Characterization of the fatty acid metabolism gene signature in gastric cancer to assist precision treatment in chemotherapy

doi:10.21203/rs.3.rs-2243913/v1

Download PDF

Research Article

Characterization of the fatty acid metabolism gene signature in gastric cancer to assist precision treatment in chemotherapy

https://doi.org/10.21203/rs.3.rs-2243913/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective

This project evaluated the efficacy of fatty acid metabolism (FAM) in gastric cancer (GC) with pII/III stage to explore potential therapeutic strategies.

Methods

Clinical features, transcriptome expression and follow-up information were downloaded from TCGA and GEO databases. "Limma" package serves to analyze differentially expressed genes (DEG). We applied "clusterprofiler" package to carry out GO and KEGG aggregation analysis on DEG. The "maftools" package was used to analyze gene transformation and association. "Coef" refers to the non-zero correlation coefficient estimated according to lasso cox regression analysis. The half maximum inhibitory concentration (IC50) is calculated according to the "prrophytic" package. We also used the "rms" package to structure nomograms.

Results

A digital model of the risk score was established by using ten FAM genes. Compare to the low-risk group, the high-risk group has a worse prognosis (p < 0.001). ROC, univariate and multivariable cox analysis all verified the above results. We also constructed nomograms based on age, sex, pathological TNM stage and risk scores. The high-risk score sample is not particularly sensitive to 5-Fu chemotherapy, and the Tumor immune dysfunction and rejection (TIDE) index of the high-risk group is higher, which indicates that the high-risk group may be more prone to immune escape.

Conclusion

The FAM gene model believes that the low-risk groups have better 5-Fu drug sensitivity, lower probability of immune escape, and better prognosis.

Gastric cancer

5-Fu resistance

prognosis

fatty acid metabolism

gene signature

According to research, 25%-40% of patients with pathological II/III (pII/III) gastric cancer (GC) experienced distant metastasis of cancer cells after surgery[1, 2]. Local recurrence and distant metastasis are the primary reasons for the failure of GC surgery[3]. For this reason, postoperative chemotherapy in pII/III GC cases is of major importance[4]. Basing on the NCCN treatment guidelines, 5-fluorouracil (5-Fu) is generally the basic chemotherapy for these cases after radical resection[5]. GC is a relatively responsive tumor to chemotherapy, but not all cases can benefit from chemotherapy[6]. Moreover, the clinical efficacy and prognostic model of 5-Fu in GC have not been well evaluated yet.

As we all know, GC is a tumor with prominent heterogeneity. Different clinical characteristics and individual differences of patients show the reflection changes of 5-Fu treatment^[6]. Recent studies have shown that abnormal fatty acid metabolism (FAM) is closely linked to clinical treatment^[7–9]. However, most of the reports in the literature focus on scattered genes^{[10, 11]}, which cannot evaluate the chemo-resistance of 5-Fu as a whole. Therefore, this study comprehensively analyzed the metabolic gene sets in TCGA and GEO databases, and it was also constructed a 5-Fu chemotherapy sensitivity and clinical prediction model.

In this study, the gene expression information, clinical characteristics and follow-up information of TCGA database were fused. In order to explore the individualized chemotherapy scheme, a FAM model of GC cases was built, and the GEO database was verified as external data.

2.1 Development of prognostic risk scores model

The expression level of normal genes related to FAM in pII/III stage GC samples was compared and 55 genes were selected from TCGA sequence. A total of 31 genes increased and 24 genes decreased (Table 1, Fig. 1A/B). Univariate cox multiple regression analysis was conducted on 17 genes related to FAM (Fig. 1C). Then, the minimum constructive folding and selection algorithm (lasso) cox multiple regression analysis was used to reduce the total number of genes. Finally, ten genes (ELOVL2, FAAH, ECH1, GAD2, MCAT, ELOVL4, ACADM, FAAH2, PCCB, and ALDH3A2) were used to create a model of the post-treatment risk score (Fig. 1D and 1E). We used the following calculation formula to calculate the risk score of each sample: risk score=(0.277777997) ELOVL2+ (-0.143591363) FAAH+(0.217450902) ECH1+ (0.541576369) GAD2+ (-0.113714283) MCAT+ (0.039701796) ELOVL4+ (-0.233549272) ACADM+ (-0.095293889) FAAH2+ (-0.050352332) PCCB+ (-0.05880813) ALDH3A2, as shown in Table 2. Then PCA served to analyze the gene expression profile in the hihg- and low-risk scoring model (Fig. 1F). The ROC of time-dependent subjects was generated within 5-year to verify the accuracy of the model(Fig. 1J).

