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
. 2023 May 28;48(5):671-681.
doi: 10.11817/j.issn.1672-7347.2023.230069.

Analysis on tumor immune microenvironment and construction of a prognosis model for immune-related skin cutaneous melanoma

[Article in English, Chinese]
Meng Wu  1   2 , Zheng Wang  3 , Jianglin Zhang  4   5
Affiliations

Analysis on tumor immune microenvironment and construction of a prognosis model for immune-related skin cutaneous melanoma

[Article in English, Chinese]
Meng Wu et al. Zhong Nan Da Xue Xue Bao Yi Xue Ban. .

Abstract
in English, Chinese

Objectives: Malignant melanoma is a highly malignant and heterogeneous skin cancer. Although immunotherapy has improved survival rates, the inhibitory effect of tumor microenvironment has weakened its efficacy. To improve survival and treatment strategies, we need to develop immune-related prognostic models. Based on the analysis of the Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO), and Sequence Read Archive (SRA) database, this study aims to establish an immune-related prognosis prediction model, and to evaluate the tumor immune microenvironment by risk score to guide immunotherapy.

Methods: Skin cutaneous melanoma (SKCM) transcriptome sequencing data and corresponding clinical information were obtained from the TCGA database, differentially expressed genes were analyzed, and prognostic models were developed using univariate Cox regression, the LASSO method, and stepwise regression. Differentially expressed genes in prognostic models confirmed by real-time reverse transcription PCR (real-time RT-PCR) and Western blotting. Survival analysis was performed by using the Kaplan-Meier method, and the effect of the model was evaluated by time-dependent receiver operating characteristic curve as well as multivariate Cox regression, and the prognostic model was validated by 2 GEO melanoma datasets. Furthermore, correlations between risk score and immune cell infiltration, Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data (ESTIMATE) score, immune checkpoint mRNA expression levels, tumor immune cycle, or tumor immune micro-environmental pathways were analyzed. Finally, we performed association analysis for risk score and the efficacy of immunotherapy.

Results: We identified 4 genes that were differentially expressed in TCGA-SKCM datasets, which were mainly associated with the tumor immune microenvironment. A prognostic model was also established based on 4 genes. Among 4 genes, the mRNA and protein levels of killer cell lectin like receptor D1 (KLRD1), leukemia inhibitory factor (LIF), and cellular retinoic acid binding protein 2 (CRABP2) genes in melanoma tissues differed significantly from those in normal skin (all P<0.01). The prognostic model was a good predictor of prognosis for patients with SKCM. The patients with high-risk scores had significantly shorter overall survival than those with low-risk scores, and consistent results were achieved in the training cohort and multiple validation cohorts (P<0.001). The risk score was strongly associated with immune cell infiltration, ESTIMATE score, immune checkpoint mRNA expression levels, tumor immune cycle, and tumor immune microenvironmental pathways (P<0.001). The correlation analysis showed that patients with the high-risk scores were in an inhibitory immune microenvironment based on the prognostic model (P<0.01).

Conclusions: The immune-related SKCM prognostic model constructed in this study can effectively predict the prognosis of SKCM patients. Considering its close correlation to the tumor immune microenvironment, the model has some reference value for clinical immunotherapy of SKCM.

目的: 恶性黑色素瘤是高度恶性和异质性的皮肤肿瘤，尽管免疫治疗的出现提高了患者的生存率，但肿瘤微环境的抑制作用却减弱了免疫治疗的效果。因此需开发特定的免疫相关的预后模型从而提高患者的生存率和治疗策略。本研究结合癌症基因组图谱(The Cancer Genome Atlas，TCGA)、高通量基因表达(Gene Expression Omnibus，GEO)以及序列读取档案(Sequence Read Archive，SRA)数据库中皮肤黑色素瘤(skin cutaneous melanoma，SKCM)的相关数据，旨在建立一个基于差异表达的免疫相关预后预测模型，并通过风险评分评估肿瘤免疫微环境用以指导临床免疫治疗。方法: 从TCGA数据库获取SKCM转录组测序数据和对应的临床信息，分析其差异表达基因，并使用单因素Cox回归、LASSO方法以及逐步回归建立预后模型。采用实时反转录聚合酶链反应(real-time reverse transcription PCR，real-time RT-PCR)和蛋白质印迹法验证预后模型中基因的表达差异。使用Kaplan-Meier方法进行生存分析，通过时间依赖受试者操作特征曲线以及多因素Cox回归评价模型的效果，并使用GEO数据库中2个SKCM数据集对预后模型进行验证。进一步分析风险评分与免疫细胞浸润、基于表达估计恶性肿瘤组织的基质细胞和免疫细胞(Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data，ESTIMATE)评分、免疫检查点mRNA表达水平、肿瘤免疫周期及肿瘤免疫微环境通路间的相关性。最后在真实世界的队列中验证风险评分与免疫治疗效果及预后的相关性。结果: TCGA-SKCM数据集的差异分析发现差异表达基因主要与肿瘤免疫微环境相关。同时得到一个基于4个基因的预后模型，在模型中SKCM组织与正常皮肤组织的人细胞维甲酸结合蛋白2 (cellular retinoic acid binding protein 2，CRABP2)、白血病抑制因子(leukemia inhibitory factor，LIF)及杀伤细胞凝集素样受体D1(killer cell lectin like receptor D1，KLRD1)基因的mRNA和蛋白质水平差异均有统计学意义(均P<0.01)。高风险评分的患者总生存期显著短于低风险评分的患者(P<0.001)，在训练队列与多个验证队列取得了一致的结果(P<0.001)。预后模型与肿瘤免疫微环境的相关性分析结果显示高风险评分的SKCM患者处于抑制性的免疫微环境(P<0.01)。结论: 构建的免疫相关SKCM预后模型可以有效预测SKCM患者的预后；结合其评分与肿瘤免疫微环境的密切相关性，该模型对于SKCM临床免疫治疗的效果预测具有一定的参考价值。.

