{"paper_id":"0c1aad9d-effe-43f0-898d-0379250d799f","body_text":"A hypoxia-related five-lncRNA signature predicts osteosarcoma prognosis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article A hypoxia-related five-lncRNA signature predicts osteosarcoma prognosis Xin Wang, Qian Bai, Bo Xin, Yunheng Tai, Yong Cai, Kailiang Zhang, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-2945434/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Recently, several long-noncoding RNAs (lncRNAs) have been identified in hypoxia-associated cancer process including osteosarcoma (OS), enabling an adaptive survival under hypoxic stress conditions. However, hypoxia-related lncRNA signatures have rarely been reported. This study aimed to screen hypoxia-associated lncRNA signatures and assess their prognostic value in OS. Methods OS-related expression data were downloaded from the GEO and TARGET databases. Hypoxia-associated mRNAs were obtained from the HALLMARKHYPOXIA database. Hypoxia-associated lncRNAs were identified by correlation analysis with hypoxia-associated mRNAs. The tumor samples were clustered into different subtypes based on these lncRNAs, followed by immune microenvironment comparison. Prognostic hypoxia-associated lncRNAs were selected via univariate Cox regression analysis, and a prognostic signature was established using LASSO regression analysis. A risk score (RS) model was constructed, followed by pathway analysis, immunocorrelation analysis, and drug susceptibility prediction. Results Thirty hypoxia-related lncRNAs were selected. The OS samples were classified into two subtypes based on lncRNAs. Nine immune cell types showed significantly different levels of infiltration between the two subtypes. Furthermore, five prognostic hypoxia-related lncRNAs were screened out through LASSO regression analyses, and an RS model was constructed. The high- and low-risk groups showed differences in prognosis, pathway, and drug susceptibility. The present study divided OS into two subtypes. A prognostic signature was constructed based on five hypoxia-related lncRNAs. Conclusions The study sucessfully identifies five hypoxia-related lncRNAs and this lncRNA signature may have significant prognostic value in OS. Osteosarcoma Long-noncoding RNA Hypoxia Prognosis Biomarker Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction Osteosarcoma (OS) is the most common primary malignant tumor of bones, which usually occurs in the metaphysis of the long bones and is common in the distal femur and proximal humerus ( 1 , 2 ). With the development of multimodal therapy, the 5-year survival rate for patients has improved to > 70% ( 3 ). However, many patients develop metastasis at initial diagnosis or following intensive therapy, and tumor recurrence is a common cause of death for patients ( 4 , 5 ). Therefore, elucidating the molecular mechanisms underlying the development of OS is of great clinical significance for improving treatment strategies. Hypoxia is a hallmark of solid tumors in humans ( 6 ). Similar to normal cells, tumor cells can activate signaling pathways to induce angiogenesis, proliferation, and apoptosis ( 7 , 8 ). However, among the pathways induced in tumor cells, those promoting tumor growth are more numerous than those promoting apoptosis. Furthermore, hypoxia may influence a more aggressive phenotype of tumor cells related to a poor prognosis in patients ( 9 ). A recent study reported that hypoxia could influence the tumor microenvironment by promoting the expression of adverse immune cytokines, increasing the differentiation potential of immunosuppressor cells, and decreasing the activity of natural killer cells and cytotoxic T lymphocytes ( 10 ). These factors ultimately result in chemotherapy resistance, metastasis, and poor patient outcomes ( 11 ). In OS, several hypoxia-related prognostic gene signatures have been reported ( 12 , 13 ). In recent years, studies have gradually revealed the response of long-noncoding RNAs (lncRNAs) to hypoxia and their regulatory role in human cancers ( 14 , 15 ). However, hypoxia-related lncRNA signatures have rarely been reported. This study aimed to screen hypoxia-associated lncRNAs in patients with OS based on the GEO and TARGET databases. Hypoxia-associated lncRNAs were identified by correlation analysis with hypoxia-associated mRNAs. Tumor samples were clustered into different subtypes based on lncRNAs. Prognostic hypoxia-associated lncRNAs were screened using univariate Cox regression analysis, and a prognostic signature was established using LASSO regression analysis. The lncRNA signature obtained may serve as a promising prognostic biomarker for OS. Methods Acquisition and preprocessing of expression profile data The GSE16088 dataset was downloaded from the GEO database. The gene expression profiles of six normal tissue samples and 14 OS tissue samples were included (detection platform: GPL96 [HG-U133A] Affymetrix Human Genome U133A Array). This dataset was used to screen for differentially expressed (DE) hypoxia-related genes. The preprocessed, standardized and log2 transformed probe expression matrix was downloaded. An annotation file was downloaded. Through one-to-one matching of gene symbols and probe numbers, probes without matching gene symbols were removed. The mean value of the different probes was used as the final expression value. The RNA-seq data (log2 (FPKM + 1)) of OS were obtained from the TARGET database from the UCSC-Xena platform ( 16 ). The clinical information of the corresponding samples, including age, sex, ethnicity, disease at diagnosis, specific tumor region, primary tumor site, definitive surgery, primary site progression, and survival information, were also downloaded. A total of 88 OS samples were included, of which 85 had prognostic information. Furthermore, mRNAs and lncRNAs were re-annotated according to the annotation file related to GENCODE V22. The gene with the annotation information of “protein_coding” was retained as the mRNA, and that with the annotation information of “antisense,” “sense_intronic,” “lincRNA,” “sense_overlapping,” “processed_transcript,” “3prime_overlapping_ncRNA,” or “non_coding” was retained as the lncRNA. The expression matrices of the mRNAs and lncRNAs in 88 samples were obtained for follow-up analysis. DE hypoxia-related gene screening Based on the GSE16088 dataset, linear regression and empirical Bayesian methods in limma 3.10.3 ( 17 ) were used to analyze the DE genes of tumor vs. normal controls. The Benjamini & Hochberg method was adopted for multiple corrections, and adjusted p < 0.05 and |logFC| > 1 were used as thresholds. Volcano mapping was performed to identify the DE genes. The pathway genes of HALLMARK_HYPOXIA were downloaded from MSigDB v7.1 ( 18 ) as hypoxia-related genes. The intersection of DE and hypoxia-related genes was used for subsequent analyses. Identification of hypoxia-related lncRNAs First, mRNAs obtained from the TARGET database were matched with DE hypoxia-related genes to obtain the expression values of DE hypoxia-related mRNAs in the 88 samples. Pearson’s correlation coefficients between all lncRNAs and DE hypoxia-related mRNAs were calculated. Pairs with p < 0.001 and |Pearson correlation coefficient| < 0.4 were selected. Furthermore, lncRNAs with expression values > 0 in > 80% of the samples and correlated with > 5 DE hypoxia-related mRNAs were selected. The corresponding lncRNA-mRNA network was visualized using Cytoscape 3.9.0 ( 19 ). lncRNAs in the network were considered hypoxia-related lncRNAs. Prediction of molecular subtypes of OS With the expression values of hypoxia-related lncRNAs in various TARGET dataset samples, ConsensusClusterPlus 1.54.0 ( 20 ) in R3.6.1 was used to analyze hypoxia-related molecular subtypes of all samples. Hypoxic-related score Based on the expression values of all mRNAs in each sample of the TARGET dataset, the hypoxia-related enrichment score was calculated using the ssGSEA algorithm of GSVA 1.36.2 ( 21 ) with all hypoxia-related genes as the enrichment background. Then, combined with the subtype information of the samples, the Wilcox test was used to compare the difference in significance of hypoxia-related scores among subtypes, and a violin plot was drawn. Correlation analysis between prognosis and clinical outcome of subtypes Based on the obtained hypoxia-related subtypes combined with the prognostic information of the samples, the KM survival curves of the subtypes were drawn, and the significance p-value was calculated using the log-rank test. Additionally, the clinical information of the subtypes was compared. For factor-type variables, the Chi-squared test was used to conduct statistical analysis, and for continuous variables, the Wilcox test was used. Comparison of immune microenvironments between subtypes In order to observe the differences in the immune microenvironment among subtypes, the CIBERSORT ( 22 ) algorithm was used to calculate the proportions of 22 types of immune cells according to the gene expression levels of the OS samples. The Wilcoxon test was used to identify the p-value among the subtypes. The R-pack MCPcounter ( 23 ) algorithm was used to compare the infiltration levels of eight immune cell and two stromal cell types based on the gene expression levels of the OS samples. Combined with the subtype grouping of samples, the Wilcox test was used to calculate the differential p-values among subtypes. In addition, the ESTIMATE ( 24 ) algorithm was used to estimate the stromal, immune, and ESTIMATE scores of tumor samples according to the expression data. The Wilcoxon test was used to calculate the differences among the subtypes. Furthermore, the expression data of HLA family genes and immune checkpoint genes in each OS sample were extracted for the comparison of expression differences among subtypes by the t-test, and a box diagram was drawn. Identification of prognostic hypoxia-related lncRNAs In TARGET-OS samples, based on the hypoxia-related lncRNAs obtained above, combined with the clinical survival and prognosis information, hypoxia-related lncRNAs significantly correlated with overall survival prognosis were identified using univariate Cox regression analysis in survival 2.41-1 ( 25 ). The threshold was set at p < 0.05. Construction of the prognostic hypoxia-related lncRNA signature Based