Table 1

Significantly different FAM genes of GC
gene	conMean	treatMean	logFC	pValue	fdr
CA6	0.015206605	0.520479254	5.097070665	0.002257531	0.004429397
PON1	0.041701628	0.365320114	3.130985577	0.000602978	0.00138789
CYP4F8	0.005522535	0.040264459	2.866104427	0.001057432	0.0022954
ACBD7	0.163900501	1.081396077	2.722002862	1.46E-12	4.32E-11
ELOVL3	0.061159044	0.378405045	2.629293553	7.88E-10	8.59E-09
HSD17B3	0.067524054	0.272480557	2.012679856	6.43E-07	3.18E-06
HAO2	0.015901595	0.062484374	1.97432396	0.012254962	0.019711294
ELOVL2	0.151538803	0.581706531	1.940604245	0.000133046	0.000351716
RDH16	0.081637923	0.29271291	1.842175003	3.94E-06	1.52E-05
OLAH	0.038716079	0.132134367	1.771000984	3.41E-05	0.000100102
FMO1	0.314431543	1.042850568	1.729714587	7.92E-06	2.71E-05
HSPH1	6.636837563	20.43059798	1.622163561	1.83E-16	2.45E-14
IL4I1	1.576618475	4.839050389	1.617890376	4.83E-09	4.45E-08
CPT1B	0.202032984	0.543073933	1.42655777	8.04E-10	8.59E-09
HSP90AA1	88.53151844	227.1085859	1.359119177	2.57E-18	6.87E-16
THEM5	0.889168268	2.196153473	1.30445051	0.002035949	0.004056704
SCD	23.0250525	56.09823855	1.284751026	6.97E-05	0.000197978
ELOVL7	2.436268909	5.564486176	1.191575092	1.31E-08	1.02E-07
PON2	11.28060022	25.1235606	1.15519711	1.60E-08	1.15E-07
ALOXE3	0.105441238	0.232197184	1.138911258	0.00090844	0.002038264
FADS2	2.270578594	4.957582361	1.126576765	0.001355701	0.002872796
ACLY	18.21212081	39.43608088	1.11461725	9.04E-11	1.51E-09
TDO2	0.339731294	0.720805617	1.085216135	2.38E-05	7.13E-05
ACSM3	1.078827479	2.237138638	1.052190491	0.000952764	0.002119901
GAPDHS	0.011201878	0.022809816	1.02591426	0.002263747	0.004429397
PTPRG	2.139773881	4.326489871	1.015738676	8.31E-09	6.93E-08
ACSBG2	0.043756478	0.088078985	1.009301229	0.000360886	0.000852713
CPT1C	0.457020173	0.918462403	1.006962818	0.005286348	0.009091622
SMS	20.62219531	41.29551109	1.001787045	5.45E-15	2.91E-13
GGT5	3.849336863	7.705346322	1.001249872	0.00032013	0.000784172
FASN	22.27897763	44.57906468	1.000683318	1.06E-07	6.59E-07
AKR1C3	46.55832128	20.94056669	-1.152738556	0.013702267	0.021647961
MAOA	29.5206935	12.51092889	-1.238537705	4.01E-10	5.10E-09
ACACB	5.349043188	2.142875038	-1.319733131	1.24E-07	7.52E-07
ACOX2	4.429241716	1.688715945	-1.391133055	6.58E-09	5.67E-08
CA2	242.7257058	87.99487882	-1.463835436	0.008753569	0.014516788
ACADL	0.306554488	0.110630497	-1.470394381	2.73E-10	3.88E-09
PTGS1	14.45552391	5.045415007	-1.51857605	1.80E-12	4.58E-11
ADH1C	107.0030202	37.24225357	-1.522639238	0.002272762	0.004429397
HACD1	3.535444856	1.218809132	-1.536419543	5.32E-06	1.91E-05
PRKAA2	2.839473161	0.966387022	-1.55495029	4.72E-06	1.77E-05
PTGR1	27.28530913	9.25298597	-1.56013348	6.48E-08	4.36E-07
AOC3	38.59309684	12.70078171	-1.60342552	2.07E-06	8.66E-06
CD36	5.826432672	1.784317672	-1.707240351	5.07E-09	4.51E-08
ELOVL4	2.161921178	0.581853039	-1.893587207	0.001370519	0.002881327
CBR1	59.89085094	15.86276563	-1.916691307	8.04E-10	8.59E-09
ADH4	15.51882124	3.75524817	-2.047038821	0.018345747	0.028313957
ADH1B	17.22083366	3.890477712	-2.146135765	1.59E-13	6.07E-12
HPGD	49.37558756	11.09523113	-2.153858194	1.89E-12	4.58E-11
CYP4B1	3.271629484	0.628787278	-2.379365438	1.37E-10	2.15E-09
ALDH3A1	72.79213029	12.34557603	-2.559788333	3.29E-06	1.31E-05
CYP1A1	1.109255627	0.17279993	-2.682417337	0.000887412	0.002007957
CYP4F22	3.770088776	0.467950263	-3.010171392	0.000331888	0.000798325
CIDEA	1.655274337	0.164249875	-3.333106163	1.60E-11	2.85E-10
ADH7	16.18068489	0.922334984	-4.132838045	5.68E-12	1.26E-10

