxic, and dystrophic tumor microenvironment. 11 , 12 Consequently, tumor cells exhibit notable variations in protein, fatty acid, and glucose metabolism from normal cells. 13 Tumor cells are revealed to be able to more readily utilize the higher energy provided by fatty acid metabolism (FAM). 14 , 15 Genes in the fatty acid metabolic pathway are known as fatty acid metabolism‐associated genes (FAMGs), whose reprogramming is an important guarantee for the continuous growth, proliferation, and migration of cancer cells. 16 , 17 In particular, FAMGs have potential mechanistic and prognostic roles in HNSCC. 18 Visweswaran et al. proposed that targeting lipid metabolism may serve as a novel approach for tumor therapy. 19 Constructing a prognostic risk model is a feasible solution for evaluating tumor prognostic indicators. Currently, Du et al. 15 have reported a new risk model based on FAM to assess the prognosis of HNSCC. However, the function of FAMGs in HNSCC remains unknown, and systematic studies still need to be established to investigate the role of FAMGs in the prognosis of HNSCC.

In this study, we constructed a predictive risk model for HSNCC based on 20 FAMGs screened in the HNSCC cohort of the TCGA database. We also investigated the relationship between this prognostic risk model and the immune response. The aim of this study was to evaluate the value of the FAM prognostic risk scoring model for the diagnosis and treatment of HNSCC in clinical settings.

2. METHODS AND MATERIALS

2.1. Data downloading and processing

A total of 548 RNA sequencing data profiles, including 504 HNSCC tumor samples and 44 normal tissue samples, were obtained as fragments per kilobase million from the TCGA–HNSC cohort in the TCGA database. Meanwhile, their clinical characteristics were downloaded, including age, sex, tumor grade, tumor stage, prognostic information, and other relevant information. Similarly, we utilized genetic information of the HNSCC patients collected in the Gene Expression Omnibus (GEO) database as a test set (GSE41613, including 94 samples).

2.2. Construction and validation of the prognostic model

According to the previous studies, 15 , 20 we obtained 309 FAMGs to construct this analysis (Appendix S1). The differentially expressed fatty acid metabolism‐related genes (DEFAMGs) between normal and HNSCC samples were identified according to the criteria of |logFC| >0.585 and false discovery rate <0.05. The prognostic DEFAMGs were determined using univariate Cox regression analysis (p < 0.05), and gene mutations in the tumor samples were also analyzed. A prognostic risk score model was created with the assistance of least absolute shrinkage and selection operator (LASSO) regression analysis, aiming to predict HNSCC patient prognosis. Tenfold cross‐validation was performed with over 1000 cycles of analysis. The risk score model was calculated using the following formula: Risk score = ∑ coefficient of genes * expression values of genes.

All subjects were divided into high‐ and low‐risk groups according to the median risk scores after each sample's risk score was calculated with the formula. To confirm the value of the above risk score model to predict patient prognosis, the log‐rank test and Kaplan–Meier survival curves were used to compare the overall survival (OS) of the patients in both groups. Subsequently, we drew the receiver operating characteristic (ROC) curve and determined the area under the curve (AUC) values to assess the efficacy of the patient prognostic model. In addition, univariate and multivariate Cox regression analyses were carried out to estimate whether the risk prediction model was an independent predictive factor when compared to the clinical features (e.g., age, sex, grade, and stage).

2.3. Establishment of the nomogram

The nomogram was constructed using the data from the TCGA‐HNSCC cohort, including age, tumor stage, and risk score. Furthermore, both calibration curves and ROC curves were drawn to examine the nomogram's validity and its prognostic value, respectively. In addition, we performed univariate and multivariate Cox regression analyses to judge whether the nomogram can be used as an independent indicator for predicting the prognosis of HNSCC.

2.4. GSVA

To further compare FAM between the high‐ and low‐risk groups, gene set variation analysis (GSVA) based on the gene profiles was conducted. GSVA is a nonparametric and unsupervised approach for calculating gene enrichment, which can be used to evaluate a specific biological process. 21 Statistics were judged significant when p < 0.05.