on the hypoxia-related lncRNAs significantly correlated with survival prognosis, the LASSO Cox regression model ( 26 ) was also applied to further screen the lncRNA combination by combining the survival prognosis information of TARGET-OS and expression value of lncRNAs in each sample using glmnet 2.0–18 ( 27 ) in R3.6.1. Then, according to the regression coefficient of lncRNA in hypoxia-related lncRNA combinations and the expression level of lncRNA in TARGET-OS samples, the following risk score (RS) model was established: RS = ∑β lncRNA × Exp lncRNA ( 1 ) where β lncRNA represents the LASSO regression coefficient of lncRNA and Exp lncRNA represents the expression level of lncRNA in the TARGET-OS dataset. According to the RS median value, all samples were divided into high-risk (≥ RS median value) and low-risk (< RS median value) groups. The association between risk grouping and actual survival prognosis was assessed using the Kaplan–Meier (KM) curve method in Survival 2.41-1 of R3.6.1. The prediction ROC curves for 1, 3, and 5 years were drawn. Analysis of independent prognostic factors To determine whether the RS model was an independent prognostic factor, age, sex, ethnicity, disease at diagnosis, specific tumor region, primary tumor site, definitive surgery, primary site progression, and RS were included in univariate Cox regression analysis, followed by multivariate Cox regression analysis. The independent prognostic factors were used to establish a nomogram. Meanwhile, the calibration and ROC curves of the nomogram were drawn to verify its validity. Analysis of molecular mechanisms between different risk groups To determine whether each hallmark gene set was significantly enriched in a particular risk group, with h.all.v7.4. symbols in the MSigDB v7.1 database as the enrichment background, GSEA of high- vs. low-risk was performed using R clusterProfiler 3.8.1 ( 28 ). The expression values of all mRNAs in TARGET-OS samples and high- and low-risk group sample information were included. Statistical significance was set at p < 0.05. Correlation analysis of model lncRNAs and immune cells Based on the differential immune cells between subtypes obtained in the previous analysis, the Spearman correlation coefficients and the corresponding significance p-value between model lncRNAs and immune cells were calculated, and a correlation heatmap was drawn. Drug sensitivity analysis Drug sensitivity was reflected by the IC50 values. IC50 can be predicted by the ridge regression model according to the GDSC cell line expression profile and TARGET-OS mRNA expression profile using the pRRophetic algorithm. Here, pRRophetic 0.5 ( 29 ) was used to predict the IC50 values based on the 138 drugs provided in this package, and the Wilcox test was used to analyze whether there were significant differences in the IC50 values of each drug between the risk groups. Prediction of the response to immunotherapy The immunotherapy response was determined using tumor immune dysfunction and elimination (TIDE) analysis. TIDE uses two main types of tumor immune escape mechanisms to predict the immune response to treatment. The TIDE values of all samples were calculated using TIDE tools ( 30 ), and the Spearman correlation between RS and TIDE scores was calculated. Correlation analysis of immune indexes The cytolytic activity (CYT) scoreis an established biomarker for predicting the presence of tumor-infiltrating lymphocytes based on gene expression. The CYT score of each sample in the high- and low-risk groups was calculated according to the CYT score calculation method of a previous study ( 31 ). Additionally, based on the expression values of CD8A and PD-L1 in each tumor sample, the CD8A/PD-L1 ratio was calculated, and the difference between the two groups was calculated using the Wilcox test. A recent study obtained a nine-gene signature of the tertiary lymphoid structure (TLS) in melanoma ( 32 ). Based on the expression values of the nine genes in each tumor sample, here, the ssGSEA algorithm was used to estimate the TLS score of each sample, and the Wilcox test was used for difference comparison. Finally, the distribution proportion of subtypes in the high- and low-risk groups was calculated, and significance was calculated using the Chi-squared test. Results DE hypoxia-related mRNA and lncRNA identification A total of 2,424 upregulated and 2,172 downregulated mRNAs were obtained from the GSE16088 dataset (Fig. 1 a). The intersection of DE mRNAs and hypoxia-related genes was further examined, and 72 DE hypoxia-related mRNAs were obtained (Fig. 1 b). Pearson correlation analysis obtained 5,981 lncRNA-mRNA correlation relationship pairs, including 2,802 lncRNAs and 69 mRNAs, below thresholds of p < 0.001 and |Pearson correlation coefficient| > 0.4, and lncRNAs with expression values > 0 in > 80% of samples. Due to the large number of lncRNAs, only those correlated with > 5 DE hypoxia-related mRNAs were selected. Finally, 203 lncRNA-mRNA relationship pairs, including 30 lncRNAs and 60 mRNAs, were retained. The network diagram is shown in Fig. 1 C. Molecular subtype analysis Based on the 30 hypoxia-related lncRNAs obtained above and their expression values in osteosarcoma samples, consistent clustering was conducted. As shown in Fig. 2A, two clusters were finally determined. The hypoxia-related scores of the samples in cluster 2 were significantly higher than those in cluster 1 (p < 0.05) (Fig. 2b). Prognosis and clinical correlation analysis of subtypes Comparison of clinical information between subtypes revealed that primary site progression was significantly different between the two subtypes (p < 0.05) (Table 1 ). Table 1 Clinical information between subtypes. Characteristics Cluster 1 (N = 45) Cluster 2 (N = 42) Total (N = 87) p-value Age 0.44 Mean ± SD 14.77 ± 5.15 15.38 ± 4.36 15.06 ± 4.77 Median[min-max] 14.05 [3.56,32.41] 15.08 [5.96,29.14] 14.49 [3.56,32.41] Gender 0.52 Female 21 (24.14%) 16 (18.39%) 37 (42.53%) Male 24 (27.59%) 26 (29.89%) 50 (57.47%) Ethnicity 1 Hispanic or Latino 6 (9.23%) 5 (7.69%) 11 (16.92%) Not Hispanic or Latino 29 (44.62%) 25 (38.46%) 54 (83.08%) Disease at diagnosis 0.81 Metastatic 12 (13.79%) 10 (11.49%) 22 (25.29%) Non-metastatic 33 (37.93%) 32 (36.78%) 65 (74.71%) Primary tumor site 1 Arm/hand 3 (3.53%) 3 (3.53%) 6 (7.06%) Leg/foot 41 (48.24%) 38 (44.71%) 79 (92.94%) Specific tumor region 0.09 Distal 17 (33.33%) 11 (21.57%) 28 (54.90%) Proximal 8 (15.69%) 15 (29.41%) 23 (45.10%) Definitive surgery 1 Amputation 2 (4.44%) 3 (6.67%) 5 (11.11%) Limb sparing 19 (42.22%) 21 (46.67%) 40 (88.89%) Primary site progression 0.02 No 11 (28.95%) 12 (31.58%) 23 (60.53%) Yes 13 (34.21%) 2 (5.26%) 15 (39.47%) KM survival curves of the two subtypes are plotted in Fig. 3a. The prognosis of cluster 1 was worse than that of cluster 2 (p < 0.05). A heatmap of the expression of 30 hypoxia-related lncRNAs in each sample was drawn. The expression patterns of the 30 hypoxia-related lncRNAs differed between the two subtypes (Fig. 3B). Comparison of immune microenvironments between subtypes Based on the CIBERSORT algorithm, the immune cell infiltration levels of six immune cells, including CD4 naïve T cells, gamma delta T cells, M0 macrophages, M1 macrophages, M2 macrophages, and resting dendritic cells, were significantly different between the two subtypes (p < 0.05) (Fig. 4a). Based on the MCPCounter algorithm, neutrophils and two stromal cell types (endothelial cells and fibroblasts) were significantly different between the two groups (p < 0.05) (Fig. 4b). The ESTIMATE algorithm revealed that the stromal score in cluster 2 was significantly higher than that in cluster 1 (p < 0.05), and there was no significant difference between the two subtypes of immune and microenvironment scores (ESTIMATE score) (Fig. 4C). Three HLA family genes (HLA-A, HLA-E, and HLA-F) showed significant differences between the two subtypes (p < 0.05) (Fig. 4d). Two immune checkpoint genes ( PDCD1LG2 and TNFRSF4 ) showed significant differences between the two subtypes (p < 0.05) (Fig. 4e). Identification of the prognostic hypoxia-related lncRNA signature Based on the 30 hypoxia-related lncRNAs, 10 were selected using univariate Cox regression analysis (Fig. 5a). Further, with the expression values of 10 prognostic hypoxia-related lncRNAs in OS samples, and the survival time and state of the samples, five optimized lncRNA signatures (CTB-4E7.1, MIR210HG, RP11-817J15.3, RP11-81A22.5, and RPARP-AS1) were screened by the LASSO Cox regression algorithm. Prognostic regression coefficients are shown in Fig. 5B. Based on the LASSO regression prognostic coefficients of the five optimized lncRNAs and their expression levels in the TARGET-OS dataset, the RS model was constructed. The RS visualization is shown in Fig. 5c. The association between risk grouping and actual disease prognostic information was evaluated using KM curves. As shown in Fig. 5d, the prognosis in the low-risk group was significantly better than that in the high-risk group. The 1-, 3-, and 5-year survival prediction ROC curves are shown in Fig. 5e. The results indicated a significant correlation between the risk groups predicted by the RS model and actual prognosis. Analysis of the independent prognostic factor Following univariate Cox regression analysis, the disease at diagnosis, primary site progression, and RS were screened (Fig. 6A). After multivariate Cox regression analysis, the disease at diagnosis and RS were considered independent prognostic factors (Fig. 6B). A nomogram was constructed for independent prognostic factors, as shown in Fig. 6C. The survival time of the samples was predicted based on the “Total Points” axis in the first line. The 1-, 3-, and 5-year survival rates predicted by the model were consistent with the actual 1-, 3-, and 5-year survival rates (Fig. 6d). The AUCs of the 1-, 3-, and 5-year ROC curves were 0.889, 0.852, and 0.857, respectively (Fig. 6e). Analysis of related molecular mechanisms between risk groups Enrichment analysis of hallmark gene sets was performed for the high- and low-risk groups. A total of 13 upregulated pathways, such as cholesterol homeostasis, hypoxia pathway, and oxidative phosphorylation, and 12 downregulated pathways, such as complement, epithelial mesenchymal transition, and inflammatory response, were identified (Fig. 7 ). Correlation analysis of model lncRNAs and immune cells Spearman correlation coefficients between the model lncRNAs and the differential immune cells between the subtypes were calculated. As shown in Fig. 8A, fibroblasts had the highest negative correlation with RP11-81A22.5, endothelial cells had the highest positive correlation with CTB-4E7.1, and M1 macrophages had the highest positive correlation with RP11-817J15.3. Drug sensitivity analysis between high- and low-risk groups A total of 39 drugs had significantly different IC50 values between the high- and low-risk groups. The IC50 value distribution box plot of the common drugs elesclomol, midostaurin, and vorinostat is shown in Fig. 8b. Prediction of the response to immunotherapy The scatterplot for the correlation between the TIDE score and RS is shown in Fig. 8c, and there was a significantly negative correlation between them; the higher the RS, the lower the TIDE score. Correlation analysis of immune indexes The CD8A/PD-L1 ratio in the high-risk group was significantly higher than that in the low-risk group, whereas the CYT and TLS scores were not significantly different between the two groups (Fig. 8d). Association analysis of the hypoxia subtype and risk grouping The distribution of the two subtypes in the high- and low-risk groups is shown in Fig. 8e. The high-risk group contained more samples from cluster 1. Discussion Tumor hypoxia is a common feature of many cancers, and the activation of hypoxia pathways is correlated with a poor prognosis in many cancer types. Recent evidence suggests the involvement of hypoxia-regulated lncRNAs in cancer cells ( 33 ). In the present study, 30 hypoxia-related lncRNAs were identified. OS samples were classified into two subtypes based on the lncRNAs. Furthermore, five prognostic hypoxia-related lncRNAs were screened through LASSO regression analyses, and an RS model was constructed. The samples in the two risk groups presented differences in prognosis, pathways, and the immune microenvironment. Growing evidence suggests that OS is heterogeneous, and some subtypes have been reported based on genomic analysis ( 34 , 35 ). In this study, OS was further classified into two subtypes based on 30 hypoxia-related lncRNAs. The prognosis of cluster 1 (low hypoxia score) was worse than that of cluster 2 (high hypoxia score). M2 macrophages had a higher infiltration level among the 22 immune cell types, and cluster 1 had a higher infiltration level of M2 macrophages than cluster 2. In tumors, M1 macrophages are regarded as antitumor effectors, and M2 macrophages as protumor modulators owing to their induction of neovascularization ( 36 ). The density of tumor-associated macrophages is correlated with invasion, metastasis, and a poor prognosis in some cancers, including OS ( 37 ). Thus, the higher infiltration level of M2 macrophages in cluster 1 was consistent with the worse prognosis of the samples in this cluster. Two immune checkpoint genes ( PDCD1LG2 and TNFRSF4 ) showed significant differences between the two subtypes. Here, it was speculated that PDCD1LG2 and TNFRSF4 may serve as biomarkers to predict the response of patients with OS to immune checkpoint blocking therapy. Five prognosis-related lncRNAs were selected from 30 hypoxia-related lncRNAs, and five prognosis-related ones were further selected: CTB-4E7.1, MIR210HG, RP11-817J15.3, RP11-81A22.5, and RPARP-AS1. CTB-4E7.1 has been demonstrated to be one of the prognostic risk signatures for OS ( 38 ). MIR210HG has been reported to promote tumor cell proliferation, invasion, and metastasis in many cancers, including OS, indicating a poor prognosis ( 39 – 41 ). RP1-261G23.7, RP11-81A22.5, and RPARP-AS1 have also been found to be DE in OS in recent studies ( 42 , 43 ). The present study is the first to provide a risk model for OS with f5 lncRNAs with prognostic relevance linked to hypoxia. After the prognostic RS model for OS was constructed, subsequent KM analysis and ROC curves further indicated the predictive value of the RS model. The high-risk group was associated with a poor prognosis. The AUC of the 5-year survival prediction ROC curve was 0.746, which was similar to other signatures of OS ( 38 , 44 ). Moreover, RS was an independent prognostic factor associated with disease at diagnosis. The nomogram constructed based on the two independent prognostic factors could accurately predict the 1-, 3-, and 5-year survival rates of patients with OS. These results suggested that this prognostic model has the potential to evaluate the prognosis of patients with OS. To further investigate the molecular mechanisms between the two risk groups, GSEA was conducted to analyze the pathways between the two groups. The hypoxia pathway was among the upregulated pathways, and epithelial-mesenchymal transition was one of the downregulated pathways. Epithelial-mesenchymal transition is associated with tumor proliferation, vascularization, and metastasis ( 45 ). Both epithelial mesenchymal transition and hypoxia are regarded as separate events that promote the invasion and metastasis of many cancer types ( 46 ). Additionally, the correlation between the risk groups and immunity was analyzed. Recent studies have examined the mechanisms of tumor immune evasion ( 47 , 48 ). Some cancers have high infiltration levels of cytotoxic T cells; however, these T cells are often in a dysfunctional state. In other cancers, immunosuppressive factors may exclude T cells from infiltrating tumors ( 49 ). TIDE was developed by Peng et al. ( 30 ) to identify the factors underlying tumor immune escape mechanisms. In this study, there was a significant negative correlation between the TIDE score and RS, indicating different responses to immunotherapy in the two risk groups. Moreover, the high-risk group contained more samples from cluster 1, which had a poor prognosis. Conclusions In conclusion, the present study divided OS into two subtypes based on 30 hypoxia-related lncRNAs. Further analyses identified five hypoxia-related lncRNAs and constructed a prognostic signature. The model lncRNAs may be involved in the progression of OS via the epithelial-mesenchymal transition and hypoxia pathways. Further experiments are required to confirm these findings. Abbreviations LncRNA Long-noncoding RNA OS Osteosarcoma RS Risk score DE differentially expressed KM Kaplan–Meier TIDE Tumor immune dysfunction and elimination CYT Cytolytic activity TLS Tertiary lymphoid structure Declarations Acknowledgments Not applicable. Authors’ contributions All authors contributed to this paper. YZ and KZ devised the study. XW and QB performed experiments and conducted data analysis. XW, BX, YT and YC conducted data preparation. YZ, KZ, XW and QB participated in authoring the manuscript. All authors have read and approved the manuscript. Funding This study was supported by the National Natural Science Foundation of China (82173249). Availability of data and materials All data generated or analysed during this study are included in this published article. Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Conflicts of Interest The authors declare that they have no competing interests. References Huang C-Y, Wei P-L, Wang J-W, Makondi PT, Huang M-T, Chen H-A, et al. Glucose-regulated protein 94 modulates the response of osteosarcoma to chemotherapy. Disease Markers. 2019;2019. Lin J, Wang X, Wang X, Wang S, Shen R, Yang Y, et al. Hypoxia increases the expression of stem cell markers in human osteosarcoma cells. Oncology Letters. 2021;21(3):1. Anderson ME. Update on survival in osteosarcoma. Orthopedic Clinics. 2016;47(1):283-92. Zhang Z, Luo G, Yu C, Yu G, Jiang R, Shi X. Retracted: MicroRNA‐493‐5p inhibits proliferation and metastasis of osteosarcoma cells by targeting Kruppel‐like factor 5. Journal of cellular physiology. 2019;234(8):13525-33. Kansara M, Teng MW, Smyth MJ, Thomas DM. Translational biology of osteosarcoma. Nature Reviews Cancer. 2014;14(11):722-35. Ruan K, Song G, Ouyang G. Role of hypoxia in the hallmarks of human cancer. Journal of cellular biochemistry. 2009;107(6):1053-62. Iyer NV, Leung SW, Semenza GL. The human hypoxia-inducible factor 1α gene: Hif1astructure and evolutionary conservation. Genomics. 1998;52(2):159-65. Guo Z, Wang X, Yang Y, Chen W, Zhang K, Teng B, et al. Hypoxic tumor-derived exosomal long noncoding RNA UCA1 promotes angiogenesis via miR-96-5p/AMOTL2 in pancreatic cancer. Molecular Therapy-Nucleic Acids. 2020;22:179-95. Vaupel P, Kallinowski F, Okunieff P. Blood flow, oxygen and nutrient supply, and metabolic microenvironment of human tumors: a review. Cancer research. 1989;49(23):6449-65. Terry S, Buart S, Chouaib S. Hypoxic stress-induced tumor and immune plasticity, suppression, and impact on tumor heterogeneity. Frontiers in immunology. 2017;8:1625. Walsh JC, Lebedev A, Aten E, Madsen K, Marciano L, Kolb HC. The clinical importance of assessing tumor hypoxia: relationship of tumor hypoxia to prognosis and therapeutic opportunities. Antioxidants & redox signaling. 2014;21(10):1516-54. Fu Y, Bao Q, Liu Z, He G, Wen J, Liu Q, et al. Development and validation of a hypoxia-associated prognostic signature related to osteosarcoma metastasis and immune infiltration. Frontiers in Cell and Developmental Biology. 2021;9:633607. Jiang F, Miao X-L, Zhang X-T, Yan F, Mao Y, Wu C-Y, et al. A hypoxia gene-based signature to predict the survival and affect the tumor immune microenvironment of osteosarcoma in children. Journal of Immunology Research. 2021;2021. Dong H, Hu J, Zou K, Ye M, Chen Y, Wu C, et al. Activation of LncRNA TINCR by H3K27 acetylation promotes Trastuzumab resistance and epithelial-mesenchymal transition by targeting MicroRNA-125b in breast Cancer. Molecular cancer. 2019;18:1-18. Han Y, Wang X, Mao E, Shen B, Huang L. Analysis of differentially expressed lncRNAs and mRNAs for the identification of hypoxia-regulated angiogenic genes in colorectal cancer by RNA-seq. Medical Science Monitor: International Medical Journal of Experimental and Clinical Research. 2019;25:2009. Goldman M, Craft B, Hastie M, Repečka K, McDade F, Kamath A, et al. The UCSC Xena platform for public and private cancer genomics data visualization and interpretation. biorxiv. 2018:326470. Smyth GK. Limma: linear models for microarray data. Bioinformatics and computational biology solutions using R and Bioconductor. 