Table 2

ten genes were selected to set up the entity model of risk score
Gene	Coef
ELOVL2	0.277777997
FAAH	-0.143591363
ECH1	0.217450902
GAD2	0.541576369
MCAT	-0.113714283
ELOVL4	0.039701796
ACADM	-0.233549272
FAAH2	-0.095293889
PCCB	-0.050352332
ALDH3A2	-0.05880813

2.2 Correlation of risk scores with clinical characteristics

There was no significant differences between the risk scores and clinical characteristics (age, gender and T stages, all p > 0.001, Fig. 2A-D). Compared against the low-risk group, the prognosis in the high-risk group was worse (Fig. 2E for TCGA set and Fig. 2f for GEO set, all p < 0.001). In univariate analysis, the key factors related to OS included TNM stage and risk score. In multivariate analysis, only risk score was an independent predictor of OS (all p < 0.001, Fig. 2G and 2H).

2.3 Construction of a nomogram for predicting survival

The sankey diagram showing the comprehensive age, sex, pathological TNM staging and post-treatment risk score was established (Fig. 3A), and the nomograph model for the prediction of stage II/III GC cases is also built (Fig. 3B). The calibration curves of 1-, 3-, and 5- years confirmed that the nomogram can accurately predict the prognosis (Fig. 3C). Univariate and multivariate cox multiple regression analysis showed that the model of post-treatment risk score was a separate post-treatment index value (Fig. 3D and E).

2.4 Relationship between FAM-related prediction models and immune targets

GSVA enrichment finds that FAM signaling is underexpressed in high-risk groups (Fig. 4A). In the high-risk subgroup, the proportion of monocytes and macrophages M2 was significantly increased (all p < 0.05, Fig. 4B). There was no significant difference in the expression level of PD-L1(Fig. 4C), CTLA4(Fig. 4D), and TMB(Fig. 4E) between the high-risk group and the low-risk group (all p > 0.001).

2.5 Immune heterogeneity and functioning immune cells

The lymphocyte depleted subtype has significantly higher risk values than other subtypes(all p > 0.05, Fig. 5A). The proportion of T cell follicular helper and mast cell resting was significantly higher in the GC high-risk group(Fig. 5B). Mast cell resting is inversely correlated with risk value of FAM gene prediction model(Fig. 5C). T cell follicular helper is positively associated with risk value in FAM gene prediction model(Fig. 5D).

2.6 Gene ontology and KEGG pathway analysis

According to the clusterprofiler package, we obtained GO and KEGG enrichment of all target FAM genes participating. The GO analysis results show that 270 GO annotations are related to the development trend of FAM. The dot chart shows the top terms of the biological process (BP, Fig. 6A), cellular component (CC, Fig. 6B) and molecular function of all molecular function (MF, Fig. 6C). Top 15 signaling pathways of KEGG enrichment analysis (Fig. 6D) of genes to pathway-pathway network(Fig. 6E). Type I/II IFN reponse was significantly increased in the FAM-related high-risk group (p < 0.05, Fig. 6F).