2.5. Correlation between the risk score and immune factors

FAM processes have been reported to be associated with immune cells and tumor progression. 22 According to CIBERSORT, which is used to assess immune cell infiltration, 23 we investigated the relationship between the predictive model and immune infiltration in HNSCC using the Wilcoxon signed‐rank test.

2.6. GO and KEGG

The differentially expressed genes (DEGs) in both the high‐ and low‐risk groups were retrieved by the R package “limma”. Moreover, the “clusterProfiler” R package was used to perform Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. KEGG is capable of identifying the main biological properties and functional pathways of DEGs.

2.7. Protein–protein interaction network

We analyzed DEGs through the STRING online database to generate protein–protein interaction (PPI) networks. The PPI network was further processed via Cytoscape software (version 3.7.2). Additionally, we used CytoHubba, a plug‐in of Cytoscape, to screen out the hub genes among the DEGs. Based on the level of hub gene expression, the samples were categorized as upregulated or downregulated. Finally, we conducted a survival analysis and compared immune cell infiltration.

2.8. Sensitivity analysis of chemotherapy drugs

The pRRophetic software package was used to predict the half‐maximal inhibitory concentration (IC50) of chemotherapeutic drugs in each sample, including cisplatin, docetaxel, gemcitabine, and paclitaxel, which are routinely utilized for the treatment of HNSCC. The differences in IC50 between the two groups were compared to evaluate the clinical efficacy of chemotherapy for the treatment of HNSCC.

3. RESULTS

3.1. Establishment and verification of the risk score model

The flow chart of this study is presented in Figure 1. According to the expression of FAMGs, we obtained 129 DEFAMGs from the gene expression matrix that were defined as differential genes. Specifically, the top 50 genes highly expressed in normal and tumor samples were selected. Heatmaps (Figure 2A) and volcano plots (Figure 2B) were drawn to demonstrate DEFAMGs between the normal and tumor samples.

FIGURE 1.

FIGURE 1

Open in a new tab

The study design and flow chart

FIGURE 2.

FIGURE 2

Open in a new tab

Construction of the prognostic model. (A, B) Heatmap and volcano plot of differentially expressed genes between HNSCC samples and normal samples; (C) Forest plot that shows 22 prognostic genes; (D) Frequency of mutations in genes associated with FAM in HSNCC patients; (E) Co‐occurrence relationship of FAM mutant genes; (F) LASSO coefficient diagram for prognostic genes; (G) Cross‐validation error curve for modeling gene identification

Based on the uni‐Cox regression analysis, 22 FAMGs were shown to be substantially linked with patient prognosis (p < 0.05) (Figure 2C). The frequencies of 22 prognostic FAMGs somatic mutation profiles were summarized in Figure 2D, achieving a total of 10.39% in 510 HNSCC samples. In addition, the correlations of these 22 FAMGs were analyzed as well, and the results were shown in Figure 2E.

Similarly, with the assistance of LASSO regression analysis, 20 FAMGs were selected to establish a predictive risk score model (Figure 2F,G). The risk score for each HNSCC patient was calculated as follows: Risk score = −(0.267068370405719 × ACAA1) + (0.163238015569013 × PRKAA2) + (0.082678299650745 × OSTC) + (0.100868138563843 × HSPH1) − (0.117924286651454 × PTGDS) + (0.157079934784576 × PHYH) + (0.0439834205998577 × PTGES3) − (0.435964503575487 × CYP8B1) − (0.694856005960078 × THRSP) − (0.0272691081848022 × ACOX3) − (2.69032042245729 × ACSBG2) + (0.1350935514509 × LDHA) + (0.0861616365857884 × HADHB) − (0.0512350179850526 × ACSM3) + (0.0443599211083752 × SMS) + (0.188524223062724 × ELOVL6) + (0.363348040751783 × ACADL) − (0.263268335672849 × ACACB) + (0.00801375207470822 × HSP90AA1) + (0.0587043322637337 × SLC25A17).