2005:397-420. Liberzon A, Subramanian A, Pinchback R, Thorvaldsdóttir H, Tamayo P, Mesirov JP. Molecular signatures database (MSigDB) 3.0. Bioinformatics. 2011;27(12):1739-40. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome research. 2003;13(11):2498-504. Wilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26(12):1572-3. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC bioinformatics. 2013;14:1-15. Kawada J-i, Takeuchi S, Imai H, Okumura T, Horiba K, Suzuki T, et al. Immune cell infiltration landscapes in pediatric acute myocarditis analyzed by CIBERSORT. Journal of Cardiology. 2021;77(2):174-8. Becht E, Giraldo NA, Lacroix L, Buttard B, Elarouci N, Petitprez F, et al. Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome biology. 2016;17(1):1-20. Yoshihara K, Shahmoradgoli M, Martínez E, Vegesna R, Kim H, Torres-Garcia W, et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nature communications. 2013;4(1):2612. Wang P, Wang Y, Hang B, Zou X, Mao J-H. A novel gene expression-based prognostic scoring system to predict survival in gastric cancer. Oncotarget. 2016;7(34):55343. Tibshirani R. The lasso method for variable selection in the Cox model. Statistics in medicine. 1997;16(4):385-95. Friedman J, Hastie T, Tibshirani R, Narasimhan B, Tay K, Simon N. glmnet: Lasso and elastic-net regularized generalized linear models. R package version. 2009;1(4):1-24. Yu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics: a journal of integrative biology. 2012;16(5):284-7. Geeleher P, Cox N, Huang RS. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PloS one. 2014;9(9):e107468. Jiang P, Gu S, Pan D, Fu J, Sahu A, Hu X, et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nature medicine. 2018;24(10):1550-8. Rooney MS, Shukla SA, Wu CJ, Getz G, Hacohen N. Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell. 2015;160(1-2):48-61. Cabrita R, Lauss M, Sanna A, Donia M, Skaarup Larsen M, Mitra S, et al. Tertiary lymphoid structures improve immunotherapy and survival in melanoma. Nature. 2020;577(7791):561-5. Chang Y-N, Zhang K, Hu Z-M, Qi H-X, Shi Z-M, Han X-H, et al. Hypoxia-regulated lncRNAs in cancer. Gene. 2016;575(1):1-8. Wang X, Wang L, Xu W, Wang X, Ke D, Lin J, et al. Classification of osteosarcoma based on immunogenomic profiling. Frontiers in Cell and Developmental Biology. 2021;9:696878. Kubista B, Klinglmueller F, Bilban M, Pfeiffer M, Lass R, Giurea A, et al. Microarray analysis identifies distinct gene expression profiles associated with histological subtype in human osteosarcoma. International orthopaedics. 2011;35:401-11. Qian B-Z, Pollard JW. Macrophage diversity enhances tumor progression and metastasis. Cell. 2010;141(1):39-51. Cersosimo F, Lonardi S, Bernardini G, Telfer B, Mandelli GE, Santucci A, et al. Tumor-associated macrophages in osteosarcoma: from mechanisms to therapy. International Journal of Molecular Sciences. 2020;21(15):5207. Liu H, Chen C, Liu L, Wang Z. A four-lncRNA risk signature for prognostic prediction of osteosarcoma. Frontiers in genetics. 2023;13:1081478. Wang A-H, Jin C-H, Cui G-Y, Li H-Y, Wang Y, Yu J-J, et al. MIR210HG promotes cell proliferation and invasion by regulating miR-503-5p/TRAF4 axis in cervical cancer. Aging (Albany NY). 2020;12(4):3205. Li Z-Y, Xie Y, Deng M, Zhu L, Wu X, Li G, et al. c-Myc-activated intronic miR-210 and lncRNA MIR210HG synergistically promote the metastasis of gastric cancer. Cancer Letters. 2022;526:322-34. Li J, Wu Q-M, Wang X-Q, Zhang C-Q. Long noncoding RNA miR210HG sponges miR-503 to facilitate osteosarcoma cell invasion and metastasis. DNA and cell biology. 2017;36(12):1117-25. Ying T, Dong J-l, Yuan C, Li P, Guo Q. The lncRNAs RP1-261G23. 7, RP11-69E11. 4 and SATB2-AS1 are a novel clinical signature for predicting recurrent osteosarcoma. Bioscience Reports. 2020;40(1). Bu X, Liu J, Ding R, Li Z. Prognostic value of a pyroptosis-related long noncoding RNA signature associated with osteosarcoma microenvironment. Journal of Oncology. 2021;2021. Yu S, Shao F, Liu H, Liu Q. A five metastasis-related long noncoding RNA risk signature for osteosarcoma survival prediction. BMC Medical Genomics. 2021;14(1):124. Nantajit D, Lin D, Li JJ. The network of epithelial–mesenchymal transition: potential new targets for tumor resistance. Journal of cancer research and clinical oncology. 2015;141:1697-713. Tam SY, Wu VW, Law HK. Hypoxia-induced epithelial-mesenchymal transition in cancers: HIF-1α and beyond. Frontiers in oncology. 2020;10:486. Gajewski TF, Schreiber H, Fu Y-X. Innate and adaptive immune cells in the tumor microenvironment. Nature immunology. 2013;14(10):1014-22. Joyce JA, Fearon DT. T cell exclusion, immune privilege, and the tumor microenvironment. Science. 2015;348(6230):74-80. Spranger S, Gajewski TF. Tumor-intrinsic oncogene pathways mediating immune avoidance. Oncoimmunology. 2016;5(3):e1086862. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-2945434\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":201302679,\"identity\":\"2e5b0b8b-f047-4cf4-99ef-47a0164bfb31\",\"order_by\":0,\"name\":\"Xin Wang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Lintong Rehabilitation and Convalescent Centre of the PLA Joint Logistics Support Force\",\"correspondingAuthor\":false,\"submittingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Xin\",\"middleName\":\"\",\"lastName\":\"Wang\",\"suffix\":\"\"},{\"id\":201302680,\"identity\":\"ff8da276-77b3-417e-a01c-830a91894115\",\"order_by\":1,\"name\":\"Qian Bai\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Army Hospital of Chinese PLA\",\"correspondingAuthor\":false,\"submittingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Qian\",\"middleName\":\"\",\"lastName\":\"Bai\",\"suffix\":\"\"},{\"id\":201302681,\"identity\":\"140dbb43-2e52-46fd-9063-021480649597\",\"order_by\":2,\"name\":\"Bo Xin\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"the 960th Hospital of the PLA Joint Logistics Support Force\",\"correspondingAuthor\":false,\"submittingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Bo\",\"middleName\":\"\",\"lastName\":\"Xin\",\"suffix\":\"\"},{\"id\":201302682,\"identity\":\"767bcdbb-71f9-444b-bfb5-23f038595dde\",\"order_by\":3,\"name\":\"Yunheng Tai\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"the 960th Hospital of the PLA Joint Logistics Support Force\",\"correspondingAuthor\":false,\"submittingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yunheng\",\"middleName\":\"\",\"lastName\":\"Tai\",\"suffix\":\"\"},{\"id\":201302683,\"identity\":\"a94dcc4f-da69-4082-95e1-8906fdda6fdc\",\"order_by\":4,\"name\":\"Yong Cai\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"the 960th Hospital of the PLA Joint Logistics Support Force\",\"correspondingAuthor\":false,\"submittingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yong\",\"middleName\":\"\",\"lastName\":\"Cai\",\"suffix\":\"\"},{\"id\":201302684,\"identity\":\"cbe8a48e-01af-4021-87ec-838e383dda8b\",\"order_by\":5,\"name\":\"Kailiang Zhang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"the 960th Hospital of the PLA Joint Logistics Support Force\",\"correspondingAuthor\":false,\"submittingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Kailiang\",\"middleName\":\"\",\"lastName\":\"Zhang\",\"suffix\":\"\"},{\"id\":201302685,\"identity\":\"2a2822c4-5e42-4d71-bcd3-270313423013\",\"order_by\":6,\"name\":\"Yong Zhou\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAx0lEQVRIiWNgGAWjYFACNhBhw4/EIU5LmmQDqVoOk6BFPiIt8eHPtvMSutPOGDB8KDvMwD+7Ab8Wwxtphw0k225LmN3OMWCcce4wg8SdAwS0zEhvkzBsu10H0sLM23aYwUAigaCW9h+JbefAtjD/JUaLvETaMYaDbQcgWhiJ0WLA8yxZsuFcMlBLWsHBnnPpPBI3CNnSnmb48UeZHVBL8sYHP8qs5fhnELLlAJBghEYHiM2DXz3IlgYQ+YegulEwCkbBKBjJAACM5kP68V55DgAAAABJRU5ErkJggg==\",\"orcid\":\"\",\"institution\":\"Orthopedic Oncology Institute of Chinese PLA, Tangdu Hospital, Air Force Medical University\",\"correspondingAuthor\":true,\"submittingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yong\",\"middleName\":\"\",\"lastName\":\"Zhou\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2023-05-17 05:14:13\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-2945434/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-2945434/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":37247232,\"identity\":\"e7c13617-6447-4626-86e7-88aa25f29bf4\",\"added_by\":\"auto\",\"created_at\":\"2023-05-19 15:06:50\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":395252,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e(a) Volcano maps of differentially expressed genes. (b) Venn diagram of significantly differentially expressed genes and hypoxia-related genes. (c) Hypoxia-related lncRNA-mRNA co-expression network (the blue circle represents downregulated hypoxia-related mRNA, the yellow circle represents upregulated hypoxia-related mRNA, the red diamond represents hypoxia-related lncRNA, the red dotted line represents a positive correlation between lncRNA and mRNA, and the green dotted line represents a negative correlation between lncRNA and mRNA. The node size represents the connection size).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-2945434/v1/a7ce11f53dc4e4df69aba384.png\"},{\"id\":37245838,\"identity\":\"9e57673b-a94a-480a-99a1-77ba6e349508\",\"added_by\":\"auto\",\"created_at\":\"2023-05-19 14:58:50\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":136004,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e(a) Tumor samples were divided into two subtypes through consistent clustering. (b) The distribution of hypoxic-related scores among different subtypes.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-2945434/v1/dae16e89ecfcdb81dbc3fe9a.png\"},{\"id\":37245841,\"identity\":\"679208d8-70a7-4796-b180-6589d339581c\",\"added_by\":\"auto\",\"created_at\":\"2023-05-19 14:58:50\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":269269,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e(a) KM survival curves for two subtypes. (b) Expression heatmaps of 30 hypoxia-related lncRNAs in each sample.