2.7 Response to chemotherapy response

The correlation between TNM staging and clinical characteristic was shown (Fig. 7A). We use the “factoextra” package to generate the optimal number of clusters, and the optimal number of clusters was generated (K = 2, Fig. 7B/C). By non-negative matrix factorization (NMF) method, patients were divided into two distinct subtypes (Fig. 7D). The IC50 was used to predict the treatment response to 5-FU in TCGA cohort. Low-risk score samples were more sensitive to 5-FU, widely used in treating stage II-III GC (Fig. 7E/F).The high risk group is liable to be more prone to immune escape (Fig. 7G).

According to NCCN diagnosis and treatment guidelines, GC patients with pII/III stage should propose 5-Fu based chemotherapy^[12].However, due to personal homogeneity, especially the difference in the metabolic microenvironment of somatic cells, some patients generally develop resistance to chemotherapy^[13]. Recent studies have found that reprogramming of somatic metabolism plays a major role in the development trend of malignant tumors^{[14, 15]}. Although most scientific research focuses on the role of single regulatory factors of FAM in GC^[10], the overall efficacy of several FAM-related genes is unclear. Tumor cells show different metabolic characteristics from normal somatic cells^[16]. This kind of connection may be related to the drug resistance of tumors to 5-FU treatment, but it has not been evaluated.

The results of the JCOG1104 trial^[17] showed that for pII GC, the use of S-1 single-agent postoperative chemotherapy has better clinical benefit, which is consistent with the ACTS-GC trial results^{[18, 19]}. According to the results of the JACCRO GC-07 trial^{[20, 21]}, multi-drug combination chemotherapy is recommended for pIII GC. Conventional regimens such as S-1 plus docetaxel or oxaliplatin combination regimen can be used^[22]. However, due to individual heterogeneity, not everyone can benefit from this kind of chemotherapy regimen^[23]. The choice of postoperative adjuvant chemotherapy should not only refer to pathological diagnosis, but also evaluate the patient's metabolism and drug resistance, so as to formulate the best individualized countermeasures.

In our research, 55 genetic genes related to FAM were selected and the number of hereditary genes was reduced. Totally, ten genetic genes were utilized to build a risk scoring model. Compared with the low-risk set, the survival time in the high-risk set was lower (all p < 0.001). The ROC of time-dependent subjects was carried out within 5 years to determine the accuracy of the risk scoring model (5 years, AUC = 0.793). Univariate and multivariate analysis verified the effectiveness of the model (all p < 0.001). Therefore, we have basically confirmed that this kind of FAM model can effectively distinguish the prognosis of patients. However, the ability of this model to individually assess 5-Fu treatment resistance is still unknown. To continue to explore the resistance to 5-Fu treatment, we first clustered patients into two distinct sub-types by NMF method. Subsequently, we further evaluated the chemo-sensitivity to 5-Fu, and it was found that the high-risk set was lower sensitive to 5-Fu chemotherapy. This may also partly explain this higher proportion of 5-Fu chemotherapy resistance in the high-risk group.

The relationship between FAM and immune microenvironment has been reported in the literature. It was found that the TIDE index of high-risk groups is higher, which indicates that high-risk groups may be more prone to immune escape. GSVA enrichment was conducted and it was found that most metabolism pathways, including FAM, got rich in the low-risk score. It was also found that in the high-risk group, the proportion of T cells follicular helper level was high, but the level of mast cells resting was low. According to the literature reports, the immune status of different subtypes may also lead to distinct sensitivity of immunotherapy. Frustratingly, we did not find any difference in immune binding sites (PD-L1 and CTLA4) and TMB. This may indicate that pII-III GC does not benefits from immunotherapy. However, these results need to be verified by future large sample, multi center clinical studies.

In order to evaluate the effectiveness of the FAM model independently, we integrated the metabolic model and clinical characteristics into the analysis. Firstly, a sankey diagram integrating age, gender, pathological TNM staging and post-treatment risk score was proposed, and a nomograph model was built. We can confirm that the low-risk score of FAM model is close to 10 points, but the score of the high-risk group is increased to about 55 points. Therefore, the prognosis of patients can be independently evaluated by this model.

The limitations of our research include : firstly, this study is based on publicly available data, some important clinicopathological features cannot be analyzed, which may lead to potential errors or bias. Secondly, the GPL6947 platform includes only a small portion of all possible mRNA, which may not represent a complete mRNA population of GC biological behavior.