Subsequently, as shown in Figure 3A,B, patients were divided into low‐ and high‐risk groups based on the median risk score, which indicated that patients can be distinguished clearly by this prognostic model. Referring to the K–M survival analysis, patients in the high‐risk group had worse OS than those in the low‐risk group (Figure 3C), which was also validated by an external test cohort (Figure 3D). Furthermore, while comparing the difference in progression‐free survival (PFS) between the two risk groups, the low‐risk group was considered to have longer PFS (Figure 3E).

FIGURE 3.

FIGURE 3

Open in a new tab

Predictive value of risk scoring models for HNSCC. (A) Principal component analysis of genes involved in FAMGs; (B) Principal component analysis of genes involved in model genes; (C) Overall survival of high‐ and low‐risk groups in the TCGA cohorts; (D) Overall survival of high‐ and low‐risk groups in the GEO cohorts; (E) Comparison of progression‐free survival between patients in the high‐ and low‐risk groups in the TCGA cohorts; (F) AUC values at 1, 3, and 5 years; (G) ROC curves of clinical features and risk scores; (H) Univariate Cox analysis for independent prognostic analysis of the risk score model; (I) Multivariate Cox analysis for independent prognostic analysis of the risk score model

As shown in Figure 3F, the one–year, three–year, and five–year AUC values on the ROC curves of the risk model were 0.689, 0.712, and 0.719, respectively, revealing much better predictive performance than other features (Figure 3G).clinical.

3.2. Establishment of the nomogram for OS

The independent prognostic analyses of the risk model mainly included multiple clinical factors, including age, sex, tumor grade, and clinical stage. Combining uni‐ and multi‐Cox regression analyses, age, clinical stage, and risk score were considered three independent prognostic factors (Figure 3H,I). Based on the above results, a nomogram combining age, clinical stage, and risk score was constructed (Figure 4A). For one–year, three–year, and five–year calibration curves (Figure 4B), it is clear that nomograms are beneficial for predicting HNSCC prognosis with high accuracy. The Cox analysis revealed that the nomogram may be considered an independent feature for prognostic prediction (Figure 4C,D).

FIGURE 4.

FIGURE 4

Open in a new tab

Establishment and verification of the nomogram. (A) Nomogram (B) Calibration curve for the nomogram; (C) Univariate Cox analysis for the nomogram; (D) Multivariate Cox analysis for the nomogram

3.3. Functional analysis of DEGs

To explore differential biological functions between the high‐ and low‐risk score groups, GSVA enrichment analysis was conducted, showing that the majority of metabolic pathways (e.g., FAM) were enriched in the high‐risk score group (Figure 5A).

FIGURE 5.

FIGURE 5

Open in a new tab

Functional analysis of DEGs. (A) Heatmap is showing the GSVA of DEGs; (B) Circle plot for GO analysis of DEGs; (C) Bar plot for GO analysis of DEGs; (D) Circle plot for KEGG analysis of DEGs; (E) Bar plot for KEGG analysis of DEGs

Moreover, we performed GO and KEGG analyses for further functional analysis of the DEGs. The DEGs were considerably enriched within the biological process category in the GO circle graph, indicating their importance (Figure 5B). The GO bar plot showed that the DEGs were highly enriched in antigen binding among the molecular functions and immunoglobulin complexes among the cellular components. In terms of the biological processes, DEGs were mainly enriched in the regulation of B‐cell activation, B‐cell receptor signaling pathway, and complement activation (Figure 5C, Appendix S1). KEGG enrichment analysis demonstrated that these genes mainly participated in cytokine–cytokine receptor interactions, Staphylococcus aureus infection, the IL‐17 signaling pathway, and viral protein interactions with cytokines and cytokine receptors (Figure 5D,E, Appendix S1).