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-2945434/v1/f2d88fe184caa4c6e4f7c888.png\"},{\"id\":37245843,\"identity\":\"b67fb0be-e5cd-4dce-9216-e3c5ae7c2382\",\"added_by\":\"auto\",\"created_at\":\"2023-05-19 14:58:50\",\"extension\":\"png\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":201382,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e(a, b) Box diagrams of immune cell distribution based on (a) the CIBERSORT algorithm and (b) the MCPCounter algorithm. *p \\u0026lt; 0.05, **p \\u0026lt; 0.01, ***p \\u0026lt; 0.001, ****p \\u0026lt; 0.0001. ns, not significant. (c) Violin plots of the immune score, stromal score, and estimate score distributions in the two subtypes (the top number indicates the significance p-value). (d, e) Box diagrams of HLA family genes and immune checkpoint genes. *p \\u0026lt; 0.05, **p \\u0026lt; 0.01, ***p \\u0026lt; 0.001, ****p \\u0026lt; 0.0001. ns, not significant.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-2945434/v1/49e040d222d51e2c3b0589b3.png\"},{\"id\":37247234,\"identity\":\"070f5c5a-d33b-4cf5-bdc6-af3a8fa659e9\",\"added_by\":\"auto\",\"created_at\":\"2023-05-19 15:06:50\",\"extension\":\"png\",\"order_by\":5,\"title\":\"Figure 5\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":164213,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e(a) Prognostic hypoxia-related lncRNAs obtained by univariate Cox regression analysis. (b) LASSO regression process parameter diagram. (c) Sample RS from the low to high distribution diagram and the survival prognosis time distribution diagram of corresponding samples. (d) Prognosis-related KM curves of the samples based on RS prediction models. The blue and orange curves represent low- and high-risk samples, respectively. (e) 1-, 3-, and 5-year survival prediction ROC curves.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-2945434/v1/32f295acd16a3132cac87b25.png\"},{\"id\":37245839,\"identity\":\"5069f126-a532-43ba-bf37-fd68c8f857c9\",\"added_by\":\"auto\",\"created_at\":\"2023-05-19 14:58:50\",\"extension\":\"png\",\"order_by\":6,\"title\":\"Figure 6\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":125576,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e(a,b) Univariate and multivariate Cox regression forest maps of clinical factors. (c) Nomogram of the survival prediction model based on independent prognostic factors. (d,e) (d) Calibration curves and (e) ROC curves of the nomogram.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"6.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-2945434/v1/3e48331bd6b645a38770e2b3.png\"},{\"id\":37245845,\"identity\":\"621da4fd-5f15-4996-b5be-fd0b70a49223\",\"added_by\":\"auto\",\"created_at\":\"2023-05-19 14:58:50\",\"extension\":\"png\",\"order_by\":7,\"title\":\"Figure 7\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":406714,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eGSEA results of high- vs. low-risk hallmark gene sets.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"7.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-2945434/v1/8874de5722047316443f20bd.png\"},{\"id\":37247233,\"identity\":\"58d16e4b-69a0-4674-99c7-14354c5b5336\",\"added_by\":\"auto\",\"created_at\":\"2023-05-19 15:06:50\",\"extension\":\"png\",\"order_by\":8,\"title\":\"Figure 8\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":181895,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e(a) Correlation heatmaps of five model lncRNAs with seven types of immune cells and two types of stromal cells. (b) Box chart of IC50 values for common drugs across high- and low-risk groups (top numbers indicate significant p-values). (c) Correlation scatterplot between the TIDE score and RS. (d) Distribution of the CYT score, TLS score, and CD8A/PD-L1 ratio between high- and low-risk groups (top numbers indicate significant p-values). (e) The distribution of two subtypes in the high- and low-risk groups.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"8.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-2945434/v1/644443d9aa5d7eb3887fb82d.png\"},{\"id\":37637092,\"identity\":\"a06eff4a-8365-4b97-a8b3-cd552dce9bab\",\"added_by\":\"auto\",\"created_at\":\"2023-05-30 07:29:38\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":2567944,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-2945434/v1/fcd28001-2ac3-490e-b48e-c4b35f725e5b.pdf\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"A hypoxia-related five-lncRNA signature predicts osteosarcoma prognosis\",\"fulltext\":[{\"header\":\"Introduction\",\"content\":\"\\u003cp\\u003eOsteosarcoma (OS) is the most common primary malignant tumor of bones, which usually occurs in the metaphysis of the long bones and is common in the distal femur and proximal humerus (\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e). With the development of multimodal therapy, the 5-year survival rate for patients has improved to \\u0026gt;\\u0026thinsp;70% (\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e). However, many patients develop metastasis at initial diagnosis or following intensive therapy, and tumor recurrence is a common cause of death for patients (\\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e). Therefore, elucidating the molecular mechanisms underlying the development of OS is of great clinical significance for improving treatment strategies.\\u003c/p\\u003e \\u003cp\\u003eHypoxia is a hallmark of solid tumors in humans (\\u003cspan citationid=\\\"CR6\\\" class=\\\"CitationRef\\\"\\u003e6\\u003c/span\\u003e). Similar to normal cells, tumor cells can activate signaling pathways to induce angiogenesis, proliferation, and apoptosis (\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR8\\\" class=\\\"CitationRef\\\"\\u003e8\\u003c/span\\u003e). However, among the pathways induced in tumor cells, those promoting tumor growth are more numerous than those promoting apoptosis. Furthermore, hypoxia may influence a more aggressive phenotype of tumor cells related to a poor prognosis in patients (\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e). A recent study reported that hypoxia could influence the tumor microenvironment by promoting the expression of adverse immune cytokines, increasing the differentiation potential of immunosuppressor cells, and decreasing the activity of natural killer cells and cytotoxic T lymphocytes (\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e). These factors ultimately result in chemotherapy resistance, metastasis, and poor patient outcomes (\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e). In OS, several hypoxia-related prognostic gene signatures have been reported (\\u003cspan citationid=\\\"CR12\\\" class=\\\"CitationRef\\\"\\u003e12\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e). In recent years, studies have gradually revealed the response of long-noncoding RNAs (lncRNAs) to hypoxia and their regulatory role in human cancers (\\u003cspan citationid=\\\"CR14\\\" class=\\\"CitationRef\\\"\\u003e14\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR15\\\" class=\\\"CitationRef\\\"\\u003e15\\u003c/span\\u003e). However, hypoxia-related lncRNA signatures have rarely been reported.\\u003c/p\\u003e \\u003cp\\u003eThis study aimed to screen hypoxia-associated lncRNAs in patients with OS based on the GEO and TARGET databases. Hypoxia-associated lncRNAs were identified by correlation analysis with hypoxia-associated mRNAs. Tumor samples were clustered into different subtypes based on lncRNAs. Prognostic hypoxia-associated lncRNAs were screened using univariate Cox regression analysis, and a prognostic signature was established using LASSO regression analysis. The lncRNA signature obtained may serve as a promising prognostic biomarker for OS.\\u003c/p\\u003e\"},{\"header\":\"Methods\",\"content\":\"\\u003cdiv id=\\\"Sec3\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eAcquisition and preprocessing of expression profile data\\u003c/h2\\u003e \\u003cp\\u003eThe GSE16088 dataset was downloaded from the GEO database. The gene expression profiles of six normal tissue samples and 14 OS tissue samples were included (detection platform: GPL96 [HG-U133A] Affymetrix Human Genome U133A Array). This dataset was used to screen for differentially expressed (DE) hypoxia-related genes. The preprocessed, standardized and log2 transformed probe expression matrix was downloaded. An annotation file was downloaded. Through one-to-one matching of gene symbols and probe numbers, probes without matching gene symbols were removed. The mean value of the different probes was used as the final expression value.\\u003c/p\\u003e \\u003cp\\u003eThe RNA-seq data (log2 (FPKM\\u0026thinsp;+\\u0026thinsp;1)) of OS were obtained from the TARGET database from the UCSC-Xena platform (\\u003cspan citationid=\\\"CR16\\\" class=\\\"CitationRef\\\"\\u003e16\\u003c/span\\u003e). The clinical information of the corresponding samples, including age, sex, ethnicity, disease at diagnosis, specific tumor region, primary tumor site, definitive surgery, primary site progression, and survival information, were also downloaded. A total of 88 OS samples were included, of which 85 had prognostic information. Furthermore, mRNAs and lncRNAs were re-annotated according to the annotation file related to GENCODE V22. The gene with the annotation information of \\u0026ldquo;protein_coding\\u0026rdquo; was retained as the mRNA, and that with the annotation information of \\u0026ldquo;antisense,\\u0026rdquo; \\u0026ldquo;sense_intronic,\\u0026rdquo; \\u0026ldquo;lincRNA,\\u0026rdquo; \\u0026ldquo;sense_overlapping,\\u0026rdquo; \\u0026ldquo;processed_transcript,\\u0026rdquo; \\u0026ldquo;3prime_overlapping_ncRNA,\\u0026rdquo; or \\u0026ldquo;non_coding\\u0026rdquo; was retained as the lncRNA. The expression matrices of the mRNAs and lncRNAs in 88 samples were obtained for follow-up analysis.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec4\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eDE hypoxia-related gene screening\\u003c/h2\\u003e \\u003cp\\u003eBased on the GSE16088 dataset, linear regression and empirical Bayesian methods in limma 3.10.3 (\\u003cspan citationid=\\\"CR17\\\" class=\\\"CitationRef\\\"\\u003e17\\u003c/span\\u003e) were used to analyze the DE genes of tumor vs. normal controls. The Benjamini \\u0026amp; Hochberg method was adopted for multiple corrections, and adjusted p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05 and |logFC| \\u0026gt; 1 were used as thresholds. Volcano mapping was performed to identify the DE genes.