In conclusion, the fatty acid prognostic risk score model can comprehensively evaluate the clinical characteristics of stage II-III GC. In addition, The model believes that owners of low-risk groups have better 5-Fu drug sensitivity, lower probability of immune escape, and better prognosis.

4.1 Data collections

Raw RNA transcriptase sequencing (RNA-seq) data information system variables, including 375 GC and 32 normal gastric tissue cases, are loaded from transcriptase sequencing (HTSeq) of TCGA database. According to GSE84437 microarray data information of GPL6947 service platform was download from GEO database. In previous systematic studies, 309 genes related to FAM were evaluated (table S1). Clinical medical information about GC patients, including age, gender, TNM stage and prognosis information, which can also be collected from TCGA and GEO databases. The annotation service platform is used for convert the Entrez gene ID of each sample into the corresponding gene marker. If the same ID is identified by multiple probes, the average value is chosen. In addition, clinical medical information of each sample in GSE84437 can be downloaded from the GEO database. In a preceding study, FAM-related genes were identification^[24].

4.2 Enrichment analyses of the DEGs

The "limma" package is used to analyze differentially expressed genes (DEGs) related to FAM in normal and GC tissue samples. Then, the "org.Hs.eg.db" package was used to convert the marker gene of each DEG (table S2). The "clusterprofiler" package was used to carry out gene ontology (GO) and kyoto encyclopedia of genes and genomes (KEGG) enrichment analysis on DEGs to show the key biological characteristics and the mode of action of tissues and cells. Finally, the enrichment analysis results are visualized by using "enrichplot" and "ggplot2" packages.

4.3 Establishment of a prognostic risk score model

According to the univariate cox multiple regression analysis, genes related to prognosis were selected from the genes related to FAM reflected in the differences. The "Maftools" package is utilized to analyze the transformation. We used "glmnet" package and lasso cox multiple regression analysis. we can further solve the relevant genes, so as to create a FAM-related gene model for predicting risk score and prognosis. Ten cross verifications were carried out to respond to the basic parameters of the gene model. The following formula serves to calculate calculate the risk performance of each sample. "coef" refers to the non-zero correlation coefficient estimated according to lasso cox regression analysis, "expgene" refers to the gene expression value obtained from the entity model of post-treatment risk score. The "survivalroc" package is used to ensure the prediction and analysis accuracy of the prognosis. "Cmscaler" package is used to classify samples according to the characteristics of TCGA samples.

4.4 Principal-component analyses (PCA) comparison

In practice, we successfully applied the "limma" package to carry out a principal component analysis, so as to grasp the significant differences between the low- and high-risk scoring groups. Firstly, PCA was performed on all DEGs. Then PCA served to analyze the gene expression profile in the risk scoring model. Finally, according to the first two principal components, the results of principal component analysis are posted on the two-dimensional diagram by using "ggplot2" package.

4.5 Gene sets variation analyses

According to the "GSVA" package, we carried out gene set variation analysis (GSVA) on the gene map to distinguish the whole biological process between two groups. The "c2.cp.kegg.v7.1.symbols" gene set in the molecular characteristics database is served as the reference gene set. FDR < 0.05 indicates a statistically meaningful aggregation pathway.

4.6 Drug susceptibility assessment of 5-Fu

The half-maximal inhibitory concentration (IC50), calculated by the “pRRophetic” package, was utilized to predict the treatment response to 5-Fu. ssGSEA was performed using “GSEABase” package to quantify the extent of the immune-related infiltration of GC sample^{[25, 26]}. Tumor immune dysfunction and exclusion (TIDE) and subMap algorithms were used to predict immune checkpoint response inhibitors of PD-1 and CTLA4.

4.7 Constructing a FAM-related gene prediction nomogram

According to the TCGA sequence, a nomogram was developed and designed by "rms" package to predict the overall survival rate (OS). Therefore, a calibration curve related to time is suggested to predict the accuracy of nomograph. In addition, multiple cox multiple linear regression analysis was used to determine whether the digital model can be used as an independent indicator. Then the area under the curve (AUC) is estimated according to the online receiver operating characteristic (ROC) curve to show the effect and use value of nomograph.