3.4. Mutation differences between high‐ and low‐risk groups

The difference in the top 15 mutated genes was compared and visualized in waterfall plots (Figure 6A,B). As the topmost mutant genes, we assessed the differences in risk score between the TP53 wild‐type and mutation‐type, which suggested that the TP53 mutation‐type has a higher risk (Figure 6C). Survival analysis revealed that low tumor mutation burden (TMB) was associated with a more favorable outcome than high TMB. Comprehensive analysis of TMB and risk score factors showed that patients in the low‐risk group with low TMB had a favorable prognosis compared to the other categories, including high TMB in the high‐risk group, high TMB in the low‐risk group, and low TMB in the high‐risk group (Figure 6D).

FIGURE 6.

FIGURE 6

Open in a new tab

Differences in mutations between the high‐risk and low‐risk groups. (A, B) Differences in gene mutations between the high‐risk and low‐risk groups; (C) Comparison of risk scores for different TP53 mutation types; (D) Survival probability analysis combining TMB and risk score

3.5. Relationship between immune factors and the risk score model

The boxplot (Figure 7A) based on CIBERSORT revealed that the low‐risk group had more infiltrating immune cells (e.g., CD8 T cells); nevertheless, resting natural killer cells, M0 and M1 macrophages, and activated mast cells were more enriched in the high‐risk group. Comparing the differences in immune functions, the low‐risk group appeared to have a more activated immune function, including checkpoint, cytolytic activity, HLA, inflammation promotion, T‐cell coinhibition, and T‐cell costimulation (Figure 7B). Therefore, considering these results of immune‐related analysis, patients in the low‐risk group may be more sensitive to immunotherapy.

FIGURE 7.

FIGURE 7

Open in a new tab

Differences in immunization between the high‐risk and low‐risk groups. (A) Immune cell infiltration in the high‐risk and low‐risk groups. (B) Comparison of immune function between the high‐risk and low‐risk groups

3.6. Assessment of chemotherapeutic effects

The IC50 value was used to predict the chemotherapeutic effects of cisplatin, docetaxel, gemcitabine, and paclitaxel. As indicated by the Spearman correlation analysis (Figure 8A–F), the risk score was negatively correlated with the IC50 values of docetaxel, gemcitabine, and paclitaxel, suggesting that patients in the high‐risk group may exhibit a better chemotherapeutic response to these chemotherapy agents. However, for cisplatin, there were no significant differences in IC50 between the two risk groups (Figure 8G,H).

FIGURE 8.

FIGURE 8

Open in a new tab

Comparison of the efficacy of four chemotherapeutic agents. (A, B) Sensitivity of high‐ and low‐risk groups to docetaxel; (C, D) Sensitivity of high‐ and low‐risk groups to gemcitabine; (E, F) Sensitivity of high‐ and low‐risk groups to paclitaxel; (G, H) Sensitivity of high‐ and low‐risk groups to cisplatin

3.7. PPI network and single gene analysis

A PPI network was established from the expression profiles of DEGs by the STRING online database (Figure 9A) and was visualized by Cytoscape software (Figure 9B). As presented in Figure 9A, genes associated with increased risk are shown in red markers, while the green markers represent genes that were enriched in the low‐risk groups. Assessed by cytoHubba, the top 10 hub genes were obtained, ranking by degree method as LCE3D, SPRR3, SPRR2E, SPRR2G, LCE3E, FLG, SPRR2B, LCE2B, KPRP, and LCE6A.

FIGURE 9.

FIGURE 9

Open in a new tab

PPI network and single gene analysis. (A) PPI network; (B) Hub gene network processed by cytoHubaa; (C) Survival probability analysis for SPRR3; (D) Effect of SPRR3 on immune cells

Survival analysis of the top five hub genes indicated that the expression of SPRR3 was notably associated with the prognosis of HNSCC patients (Figure 9C). Furthermore, the relationship between SPRR3 expression and immune cell infiltration was compared between the low‐ and high‐expression groups (Figure 9D). The numbers of naive B cells, plasma cells, regulatory T cells (Tregs), and neutrophils were markedly infiltrated in tumors with high SPRR3 expression; nevertheless, macrophage M2 was notably higher in tumors with low SPRR3 expression.

