Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Jul 9;10(19):8633-8647.
doi: 10.7150/thno.47938. eCollection 2020.

Identification of a Tumor Microenvironment-relevant Gene set-based Prognostic Signature and Related Therapy Targets in Gastric Cancer

Wang-Yu Cai  1   2 , Zi-Nan Dong  1   2 , Xiao-Teng Fu  3 , Ling-Yun Lin  1   2 , Lin Wang  2 , Guo-Dong Ye  4 , Qi-Cong Luo  2   5 , Yu-Chao Chen  6
Affiliations

Identification of a Tumor Microenvironment-relevant Gene set-based Prognostic Signature and Related Therapy Targets in Gastric Cancer

Wang-Yu Cai et al. Theranostics. .

Abstract

Rationale: The prognosis of gastric cancer (GC) patients is poor, and there is limited therapeutic efficacy due to genetic heterogeneity and difficulty in early-stage screening. Here, we developed and validated an individualized gene set-based prognostic signature for gastric cancer (GPSGC) and further explored survival-related regulatory mechanisms as well as therapeutic targets in GC. Methods: By implementing machine learning, a prognostic model was established based on gastric cancer gene expression datasets from 1699 patients from five independent cohorts with reported full clinical annotations. Analysis of the tumor microenvironment, including stromal and immune subcomponents, cell types, panimmune gene sets, and immunomodulatory genes, was carried out in 834 GC patients from three independent cohorts to explore regulatory survival mechanisms and therapeutic targets related to the GPSGC. To prove the stability and reliability of the GPSGC model and therapeutic targets, multiplex fluorescent immunohistochemistry was conducted with tissue microarrays representing 186 GC patients. Based on multivariate Cox analysis, a nomogram that integrated the GPSGC and other clinical risk factors was constructed with two training cohorts and was verified by two validation cohorts. Results: Through machine learning, we obtained an optimal risk assessment model, the GPSGC, which showed higher accuracy in predicting survival than individual prognostic factors. The impact of the GPSGC score on poor survival of GC patients was probably correlated with the remodeling of stromal components in the tumor microenvironment. Specifically, TGFβ and angiogenesis-related gene sets were significantly associated with the GPSGC risk score and poor outcome. Immunomodulatory gene analysis combined with experimental verification further revealed that TGFβ1 and VEGFB may be developed as potential therapeutic targets of GC patients with poor prognosis according to the GPSGC. Furthermore, we developed a nomogram based on the GPSGC and other clinical variables to predict the 3-year and 5-year overall survival for GC patients, which showed improved prognostic accuracy than clinical characteristics only. Conclusion: As a tumor microenvironment-relevant gene set-based prognostic signature, the GPSGC model provides an effective approach to evaluate GC patient survival outcomes and may prolong overall survival by enabling the selection of individualized targeted therapy.

Keywords: gastric cancer; machine learning; prognostic signature; targeted therapy; tumor microenvironment.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: The authors have declared that no competing interest exists.