4.8 Statistical analyses

The wilcoxon rank test was used to compare the differences between clinical features. Kaplan-Meier analysis served to examine examine the difference in survival between the low- and high-risk groups. Multivariate cox regression analysis serves to clearly predict and analyze the distinctive index value of OS. The ROC curve is taken to evaluate the estimated actual effect of post-treatment risk scoring method and nomogram. All data were analysed by R4.2.1.

Gastric cancer

pathological II/III

pII/III

5-Fu

5-fluorouracil

FAM

fatty acid metabolism

TMB

tumor mutation burden

TCGA

The Cancer Genome Atlas

GEO

Gene Expression Omnibus

PCA

Principal-component analyses

IC50

The half-maximal inhibitory concentration

overall survival

DEG

differentially expressed gene

FDR

false discovery rate

gene ontology

biological process

cellular component

molecular function

KEGG

Kyoto Encyclopedia of Genes and Genomes

GSEA

gene set enrichment analysis

LASSO

least absolute shrinkage and selection operator

ROC

receiver operator characteristic

AUC

area under the curve

ssGSEA

single-sample gene set enrichment analysis

hazard ratio

confidence interval

TIDE

Tumor Immune Dysfunction and Exclusion

NCCN

The National Comprehensive Cancer Network

RNA-seq

RNA transcriptase sequencing

HTSeq

the transcriptase sequencing

GSVA

gene set gene variation analysis

IFN

interferon

overall survival

NMF

non-negative matrix factorization.

Data Availability:

The data used to support the findings of this study are included within the supplementary information files.

Conflicts of Interest:

The authors declare no conﬂicts of interest.

Funding Statement:

This study was supported by Fujian Provincial health technology project (Grant No. 2017-ZQN-18, 2021GGA041), supported by grants from Fujian science and technology project (Grant No. 2019J05139, 2019J01200).

Acknowledgments:

Not applicable.

Ethics & Consent to participate

Not applicable.

Author contributions

ZWD and XJ wrote the main manuscript, XJ and ZLY performed the analyses. XJ, and ZLY collected data and prepared all Fig.s and tables. XJ and ZWD conceived of and designed the study. All authors reviewed and approved the manuscript.