4. DISCUSSION

HNSCC is a common malignant and aggressive cancer of the head and neck with a high incidence. 1 , 2 , 3 , 4 , 24 Its atypical early clinical symptoms and insidious disease progression have increased morbidity and mortality over the years. 25 , 26 Cancer cells and immune cells are prone to metabolic reprogramming under the regulation of various factors in the tumor microenvironment (TME), leading to abnormal activation of oncogenic signaling pathways. FAM is one of the key metabolic pathways in HNSCC reprogramming. However, studies demonstrating the relationship between FAMGs and HNSCC are still limited so far.

In this study, we established a risk score model for HNSCC prediction by analyzing the differential expression of FAMGs in both tumor and normal samples of the TCGA‐HNSC cohort. By analyzing the ROC curve, it is evident that this model is superior to the other clinical factors for survival rate prediction of HNSCC patients. The nomogram analyzed by combining this model with clinical features is highly stable and accurate, contributing to higher predictive potential.

Furthermore, we screened 20 genes as prognostic markers following DEFAMGs analysis. HPSH1 has been reported to be an early detected gene that can be used as a biomarker for HNSCC identification. 27 Similarly, the mRNA expression levels of LDHA isozymes were significantly greater in HNSCC compared to the controls. 28 Martin et al. 29 reported that ELOVL6 was involved in fatty acid elongation, and its low expression could improve the prognosis of HNSCC. Moreover, Thomas and colleagues developed a new risk signature model with genes such as PRKAA2 to predict HNSCC prognosis. 30 In fact, PRKAA2 is a related gene in the mTOR signaling pathway associated with chemotherapy resistance. 31 In the literature, PRKAA2 has been shown to play a key role in the PKA‐AMPK pathway, and its downregulation under certain drugs may prevent the occurrence of oral cancer. 32 In this study, we found the levels of gene mutations in both high‐ and low‐risk groups varied. Gene mutations may affect the risk score results. For instance, TP53 gene mutation will improve the risk score. The number of gene mutations present in the tumor is usually expressed by the TMB. 33 , 34 , 35 Consequently, a combined analysis of TMB and risk score indicated that the prognosis of HNSCC patients with low TMB and low risk was relatively good, while high TMB and high risk significantly reduced the survival probability of the patients.

It has been reported that the interaction between the surrounding tissue matrix and immune cells in the TME may be responsible for the high recurrence and metastasis rates of HNSCC. Therefore, analyzing immune infiltration in patients with high‐ and low‐risk scores is beneficial to judge the efficacy of immunotherapy on HNSCC. Our results showed that patients with low‐risk score had more infiltration of immune cells, such as CD8 T cells and Treg cells, than patients with low‐risk score. By inhibiting immune checkpoints such as PD‐1, CD8 T cells enhance cellular immune activity and improve sensitivity to immunotherapy. Treg cells, as inhibitory immune cells, were found to be elevated, which also demonstrated that patients with low‐risk scores were more suitable for immunotherapy. Additionally, we found that low‐risk groups had a more active immune function, which further demonstrated that immunotherapy may benefit patients with low‐risk scores. Moreover, we examined differences in responses to four common chemotherapy drugs in high‐risk groups and low‐risk groups. The results showed that docetaxel and gemcitabine exhibited higher IC50 values in the low‐risk groups. IC50 is the drug concentration that induces apoptosis in 50% of tumor cells, which can reflect the degree of drug resistance of tumor cells. 36 , 37 , 38 This result suggests that chemotherapeutic drug therapy may be more suitable for patients with high‐risk scores.