Keywords: bioinformatics; prognostic model; skin cutaneous melanoma; tumor immune microenvironment.

PubMed Disclaimer

Conflict of interest statement

作者声称无任何利益冲突。

Figures

图1
图1
TCGA-SCKM数据集DEGs筛选及GO分析和KEGG通路富集结果 Figure 1 Screening of DEGs in TCGA-SCKM dataset and the results of GO enrichment analysis and KEGG pathway enrichment A: Comparison of up-regulated and down-regulated genes in the TCGA-SKCM database between primary and metastatic melanoma; B: Analysis of differential genes using GO enrichment analysis; C: KEGG pathway enrichment analysis of DEGs. TCGA-SCKM:The Cancer Genome Atlas-Skin cutaneous melanoma; DEGs: Differentially-expressed genes; GO:Gene Ontology; KEGG: Kyoto Encyclopedia of Genes and Genomes; FC: Fold change; FDR: False discovery rate.
图2
图2
构建免疫相关预后模型 Figure 2 Construction of immune-related prognostic model A: Genetic interaction between immune-related and differential gene sets; B: Analysis of LASSO-modeled prognostic genes; C: Cross-validation of LASSO-modeled prognostic genes.
图3
图3
在皮肤黑色素瘤组织样本中验证预后模型相关基因的mRNA和蛋白水平表达 Figure 3 mRNA and protein expression of prognostic model-associated genes in skin cutaneous melanoma tissue samples A: Validation of mRNA and protein levels of LIF and KLRD1 in melanoma lesions and control (Ctrl) skin using real-time reverse transcription PCR and Western blotting (n=5, one sample excluded due to tissue degradation); B: Comparison of CRABP2 mRNA and protein levels for melanoma lesions and control skin (n=5, one sample excluded due to tissue degradation). Data are presented as mean±standard deviation, *P<0.05, **P<0.01, ***P<0.001.
图4
图4
训练队列TCGA-SCKM Figure 4 Training cohort of TCGA-SCKM A: Classification of high- and low-risk groups according to prognostic risk models for the training cohort; B: Survival analysis using prognostic hazard models for the training cohort of TCGA-SCKM; C: Analysis of TCGA-SCKM training cohort survival using prognostic risk models at 1, 3, 5, 10, and 15 years; D: Multivariate Cox analysis indicating that risk score is an independent predictor of training cohort survival. TCGA-SCKM: The Cancer Genome Atlas-skin cutaneous melanoma; AUC: Area under the curve.
图5
图5
验证队列GSE54467和GSE65904 Figure 5 Verification cohort of GSE54467 and GSE65904 A: With the prognostic hazard model, the Kaplan-Meier plot shows the survival probability of the validation cohort GSE544667, ROC curves are shown for survival analyses using 1, 3, 5, and 10-year time-dependent risk models in the validation cohort GSE544667, and prognostic risk models for high- and low-risk groups is validated in GSE544667. B: Kaplan-Meier plot shows the survival probability of the validation cohort GSE65904 using the prognostic risk model. ROC curves are shown for survival analyses using 1, 3, 5, and 10-year time-dependent risk models in the validation cohort GSE65904, and prognostic risk model for high- and low-risk groups is validated in GSE65904.
附图1
附图1. 风险评分与肿瘤免疫微环境的相关性Supplementary Figure 1 Correlation of risk score and tumor immune microenvironment
A: Comparison of 22 types of immune cells infiltration between the high- and low-risk groups of the TCGA-SCKM cohort; B: ESTIMATE algorithm comparison of immune infiltration scores in high- and low-risk groups; C: Relative expression levels of 20 immune checkpoint mRNAs in the high- and low-risk groups; D: Tumor microenvironment pathway scores and tumor immune cycle scores in the high- and low-risk groups.
附图2
附图2. 风险评分预测免疫治疗效果Supplementary Figure 2 Risk score in efficacy prediction of immunotherapy
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Bray F, Ferlay J, Soerjomataram I, et al. . Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin, 2018, 68(6): 394-424. 10.3322/caac.21492. - DOI - PubMed
    1. Spagnolo F, Boutros A, Tanda E, et al. . The adjuvant treatment revolution for high-risk melanoma patients[J]. Semin Cancer Biol, 2019, 59: 283-289. 10.1016/j.semcancer.2019.08.024. - DOI - PubMed
    1. Binnewies M, Roberts EW, Kersten K, et al. . Understanding the tumor immune microenvironment (TIME) for effective therapy[J]. Nat Med, 2018, 24(5): 541-550. 10.1038/s41591-018-0014-x. - DOI - PMC - PubMed
    1. De Jaeghere EA, Denys HG, De Wever O. Fibroblasts fuel immune escape in the tumor microenvironment[J]. Trends Cancer, 2019, 5(11): 704-723. 10.1016/j.trecan.2019.09.009. - DOI - PubMed
    1. Fortes C, Mastroeni S, Mannooranparampil TJ, et al. . Tumor-infiltrating lymphocytes predict cutaneous melanoma survival[J]. Melanoma Res, 2015, 25(4): 306-311. 10.1097/CMR.0000000000000164. - DOI - PubMed