Aurello P, Minervini A, Pace M, D'Angelo F, Nigri G, Antolino L, Valabrega S, Ramacciato G, Petrucciani N. The Role of Surgery in the Treatment of Metachronous Liver Metastasis from Gastric Cancer: A Systematic Review. Anticancer Res 2022, 42(1): 25–33.
Abdelfatah MM, Barakat M, Othman MO, Grimm IS, Uedo N. The incidence of lymph node metastasis in submucosal early gastric cancer according to the expanded criteria: a systematic review. Surg Endosc 2019, 33(1): 26–32.
Aoyama T, Komori K, Tamagawa A, Nakazano M, Hara K, Hashimoto I, Tamagawa H, Segami K, Maezawa Y, Kano K, Oshima T, Yukawa N, Rino Y. Clinical Influence of the Lymph Node Ratio on Lymph Node Metastasis-positive Gastric Cancer Patients Who Receive Curative Treatment. In Vivo 2022, 36(2): 994–1000.
Tanaka H, Yoshii M, Imai T, Tamura T, Toyokawa T, Muguruma K, Hirakawa K, Ohira M. Clinical significance of coexisting histological diffuse type in stage II/III gastric cancer. Mol Clin Oncol 2021, 15(5): 234.
Maspero M, Sposito C, Benedetti A, Virdis M, Di Bartolomeo M, Milione M, Mazzaferro V. Impact of Surgical Margins on Overall Survival after Gastrectomy for Gastric Cancer: A Validation of Japanese Gastric Cancer Association Guidelines on a Western Series. Ann Surg Oncol 2022, 29(5): 3096–3108.
Siebenhuner AR, De Dosso S, Helbling D, Astaras C, Szturz P, Moosmann P, Pederiva S, Winder T, Von Burg P, Borner M. Advanced Gastric Cancer: Current Treatment Landscape and a Future Outlook for Sequential and Personalized Guide: Swiss Expert Statement Article. Oncol Res Treat 2021, 44(9): 485–494.
Romo-Perez A, Dominguez-Gomez G, Chavez-Blanco A, Taja-Chayeb L, Gonzalez-Fierro A, Diaz-Romero C, Lopez-Basave HN, Duenas-Gonzalez A. Progress in Metabolic Studies of Gastric Cancer and Therapeutic Implications. Curr Cancer Drug Targets 2022.
Tokes AM, Vari-Kakas S, Kulka J, Torocsik B. Tumor Glucose and Fatty Acid Metabolism in the Context of Anthracycline and Taxane-Based (Neo)Adjuvant Chemotherapy in Breast Carcinomas. Front Oncol 2022, 12: 850401.
Lyu Q, Zhu W, Wei T, Ding W, Cao M, Wang Q, Guo L, Luo P, Zhang J. High mutations in fatty acid metabolism contribute to a better prognosis of small-cell lung cancer patients treated with chemotherapy. Cancer Med 2021, 10(21): 7863–7876.
Dong SR, Ju XL, Yang WZ. STAT5A reprograms fatty acid metabolism and promotes tumorigenesis of gastric cancer cells. Eur Rev Med Pharmacol Sci 2019, 23(19): 8360–8370.
Jiang Z, Shen H, Tang B, Yu Q, Ji X, Wang L. Quantitative proteomic analysis reveals that proteins required for fatty acid metabolism may serve as diagnostic markers for gastric cancer. Clin Chim Acta 2017, 464: 148–154.
Ajani JA, D'Amico TA, Bentrem DJ, Chao J, Corvera C, Das P, Denlinger CS, Enzinger PC, Fanta P, Farjah F, Gerdes H, Gibson M, Glasgow RE, Hayman JA, Hochwald S, Hofstetter WL, Ilson DH, Jaroszewski D, Johung KL, Keswani RN, Kleinberg LR, Leong S, Ly QP, Matkowskyj KA, McNamara M, Mulcahy MF, Paluri RK, Park H, Perry KA, Pimiento J, Poultsides GA, Roses R, Strong VE, Wiesner G, Willett CG, Wright CD, McMillian NR, Pluchino LA. Esophageal and Esophagogastric Junction Cancers, Version 2.2019, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2019, 17(7): 855–883.
Margalit O, Yu S, Shacham-Shmueli E, Strauss G, Yang YX, Lawerence YR, Reiss KA, Golan T, Halpern N, Aderka D, Giantonio B, Mamtani R, Boursi B. Benefit for single-agent adjuvant chemotherapy in elderly patients with locally advanced gastric adenocarcinoma. J Cancer Res Clin Oncol 2022.
da Silva EL, Mesquita FP, de Sousa Portilho AJ, Bezerra ECA, Daniel JP, Aranha ESP, Farran S, de Vasconcellos MC, de Moraes MEA, Moreira-Nunes CA, Montenegro RC. Differences in glucose concentration shows new perspectives in gastric cancer metabolism. Toxicol In Vitro 2022: 105357.
Rabben HL, Andersen GT, Olsen MK, Overby A, Ianevski A, Kainov D, Wang TC, Lundgren S, Gronbech JE, Chen D, Zhao CM. Neural signaling modulates metabolism of gastric cancer. iScience 2021, 24(2): 102091.
Di Gregorio J, Petricca S, Iorio R, Toniato E, Flati V. Mitochondrial and metabolic alterations in cancer cells. Eur J Cell Biol 2022, 101(3): 151225.
Yoshikawa T, Terashima M, Mizusawa J, Nunobe S, Nishida Y, Yamada T, Kaji M, Fukushima N, Hato S, Choda Y, Yabusaki H, Yoshida K, Ito S, Takeno A, Yasuda T, Kawachi Y, Katayama H, Fukuda H, Boku N, Sano T, Sasako M. Four courses versus eight courses of adjuvant S-1 for patients with stage II gastric cancer (JCOG1104 [OPAS-1]): an open-label, phase 3, non-inferiority, randomised trial. Lancet Gastroenterol Hepatol 2019, 4(3): 208–216.
Ito S, Ohashi Y, Sasako M. Survival after recurrence in patients with gastric cancer who receive S-1 adjuvant chemotherapy: exploratory analysis of the ACTS-GC trial. BMC Cancer 2018, 18(1): 449.
Terashima M, Ichikawa W, Ochiai A, Kitada K, Kurahashi I, Sakuramoto S, Katai H, Sano T, Imamura H, Sasako M, Group A-G. TOP2A, GGH, and PECAM1 are associated with hematogenous, lymph node, and peritoneal recurrence in stage II/III gastric cancer patients enrolled in the ACTS-GC study. Oncotarget 2017, 8(34): 57574–57582.
Kakeji Y, Yoshida K, Kodera Y, Kochi M, Sano T, Ichikawa W, Lee SW, Shibahara K, Shikano T, Kataoka M, Ishiguro A, Ojima H, Sakai Y, Musha N, Takase T, Kimura T, Takeuchi M, Fujii M. Three-year outcomes of a randomized phase III trial comparing adjuvant chemotherapy with S-1 plus docetaxel versus S-1 alone in stage III gastric cancer: JACCRO GC-07. Gastric Cancer 2022, 25(1): 188–196.
Yoshida K, Kodera Y, Kochi M, Ichikawa W, Kakeji Y, Sano T, Nagao N, Takahashi M, Takagane A, Watanabe T, Kaji M, Okitsu H, Nomura T, Matsui T, Yoshikawa T, Matsuyama J, Yamada M, Ito S, Takeuchi M, Fujii M. Addition of Docetaxel to Oral Fluoropyrimidine Improves Efficacy in Patients With Stage III Gastric Cancer: Interim Analysis of JACCRO GC-07, a Randomized Controlled Trial. J Clin Oncol 2019, 37(15): 1296–1304.
Ajani JA, D'Amico TA, Bentrem DJ, Chao J, Cooke D, Corvera C, Das P, Enzinger PC, Enzler T, Fanta P, Farjah F, Gerdes H, Gibson MK, Hochwald S, Hofstetter WL, Ilson DH, Keswani RN, Kim S, Kleinberg LR, Klempner SJ, Lacy J, Ly QP, Matkowskyj KA, McNamara M, Mulcahy MF, Outlaw D, Park H, Perry KA, Pimiento J, Poultsides GA, Reznik S, Roses RE, Strong VE, Su S, Wang HL, Wiesner G, Willett CG, Yakoub D, Yoon H, McMillian N, Pluchino LA. Gastric Cancer, Version 2.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2022, 20(2): 167–192.
Hirahara N, Matsubara T, Kaji S, Yamamoto T, Hyakudomi R, Takai K, Ishitobi K, Uchida Y, Tajima Y. Phase II feasibility study of adjuvant chemotherapy with docetaxel/cisplatin/S-1 followed by S-1 for stage III gastric cancer. BMC Cancer 2021, 21(1): 1073.
Yang C, Huang X, Liu Z, Qin W, Wang C. Metabolism-associated molecular classification of hepatocellular carcinoma. Mol Oncol 2020, 14(4): 896–913.
Charoentong P, Finotello F, Angelova M, Mayer C, Efremova M, Rieder D, Hackl H, Trajanoski Z. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell Rep 2017, 18(1): 248–262.
Ru B, Wong CN, Tong Y, Zhong JY, Zhong SSW, Wu WC, Chu KC, Wong CY, Lau CY, Chen I, Chan NW, Zhang J. TISIDB: an integrated repository portal for tumor-immune system interactions. Bioinformatics 2019, 35(20): 4200–4202.