According to our analysis, SPRR3 is an essential gene for HNSCC since it is associated with HNSCC prognosis. Previous studies on oral squamous cell carcinoma (OSCC) have shown that SPRR3 expression decreases with increasing OSCC malignancy. 39 For patients with esophageal squamous cells, SPRR3 has also been reported to be downregulated. 40 Additionally, our results demonstrated more immune cells infiltrating in the groups with higher SPRR3 expression, including CD8 T cells and Treg cells. Luo et al. 41 found that SPRR3 may act as a radiosensitizer to influence tumor sensitivity to chemoradiotherapy. Overall, these results may provide potential guidance for the treatment of HNSCC.

Generally, this prognostic scoring model based on FAMGs can be used to predict the prognosis of HNSCC. Furthermore, this model can predict tumor sensitivity to chemotherapeutic drugs and feasibility of immunotherapy for HNSCC. Consequently, our study provides an accurate and reliable predictive model for HNSCC prognosis and sensitivity to chemotherapy and immunotherapy. These findings may offer personalized treatment strategies and increase the efficacy of clinical therapy for HNSCC, which encourages more personalized cancer therapies in the future.

However, this study still has some limitations. First, although the risk prediction model has a high success rate, a lack of clinical data can affect the predictive value of this model to some extent. Second, our study is bioinformatics‐based and lacks biological trials to further support the conclusions. Nonetheless, the data analyzed in our study were from external cohorts, and thus the accuracy and reliability of our results were reliable.

5. CONCLUSION

In summary, a FAM prognostic risk score model was successfully established. This model can not only predict the survival of HNSCC patients but also assess the sensitivity and effectiveness of patients to immunotherapy and chemotherapy. This provides a certain guiding value for clinical practice and individualized treatment for HNSCC.

AUTHOR CONTRIBUTIONS

All persons designated as the authors have participated sufficiently in the work to take public responsibility for the content of the manuscript. All the authors ensure that they all gave substantial contributions.

FUNDING INFORMATION

This study was supported by the National Natural Science Foundation of China (No.81670920), Zhejiang Provincial Medical and Health Science Research Foundation (No.2020RC107, No.2020KY274, and No.2022KY1086), Ningbo Natural Science Foundation (No.2018A610363), and Ningbo Public Science Research Foundation (No.2021S170).

Supporting information

Appendix S1

Click here for additional data file. (1.8MB, zip)

Yuan Z, Huang J, Teh BM, Hu S, Hu Y, Shen Y. Exploration of a predictive model based on genes associated with fatty acid metabolism and clinical treatment for head and neck squamous cell carcinoma. J Clin Lab Anal. 2022;36:e24722. doi: 10.1002/jcla.24722

Zhechen Yuan and Juntao Huang contributed equally to this work.

DATA AVAILABILITY STATEMENT

The data used to support the findings of this study comes from the public database and are available from the corresponding author upon request.