\\u003c/p\\u003e \\u003cp\\u003eThe pathway genes of HALLMARK_HYPOXIA were downloaded from MSigDB v7.1 (\\u003cspan citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e) as hypoxia-related genes. The intersection of DE and hypoxia-related genes was used for subsequent analyses.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec5\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eIdentification of hypoxia-related lncRNAs\\u003c/h2\\u003e \\u003cp\\u003eFirst, mRNAs obtained from the TARGET database were matched with DE hypoxia-related genes to obtain the expression values of DE hypoxia-related mRNAs in the 88 samples. Pearson\\u0026rsquo;s correlation coefficients between all lncRNAs and DE hypoxia-related mRNAs were calculated. Pairs with p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.001 and |Pearson correlation coefficient| \\u0026lt; 0.4 were selected. Furthermore, lncRNAs with expression values\\u0026thinsp;\\u0026gt;\\u0026thinsp;0 in \\u0026gt;\\u0026thinsp;80% of the samples and correlated with \\u0026gt;\\u0026thinsp;5 DE hypoxia-related mRNAs were selected. The corresponding lncRNA-mRNA network was visualized using Cytoscape 3.9.0 (\\u003cspan citationid=\\\"CR19\\\" class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e). lncRNAs in the network were considered hypoxia-related lncRNAs.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec6\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003ePrediction of molecular subtypes of OS\\u003c/h2\\u003e \\u003cp\\u003eWith the expression values of hypoxia-related lncRNAs in various TARGET dataset samples, ConsensusClusterPlus 1.54.0 (\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e) in R3.6.1 was used to analyze hypoxia-related molecular subtypes of all samples.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec7\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eHypoxic-related score\\u003c/h2\\u003e \\u003cp\\u003eBased on the expression values of all mRNAs in each sample of the TARGET dataset, the hypoxia-related enrichment score was calculated using the ssGSEA algorithm of GSVA 1.36.2 (\\u003cspan citationid=\\\"CR21\\\" class=\\\"CitationRef\\\"\\u003e21\\u003c/span\\u003e) with all hypoxia-related genes as the enrichment background. Then, combined with the subtype information of the samples, the Wilcox test was used to compare the difference in significance of hypoxia-related scores among subtypes, and a violin plot was drawn.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec8\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eCorrelation analysis between prognosis and clinical outcome of subtypes\\u003c/h2\\u003e \\u003cp\\u003eBased on the obtained hypoxia-related subtypes combined with the prognostic information of the samples, the KM survival curves of the subtypes were drawn, and the significance p-value was calculated using the log-rank test. Additionally, the clinical information of the subtypes was compared. For factor-type variables, the Chi-squared test was used to conduct statistical analysis, and for continuous variables, the Wilcox test was used.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec9\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eComparison of immune microenvironments between subtypes\\u003c/h2\\u003e \\u003cp\\u003eIn order to observe the differences in the immune microenvironment among subtypes, the CIBERSORT (\\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e) algorithm was used to calculate the proportions of 22 types of immune cells according to the gene expression levels of the OS samples. The Wilcoxon test was used to identify the p-value among the subtypes.\\u003c/p\\u003e \\u003cp\\u003eThe R-pack MCPcounter (\\u003cspan citationid=\\\"CR23\\\" class=\\\"CitationRef\\\"\\u003e23\\u003c/span\\u003e) algorithm was used to compare the infiltration levels of eight immune cell and two stromal cell types based on the gene expression levels of the OS samples. Combined with the subtype grouping of samples, the Wilcox test was used to calculate the differential p-values among subtypes.\\u003c/p\\u003e \\u003cp\\u003eIn addition, the ESTIMATE (\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e) algorithm was used to estimate the stromal, immune, and ESTIMATE scores of tumor samples according to the expression data. The Wilcoxon test was used to calculate the differences among the subtypes.\\u003c/p\\u003e \\u003cp\\u003eFurthermore, the expression data of HLA family genes and immune checkpoint genes in each OS sample were extracted for the comparison of expression differences among subtypes by the t-test, and a box diagram was drawn.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec10\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eIdentification of prognostic hypoxia-related lncRNAs\\u003c/h2\\u003e \\u003cp\\u003eIn TARGET-OS samples, based on the hypoxia-related lncRNAs obtained above, combined with the clinical survival and prognosis information, hypoxia-related lncRNAs significantly correlated with overall survival prognosis were identified using univariate Cox regression analysis in survival 2.41-1 (\\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e). The threshold was set at p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec11\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eConstruction of the prognostic hypoxia-related lncRNA signature\\u003c/h2\\u003e \\u003cp\\u003eBased on the hypoxia-related lncRNAs significantly correlated with survival prognosis, the LASSO Cox regression model (\\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e) was also applied to further screen the lncRNA combination by combining the survival prognosis information of TARGET-OS and expression value of lncRNAs in each sample using glmnet 2.0\\u0026ndash;18 (\\u003cspan citationid=\\\"CR27\\\" class=\\\"CitationRef\\\"\\u003e27\\u003c/span\\u003e) in R3.6.1. Then, according to the regression coefficient of lncRNA in hypoxia-related lncRNA combinations and the expression level of lncRNA in TARGET-OS samples, the following risk score (RS) model was established:\\u003c/p\\u003e \\u003cp\\u003eRS = \\u0026sum;β\\u003csub\\u003elncRNA\\u003c/sub\\u003e\\u0026thinsp;\\u0026times;\\u0026thinsp;Exp\\u003csub\\u003elncRNA\\u003c/sub\\u003e (\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e)\\u003c/p\\u003e \\u003cp\\u003ewhere β\\u003csub\\u003elncRNA\\u003c/sub\\u003e represents the LASSO regression coefficient of lncRNA and Exp\\u003csub\\u003elncRNA\\u003c/sub\\u003e represents the expression level of lncRNA in the TARGET-OS dataset.\\u003c/p\\u003e \\u003cp\\u003eAccording to the RS median value, all samples were divided into high-risk (\\u0026ge;\\u0026thinsp;RS median value) and low-risk (\\u0026lt;\\u0026thinsp;RS median value) groups. The association between risk grouping and actual survival prognosis was assessed using the Kaplan\\u0026ndash;Meier (KM) curve method in Survival 2.41-1 of R3.6.1. The prediction ROC curves for 1, 3, and 5 years were drawn.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec12\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eAnalysis of independent prognostic factors\\u003c/h2\\u003e \\u003cp\\u003eTo determine whether the RS model was an independent prognostic factor, age, sex, ethnicity, disease at diagnosis, specific tumor region, primary tumor site, definitive surgery, primary site progression, and RS were included in univariate Cox regression analysis, followed by multivariate Cox regression analysis. The independent prognostic factors were used to establish a nomogram. Meanwhile, the calibration and ROC curves of the nomogram were drawn to verify its validity.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec13\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eAnalysis of molecular mechanisms between different risk groups\\u003c/h2\\u003e \\u003cp\\u003eTo determine whether each hallmark gene set was significantly enriched in a particular risk group, with h.all.v7.4. symbols in the MSigDB v7.1 database as the enrichment background, GSEA of high- vs. low-risk was performed using R clusterProfiler 3.8.1 (\\u003cspan citationid=\\\"CR28\\\" class=\\\"CitationRef\\\"\\u003e28\\u003c/span\\u003e). The expression values of all mRNAs in TARGET-OS samples and high- and low-risk group sample information were included. Statistical significance was set at p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec14\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eCorrelation analysis of model lncRNAs and immune cells\\u003c/h2\\u003e \\u003cp\\u003eBased on the differential immune cells between subtypes obtained in the previous analysis, the Spearman correlation coefficients and the corresponding significance p-value between model lncRNAs and immune cells were calculated, and a correlation heatmap was drawn.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec15\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eDrug sensitivity analysis\\u003c/h2\\u003e \\u003cp\\u003eDrug sensitivity was reflected by the IC50 values. IC50 can be predicted by the ridge regression model according to the GDSC cell line expression profile and TARGET-OS mRNA expression profile using the pRRophetic algorithm. Here, pRRophetic 0.5 (\\u003cspan citationid=\\\"CR29\\\" class=\\\"CitationRef\\\"\\u003e29\\u003c/span\\u003e) was used to predict the IC50 values based on the 138 drugs provided in this package, and the Wilcox test was used to analyze whether there were significant differences in the IC50 values of each drug between the risk groups.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec16\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003ePrediction of the response to immunotherapy\\u003c/h2\\u003e \\u003cp\\u003eThe immunotherapy response was determined using tumor immune dysfunction and elimination (TIDE) analysis. TIDE uses two main types of tumor immune escape mechanisms to predict the immune response to treatment. The TIDE values of all samples were calculated using TIDE tools (\\u003cspan citationid=\\\"CR30\\\" class=\\\"CitationRef\\\"\\u003e30\\u003c/span\\u003e), and the Spearman correlation between RS and TIDE scores was calculated.