Figures

Figure 1
Figure 1
Study flowchart. DEGs: differentially expressed genes; GPSGC: gene expression-based prognostic signature for gastric cancer; KM: Kaplan-Meier; mfIHC: multiplex fluorescent immunohistochemistry; TME: tumor microenvironment.
Figure 2
Figure 2
Generation of the GPSGC model from TCGA and ACRG training cohorts. (A) Kaplan-Meier curves for the high (n = 78) and low (n = 264) GPSGC risk score patient groups in the TCGA-STAD cohort. Log-rank test, P < 0.0001. (B) Kaplan-Meier curves for the high (n = 92) and low (n = 208) GPSGC risk score patient groups in the ACRG cohort. Log-rank test, P < 0.0001. (C-D) The relationships between the expression of three prognostic genes (upper) and GPSGC risk score distribution with survival status (bottom) in the TCGA-STAD (C) and ACRG (D) cohorts are shown; the X axis is sorted by GPSGC risk scores. Patients were divided into high-risk and low-risk groups with GPSGC risk score = 0.15 utilized as the cutoff value.
Figure 3
Figure 3
Kaplan-Meier curves of overall survival or recurrence-free survival according to GPSGC risk score in different gastric cancer validation cohorts. (A) GSE15459 (n = 192), (B) GSE26253 (n = 432), and (C) GSE84437 (n = 433). The provided P values are from log-rank tests. Patients were divided into high-risk and low-risk groups with GPSGC risk score = 0.15 utilized as the cutoff value.
Figure 4
Figure 4
Association of TME subcomponents with GPSGC risk score and outcome in patients with gastric cancer. The 834 gastric cancer patients involved in the analysis were from the TCGA-STAD (n = 342), ACRG (n = 300) and GSE15459 (n = 192) cohorts. (A) Scatter plots depicting the low positive correlation between immune score and GPSGC risk score in human gastric cancer samples. The fitted curve of the relation between immune score and GPSGC risk score was obtained by locally weighted scatterplot smoothing (LOWESS). Pearson's correlation coefficient is shown in the graphs (P < 1.0×10-6). (B) Scatter plots depicting the strong positive correlation between stromal score and GPSGC risk score in human gastric cancer samples. The fitted curve of the relation between stromal score and GPSGC risk score was obtained by LOWESS. Pearson's correlation coefficient is shown in the graphs (P < 1.0×10-6). (C) Kaplan-Meier curves for overall survival of 834 gastric cancer patients according to immune score. Log-rank test, P = 0.51. (D) Kaplan-Meier curves for overall survival of 834 gastric cancer patients according to stromal score. Log-rank test, P = 0.0031.
Figure 5
Figure 5
Association of TME cell types, panimmune gene sets, immunomodulatory (IM) genes with GPSGC risk score and outcome in patients with gastric cancer. The 834 gastric cancer patients involved in the analysis were from the TCGA-STAD (n = 342), ACRG (n = 300) and GSE15459 (n = 192) cohorts. (A) Among the 64 TME cell types, those significantly related to overall survival (log-rank test, P < 0.05) and GPSGC risk score (Pearson's correlation test, |r| ≥ 0.40, P < 0.05) are listed. The square data markers indicate estimated hazard ratios. The error bars represent 95% CIs. Pearson's correlation coefficients between 9 TME cell types and stromal scores are also shown (P < 0.05). (B) Among the 110 panimmune gene sets, those significantly related to overall survival (log-rank test, P < 0.05) and GPSGC risk score (Pearson's correlation test, |r| ≥ 0.40, P < 0.05) are listed. The square data markers indicate estimated hazard ratios. The error bars represent 95% CIs. Pearson's correlation coefficients between 10 panimmune gene sets and stromal scores are also shown (P < 0.05). (C) Among the 60 immunomodulatory genes, those significantly related to GPSGC risk score (Pearson's correlation test, |r| ≥ 0.40, P < 0.05) are listed. The overall survival analysis of 6 immunomodulatory genes is presented. The numbers marked in red denote estimates with a log-rank test P-value < 0.05. The square data markers indicate estimated hazard ratios. The error bars represent the 95% CIs. Pearson's correlation coefficients between the 6 immunomodulatory genes and stromal scores are also shown (P < 0.05).
Figure 6
Figure 6
Multiplex fluorescent immunohistochemistry (mfIHC) analysis of the relationship between GPSGC risk score, therapeutic target expression and overall survival with GC tissue microarray data. (A) Kaplan-Meier curves for high (n = 59) and low (n = 127) GPSGC risk score patient groups in GC tissue microarray data. Log-rank test, P < 0.0001. (B) Kaplan-Meier curves for high (n = 91) and low (n = 95) TGFβ1 expression patient groups in GC tissue microarray data. Log-rank test, P = 0.0006. (C) Kaplan-Meier curves for high (n = 97) and low (n = 89) VEGFB expression patient groups in GC tissue microarray data. Log-rank test, P = 0.0018. (D) mfIHC showed the protein expression and localization of VCAN (green), CLIP4 (cyan), and MATN3 (yellow) and the therapeutic targets TGFβ1 (orange) and VEGFB (red) in GC tissue. DAPI: blue; scale bar: 50 µm. (E) Scatter plots depicting the positive correlation between GPSGC risk score and TGFβ1 expression in GC tissue microarray data. Pearson's correlation coefficient is shown in the graphs (n = 186, P < 0.0001). (F) Scatter plots depicting the positive correlation between GPSGC risk score and VEGFB expression in GC tissue microarray data. Pearson's correlation coefficient is shown in the graphs (n = 186, P < 0.0001).
Figure 7
Figure 7
Multivariate Cox analysis evaluating independently predictive ability of the GPSGC and other clinical risk factors for OS. The square data markers indicate estimated hazard ratios. The error bars represent 95% CIs.
Figure 8
Figure 8
Construction and evaluation of a nomogram based on the GPSGC to predict the 3-year and 5-year overall survival for GC patients. (A) Nomogram was constructed with the TCGA-STAD and ACRG training cohorts (n = 642) for predicting the probability of 3-year and 5-year OS for GC patients. (B) Calibration plot of the nomogram for predicting the probability of OS at 3, and 5 years in GSE15459 validation cohort (n = 192). (C) Calibration plot of the nomogram for predicting the probability of OS at 3, and 5 years in the experimental tissue array validation cohort (n = 186). The grey line represents the ideal nomogram, and the red line represents the observed nomogram.
See this image and copyright information in PMC

References

    1. 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. - PubMed
    1. Mattiuzzi C, Lippi G. Cancer statistics: a comparison between World Health Organization (WHO) and Global Burden of Disease (GBD) Eur J Public Health. 2019. - PubMed
    1. Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F. et al. Cancer statistics in China, 2015. CA: a cancer journal for clinicians. 2016;66:115–32. - PubMed
    1. Van Cutsem E, Sagaert X, Topal B, Haustermans K, Prenen H. Gastric cancer. Lancet. 2016;388:2654–64. - PubMed
    1. Gao JP, Xu W, Liu WT, Yan M, Zhu ZG. Tumor heterogeneity of gastric cancer: From the perspective of tumor-initiating cell. World J Gastroenterol. 2018;24:2567–81. - PMC - PubMed

Publication types

MeSH terms

Substances

爱游戏平台网页版中国有限公司 | c7电子娱乐游戏网站中国有限公司 | 米兰体育在线登录中国有限公司 | ky平台中国有限公司 | 九游手机平台-九游(中国) | 江南注册大厅 | 半岛网页版界面 | 开云手机入口 | 半岛电子（中国）官方网站 |