REFERENCES

  • 1. Huang J, Xu Z, Teh BM, et al. Construction of a necroptosis‐related lncRNA signature to predict the prognosis and immune microenvironment of head and neck squamous cell carcinoma. J Clin Lab Anal. 2022;36:e24480. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Chow LQM. Head and neck cancer. N Engl J Med. 2020;382:60‐72. [DOI] [PubMed] [Google Scholar]
  • 3. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68:394‐424. [DOI] [PubMed] [Google Scholar]
  • 4. Ferlay J, Colombet M, Soerjomataram I, et al. Estimating the global cancer incidence and mortality in 2018: GLOBOCAN sources and methods. Int J Cancer. 2019;144:1941‐1953. [DOI] [PubMed] [Google Scholar]
  • 5. Rasmussen JH, Lelkaitis G, Hkansson K, et al. Intratumor heterogeneity of PD‐L1 expression in head and neck squamous cell carcinoma. Br J Cancer. 2019;120:1003‐1006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Zhong Q, Fang J, Huang Z, et al. A response prediction model for taxane, cisplatin, and 5‐fluorouracil chemotherapy in hypopharyngeal carcinoma. Sci Rep. 2018;8:12675. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Gu X, Wang L, Boldrup L, et al. AP001056.1, A prognosis‐related enhancer RNA in squamous cell carcinoma of the head and neck. Cancers (Basel). 2019;11:347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Huang J, Xu Z, Yuan Z, et al. Identification of cuproptosis‐related subtypes and characterization of the tumor microenvironment landscape in head and neck squamous cell carcinoma. J Clin Lab Anal. 2022;36:e24638. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Jung K, Narwal M, Min SY, Keam B, Kang H. Squamous cell carcinoma of head and neck: what internists should know. Korean J Intern Med. 2020;35:1031‐1044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Zhang XM, Song LJ, Shen J, et al. Prognostic and predictive values of immune infiltrate in patients with head and neck squamous cell carcinoma. Hum Pathol. 2018;82:104‐112. [DOI] [PubMed] [Google Scholar]
  • 11. Beloribi‐Djefaflia S, Vasseur S, Guillaumond F. Lipid metabolic reprogramming in cancer cells. Oncogenesis. 2016;5:e189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Ruocco MR, Avagliano A, Granato G, et al. Metabolic flexibility in melanoma: a potential therapeutic target. Semin Cancer Biol. 2019;59:187‐207. [DOI] [PubMed] [Google Scholar]
  • 13. Yu T, Wang Y, Fan Y, et al. CircRNAs in cancer metabolism: a review. J Hematol Oncol. 2019;12:90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Li Z, Zhang H. Reprogramming of glucose, fatty acid and amino acid metabolism for cancer progression. Cell Mol Life Sci. 2016;73:377‐392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Du P, Chai Y, Zong S, et al. Identification of a prognostic model based on fatty acid metabolism‐related genes of head and neck squamous cell carcinoma. Front Genet. 2022;13:888764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Corbet C, Feron O. Emerging roles of lipid metabolism in cancer progression. Curr Opin Clin Nutr Metab Care. 2017;20:254‐260. [DOI] [PubMed] [Google Scholar]
  • 17. Cao Y. Adipocyte and lipid metabolism in cancer drug resistance. J Clin Invest. 2019;129:3006‐3017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Cheng C, Geng F, Cheng X, Guo D. Lipid metabolism reprogramming and its potential targets in cancer. Cancer Commun (Lond). 2018;38:27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Visweswaran M, Arfuso F, Warrier S, Dharmarajan A. Aberrant lipid metabolism as an emerging therapeutic strategy to target cancer stem cells. Stem Cells. 2020;38:6‐14. [DOI] [PubMed] [Google Scholar]
  • 20. Huang D, Tang E, Zhang T, Xu G. Characteristics of fatty acid metabolism in lung adenocarcinoma to guide clinical treatment. Front Immunol. 2022;13:916284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA‐seq data. BMC Bioinformatics. 2013;14:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Biswas SK. Metabolic reprogramming of immune cells in cancer progression. Immunity. 2015;43:435‐449. [DOI] [PubMed] [Google Scholar]
  • 23. Newman AM, Liu CL, Green MR, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12:453‐457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin. 2019;69:7‐34. [DOI] [PubMed] [Google Scholar]