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec17\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eCorrelation analysis of immune indexes\\u003c/h2\\u003e \\u003cp\\u003eThe cytolytic activity (CYT) scoreis an established biomarker for predicting the presence of tumor-infiltrating lymphocytes based on gene expression. The CYT score of each sample in the high- and low-risk groups was calculated according to the CYT score calculation method of a previous study (\\u003cspan citationid=\\\"CR31\\\" class=\\\"CitationRef\\\"\\u003e31\\u003c/span\\u003e). Additionally, based on the expression values of CD8A and PD-L1 in each tumor sample, the CD8A/PD-L1 ratio was calculated, and the difference between the two groups was calculated using the Wilcox test.\\u003c/p\\u003e \\u003cp\\u003eA recent study obtained a nine-gene signature of the tertiary lymphoid structure (TLS) in melanoma (\\u003cspan citationid=\\\"CR32\\\" class=\\\"CitationRef\\\"\\u003e32\\u003c/span\\u003e). Based on the expression values of the nine genes in each tumor sample, here, the ssGSEA algorithm was used to estimate the TLS score of each sample, and the Wilcox test was used for difference comparison. Finally, the distribution proportion of subtypes in the high- and low-risk groups was calculated, and significance was calculated using the Chi-squared test.\\u003c/p\\u003e \\u003c/div\\u003e\"},{\"header\":\"Results\",\"content\":\"\\u003cdiv id=\\\"Sec19\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eDE hypoxia-related mRNA and lncRNA identification\\u003c/h2\\u003e\\n \\u003cp\\u003eA total of 2,424 upregulated and 2,172 downregulated mRNAs were obtained from the GSE16088 dataset (Fig. \\u003cspan class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003ea). The intersection of DE mRNAs and hypoxia-related genes was further examined, and 72 DE hypoxia-related mRNAs were obtained (Fig. \\u003cspan class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003eb).\\u003c/p\\u003e\\n \\u003cp\\u003ePearson correlation analysis obtained 5,981 lncRNA-mRNA correlation relationship pairs, including 2,802 lncRNAs and 69 mRNAs, below thresholds of p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.001 and |Pearson correlation coefficient| \\u0026gt; 0.4, and lncRNAs with expression values\\u0026thinsp;\\u0026gt;\\u0026thinsp;0 in \\u0026gt;\\u0026thinsp;80% of samples. Due to the large number of lncRNAs, only those correlated with \\u0026gt;\\u0026thinsp;5 DE hypoxia-related mRNAs were selected. Finally, 203 lncRNA-mRNA relationship pairs, including 30 lncRNAs and 60 mRNAs, were retained. The network diagram is shown in Fig. \\u003cspan class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003eC.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec20\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eMolecular subtype analysis\\u003c/h2\\u003e\\n \\u003cp\\u003eBased on the 30 hypoxia-related lncRNAs obtained above and their expression values in osteosarcoma samples, consistent clustering was conducted. As shown in Fig. 2A, two clusters were finally determined. The hypoxia-related scores of the samples in cluster 2 were significantly higher than those in cluster 1 (p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05) (Fig. 2b).\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec21\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003ePrognosis and clinical correlation analysis of subtypes\\u003c/h2\\u003e\\n \\u003cp\\u003eComparison of clinical information between subtypes revealed that primary site progression was significantly different between the two subtypes (p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05) (Table \\u003cspan class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e).\\u0026nbsp;\\u003c/p\\u003e\\n \\u003ctable id=\\\"Tab1\\\" border=\\\"1\\\"\\u003e\\n \\u003ccaption language=\\\"En\\\"\\u003e\\n \\u003cdiv class=\\\"CaptionNumber\\\"\\u003eTable 1\\u003c/div\\u003e\\n \\u003cdiv class=\\\"CaptionContent\\\"\\u003e\\n \\u003cp\\u003eClinical information between subtypes.\\u003c/p\\u003e\\n \\u003c/div\\u003e\\n \\u003c/caption\\u003e\\n \\u003cthead\\u003e\\n \\u003ctr\\u003e\\n \\u003cth align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eCharacteristics\\u003c/p\\u003e\\n \\u003c/th\\u003e\\n \\u003cth align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eCluster 1 (N\\u0026thinsp;=\\u0026thinsp;45)\\u003c/p\\u003e\\n \\u003c/th\\u003e\\n \\u003cth align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eCluster 2 (N\\u0026thinsp;=\\u0026thinsp;42)\\u003c/p\\u003e\\n \\u003c/th\\u003e\\n \\u003cth align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eTotal (N\\u0026thinsp;=\\u0026thinsp;87)\\u003c/p\\u003e\\n \\u003c/th\\u003e\\n \\u003cth align=\\\"left\\\"\\u003e\\n \\u003cp\\u003ep-value\\u003c/p\\u003e\\n \\u003c/th\\u003e\\n \\u003c/tr\\u003e\\n \\u003c/thead\\u003e\\n \\u003ctbody\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eAge\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e0.44\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eMean\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;SD\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e14.77\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;5.15\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e15.38\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;4.36\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e15.06\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;4.77\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eMedian[min-max]\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e14.05 [3.56,32.41]\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e15.08 [5.96,29.14]\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e14.49 [3.56,32.41]\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eGender\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e0.52\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eFemale\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e21 (24.14%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e16 (18.39%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e37 (42.53%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eMale\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e24 (27.59%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e26 (29.89%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e50 (57.47%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eEthnicity\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e1\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eHispanic or Latino\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e6 (9.23%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e5 (7.69%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e11 (16.92%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eNot Hispanic or Latino\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e29 (44.62%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e25 (38.46%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e54 (83.08%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eDisease at diagnosis\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e0.81\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eMetastatic\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e12 (13.79%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e10 (11.49%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e22 (25.29%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eNon-metastatic\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e33 (37.93%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e32 (36.78%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e65 (74.71%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003ePrimary tumor site\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e1\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eArm/hand\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e3 (3.53%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e3 (3.53%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e6 (7.06%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eLeg/foot\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e41 (48.24%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e38 (44.71%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e79 (92.94%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eSpecific tumor region\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e0.09\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eDistal\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e17 (33.33%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e11 (21.57%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e28 (54.90%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eProximal\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e8 (15.69%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e15 (29.41%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e23 (45.10%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eDefinitive surgery\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e1\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eAmputation\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e2 (4.44%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e3 (6.67%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e5 (11.11%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eLimb sparing\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e19 (42.22%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e21 (46.67%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e40 (88.89%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003ePrimary site progression\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e0.02\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eNo\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e11 (28.95%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e12 (31.58%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e23 (60.53%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003eYes\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e13 (34.21%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e2 (5.26%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\n \\u003cp\\u003e15 (39.47%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd align=\\\"left\\\"\\u003e\\u0026nbsp;\\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003c/tbody\\u003e\\n \\u003c/table\\u003e\\n \\u003cp\\u003eKM survival curves of the two subtypes are plotted in Fig. 3a. The prognosis of cluster 1 was worse than that of cluster 2 (p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05). A heatmap of the expression of 30 hypoxia-related lncRNAs in each sample was drawn. The expression patterns of the 30 hypoxia-related lncRNAs differed between the two subtypes (Fig. 3B).