No competing interests reported.

TableS1.xlsx
Supplementary table S1.Expression levels of fatty acid metabolism genes in TCGA database
TableS2.xls
Supplementary table S2.Differential analysis of fatty acid metabolism genes using "limma", set values as logFCfilter=1, fdrFilter=0.05

Download PDF

Version 1

posted

You are reading this latest preprint version

Characterization of the fatty acid metabolism gene signature in gastric cancer to assist precision treatment in chemotherapy

Status:

Version 1

Abstract

Objective

Methods

Results

Conclusion

Figures

1. Introduction

2. Results

2.1 Development of prognostic risk scores model

2.2 Correlation of risk scores with clinical characteristics

2.3 Construction of a nomogram for predicting survival

2.4 Relationship between FAM-related prediction models and immune targets

2.5 Immune heterogeneity and functioning immune cells

2.6 Gene ontology and KEGG pathway analysis

2.7 Response to chemotherapy response

3. Discussion

4. Materials And Methods

4.1 Data collections

4.2 Enrichment analyses of the DEGs

4.3 Establishment of a prognostic risk score model

4.4 Principal-component analyses (PCA) comparison

4.5 Gene sets variation analyses

4.6 Drug susceptibility assessment of 5-Fu

4.7 Constructing a FAM-related gene prediction nomogram

4.8 Statistical analyses

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1