  • 25. Chatfield‐Reed K, Gui S, O'Neill WQ, Teknos TN, Pan Q. HPV33+ HNSCC is associated with poor prognosis and has unique genomic and immunologic landscapes. Oral Oncol. 2020;100:104488. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Li M, Li X, Zhang Y, et al. Micropeptide MIAC inhibits HNSCC progression by interacting with aquaporin 2. J Am Chem Soc. 2020;142:6708‐6716. [DOI] [PubMed] [Google Scholar]
  • 27. Babu N, Mohan S, Nanjappa V, et al. Identification of potential biomarkers of head and neck squamous cell carcinoma using iTRAQ based quantitative proteomic approach. Data Brief. 2018;19:1124‐1130. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Mohtasham N, Ayatollahi H, Saghravanian N, et al. Evaluation of tissue and serum expression levels of lactate dehydrogenase isoenzymes in patients with head and neck squamous cell carcinoma. Anticancer Agents Med Chem. 2019;19:2072‐2078. [DOI] [PubMed] [Google Scholar]
  • 29. Prusinkiewicz MA, Gameiro SF, Ghasemi F, et al. Survival‐associated metabolic genes in human papillomavirus‐positive head and neck cancers. Cancers (Basel). 2020;12:253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Zhou H, He Y, Li L, Wu C, Hu G. Identification novel prognostic signatures for head and neck squamous cell carcinoma based on ceRNA network construction and immune infiltration analysis. Int J Med Sci. 2021;18:1297‐1311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Xu CZ, Shi RJ, Chen D, et al. Potential biomarkers for paclitaxel sensitivity in hypopharynx cancer cell. Int J Clin Exp Pathol. 2013;6:2745‐2756. [PMC free article] [PubMed] [Google Scholar]
  • 32. Knobloch TJ, Ryan NM, Bruschweiler‐Li L, et al. Metabolic regulation of glycolysis and AMP activated protein kinase pathways during black raspberry‐mediated oral cancer chemoprevention. Metabolites. 2019;9:140. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Zhang L, Li B, Peng Y, et al. The prognostic value of TMB and the relationship between TMB and immune infiltration in head and neck squamous cell carcinoma: a gene expression‐based study. Oral Oncol. 2020;110:104943. [DOI] [PubMed] [Google Scholar]
  • 34. Merino DM, McShane LM, Fabrizio D, et al. Establishing guidelines to harmonize tumor mutational burden (TMB): in silico assessment of variation in TMB quantification across diagnostic platforms: phase I of the friends of cancer research TMB harmonization project. J Immunother Cancer. 2020;8:e000147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Liu L, Bai X, Wang J, et al. Combination of TMB and CNA stratifies prognostic and predictive responses to immunotherapy across metastatic cancer. Clin Cancer Res. 2019;25:7413‐7423. [DOI] [PubMed] [Google Scholar]
  • 36. Joo M, Park A, Kim K, et al. A deep learning model for cell growth inhibition IC50 prediction and its application for gastric cancer patients. Int J Mol Sci. 2019;20:6276. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. González‐Larraza PG, López‐Goerne TM, Padilla‐Godínez FJ, González‐López MA, Hamdan‐Partida A, Gómez E. IC50 evaluation of platinum nanocatalysts for cancer treatment in fibroblast, HeLa, and DU‐145 cell lines. ACS Omega. 2020;5:25381‐25389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Turturro F, Von Burton G, Friday E. Hyperglycemia‐induced thioredoxin‐interacting protein expression differs in breast cancer‐derived cells and regulates paclitaxel IC50. Clin Cancer Res. 2007;13:3724‐3730. [DOI] [PubMed] [Google Scholar]
  • 39. Yu L, Yang Z, Liu Y, et al. Identification of SPRR3 as a novel diagnostic/prognostic biomarker for oral squamous cell carcinoma via RNA sequencing and bioinformatic analyses. PeerJ. 2020;8:e9393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. de A Simão T, Souza‐Santos PT, de Oliveira DS, et al. Quantitative evaluation of SPRR3 expression in esophageal squamous cell carcinoma by qPCR and its potential use as a biomarker. Exp Mol Pathol. 2011;91:584‐589. [DOI] [PubMed] [Google Scholar]
  • 41. Luo A, Chen H, Ding F, et al. Small proline‐rich repeat protein 3 enhances the sensitivity of esophageal cancer cells in response to DNA damage‐induced apoptosis. Mol Oncol. 2013;7:955‐967. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix S1

Click here for additional data file. (1.8MB, zip)

Data Availability Statement

The data used to support the findings of this study comes from the public database and are available from the corresponding author upon request.

Articles from Journal of Clinical Laboratory Analysis are provided here courtesy of Wiley

RESOURCES