\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec22\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eComparison of immune microenvironments between subtypes\\u003c/h2\\u003e\\n \\u003cp\\u003eBased on the CIBERSORT algorithm, the immune cell infiltration levels of six immune cells, including CD4 na\\u0026iuml;ve T cells, gamma delta T cells, M0 macrophages, M1 macrophages, M2 macrophages, and resting dendritic cells, were significantly different between the two subtypes (p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05) (Fig. 4a). Based on the MCPCounter algorithm, neutrophils and two stromal cell types (endothelial cells and fibroblasts) were significantly different between the two groups (p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05) (Fig. 4b). The ESTIMATE algorithm revealed that the stromal score in cluster 2 was significantly higher than that in cluster 1 (p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05), and there was no significant difference between the two subtypes of immune and microenvironment scores (ESTIMATE score) (Fig. 4C).\\u003c/p\\u003e\\n \\u003cp\\u003eThree HLA family genes (HLA-A, HLA-E, and HLA-F) showed significant differences between the two subtypes (p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05) (Fig.\\u0026nbsp;4d). Two immune checkpoint genes (\\u003cem\\u003ePDCD1LG2\\u003c/em\\u003e and \\u003cem\\u003eTNFRSF4\\u003c/em\\u003e) showed significant differences between the two subtypes (p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05) (Fig.\\u0026nbsp;4e).\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec23\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eIdentification of the prognostic hypoxia-related lncRNA signature\\u003c/h2\\u003e\\n \\u003cp\\u003eBased on the 30 hypoxia-related lncRNAs, 10 were selected using univariate Cox regression analysis (Fig. 5a). Further, with the expression values of 10 prognostic hypoxia-related lncRNAs in OS samples, and the survival time and state of the samples, five optimized lncRNA signatures (CTB-4E7.1, MIR210HG, RP11-817J15.3, RP11-81A22.5, and RPARP-AS1) were screened by the LASSO Cox regression algorithm. Prognostic regression coefficients are shown in Fig. 5B. Based on the LASSO regression prognostic coefficients of the five optimized lncRNAs and their expression levels in the TARGET-OS dataset, the RS model was constructed. The RS visualization is shown in Fig. 5c. The association between risk grouping and actual disease prognostic information was evaluated using KM curves. As shown in Fig. 5d, the prognosis in the low-risk group was significantly better than that in the high-risk group. The 1-, 3-, and 5-year survival prediction ROC curves are shown in Fig. 5e. The results indicated a significant correlation between the risk groups predicted by the RS model and actual prognosis.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec24\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eAnalysis of the independent prognostic factor\\u003c/h2\\u003e\\n \\u003cp\\u003eFollowing univariate Cox regression analysis, the disease at diagnosis, primary site progression, and RS were screened (Fig. 6A). After multivariate Cox regression analysis, the disease at diagnosis and RS were considered independent prognostic factors (Fig. 6B). A nomogram was constructed for independent prognostic factors, as shown in Fig. 6C. The survival time of the samples was predicted based on the \\u0026ldquo;Total Points\\u0026rdquo; axis in the first line. The 1-, 3-, and 5-year survival rates predicted by the model were consistent with the actual 1-, 3-, and 5-year survival rates (Fig. 6d). The AUCs of the 1-, 3-, and 5-year ROC curves were 0.889, 0.852, and 0.857, respectively (Fig. 6e).\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec25\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eAnalysis of related molecular mechanisms between risk groups\\u003c/h2\\u003e\\n \\u003cp\\u003eEnrichment analysis of hallmark gene sets was performed for the high- and low-risk groups. A total of 13 upregulated pathways, such as cholesterol homeostasis, hypoxia pathway, and oxidative phosphorylation, and 12 downregulated pathways, such as complement, epithelial mesenchymal transition, and inflammatory response, were identified (Fig. \\u003cspan class=\\\"InternalRef\\\"\\u003e7\\u003c/span\\u003e).\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec26\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eCorrelation analysis of model lncRNAs and immune cells\\u003c/h2\\u003e\\n \\u003cp\\u003eSpearman correlation coefficients between the model lncRNAs and the differential immune cells between the subtypes were calculated. As shown in Fig. 8A, fibroblasts had the highest negative correlation with RP11-81A22.5, endothelial cells had the highest positive correlation with CTB-4E7.1, and M1 macrophages had the highest positive correlation with RP11-817J15.3.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec27\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eDrug sensitivity analysis between high- and low-risk groups\\u003c/h2\\u003e\\n \\u003cp\\u003eA total of 39 drugs had significantly different IC50 values between the high- and low-risk groups. The IC50 value distribution box plot of the common drugs elesclomol, midostaurin, and vorinostat is shown in Fig.\\u0026nbsp;8b.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec28\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003ePrediction of the response to immunotherapy\\u003c/h2\\u003e\\n \\u003cp\\u003eThe scatterplot for the correlation between the TIDE score and RS is shown in Fig.\\u0026nbsp;8c, and there was a significantly negative correlation between them; the higher the RS, the lower the TIDE score.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec29\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eCorrelation analysis of immune indexes\\u003c/h2\\u003e\\n \\u003cp\\u003eThe CD8A/PD-L1 ratio in the high-risk group was significantly higher than that in the low-risk group, whereas the CYT and TLS scores were not significantly different between the two groups (Fig.\\u0026nbsp;8d).\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec30\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eAssociation analysis of the hypoxia subtype and risk grouping\\u003c/h2\\u003e\\n \\u003cp\\u003eThe distribution of the two subtypes in the high- and low-risk groups is shown in Fig.\\u0026nbsp;8e. The high-risk group contained more samples from cluster 1.\\u003c/p\\u003e\\n\\u003c/div\\u003e\"},{\"header\":\"Discussion\",\"content\":\"\\u003cp\\u003eTumor hypoxia is a common feature of many cancers, and the activation of hypoxia pathways is correlated with a poor prognosis in many cancer types. Recent evidence suggests the involvement of hypoxia-regulated lncRNAs in cancer cells (\\u003cspan citationid=\\\"CR33\\\" class=\\\"CitationRef\\\"\\u003e33\\u003c/span\\u003e). In the present study, 30 hypoxia-related lncRNAs were identified. OS samples were classified into two subtypes based on the lncRNAs. Furthermore, five prognostic hypoxia-related lncRNAs were screened through LASSO regression analyses, and an RS model was constructed. The samples in the two risk groups presented differences in prognosis, pathways, and the immune microenvironment.\\u003c/p\\u003e \\u003cp\\u003eGrowing evidence suggests that OS is heterogeneous, and some subtypes have been reported based on genomic analysis (\\u003cspan citationid=\\\"CR34\\\" class=\\\"CitationRef\\\"\\u003e34\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR35\\\" class=\\\"CitationRef\\\"\\u003e35\\u003c/span\\u003e). In this study, OS was further classified into two subtypes based on 30 hypoxia-related lncRNAs. The prognosis of cluster 1 (low hypoxia score) was worse than that of cluster 2 (high hypoxia score). M2 macrophages had a higher infiltration level among the 22 immune cell types, and cluster 1 had a higher infiltration level of M2 macrophages than cluster 2. In tumors, M1 macrophages are regarded as antitumor effectors, and M2 macrophages as protumor modulators owing to their induction of neovascularization (\\u003cspan citationid=\\\"CR36\\\" class=\\\"CitationRef\\\"\\u003e36\\u003c/span\\u003e). The density of tumor-associated macrophages is correlated with invasion, metastasis, and a poor prognosis in some cancers, including OS (\\u003cspan citationid=\\\"CR37\\\" class=\\\"CitationRef\\\"\\u003e37\\u003c/span\\u003e). Thus, the higher infiltration level of M2 macrophages in cluster 1 was consistent with the worse prognosis of the samples in this cluster. Two immune checkpoint genes (\\u003cem\\u003ePDCD1LG2\\u003c/em\\u003e and \\u003cem\\u003eTNFRSF4\\u003c/em\\u003e) showed significant differences between the two subtypes. Here, it was speculated that \\u003cem\\u003ePDCD1LG2\\u003c/em\\u003e and \\u003cem\\u003eTNFRSF4\\u003c/em\\u003e may serve as biomarkers to predict the response of patients with OS to immune checkpoint blocking therapy.\\u003c/p\\u003e \\u003cp\\u003eFive prognosis-related lncRNAs were selected from 30 hypoxia-related lncRNAs, and five prognosis-related ones were further selected: CTB-4E7.1, MIR210HG, RP11-817J15.3, RP11-81A22.5, and RPARP-AS1. CTB-4E7.1 has been demonstrated to be one of the prognostic risk signatures for OS (\\u003cspan citationid=\\\"CR38\\\" class=\\\"CitationRef\\\"\\u003e38\\u003c/span\\u003e). MIR210HG has been reported to promote tumor cell proliferation, invasion, and metastasis in many cancers, including OS, indicating a poor prognosis (\\u003cspan additionalcitationids=\\\"CR40\\\" citationid=\\\"CR39\\\" class=\\\"CitationRef\\\"\\u003e39\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR41\\\" class=\\\"CitationRef\\\"\\u003e41\\u003c/span\\u003e). RP1-261G23.7, RP11-81A22.5, and RPARP-AS1 have also been found to be DE in OS in recent studies (\\u003cspan citationid=\\\"CR42\\\" class=\\\"CitationRef\\\"\\u003e42\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR43\\\" class=\\\"CitationRef\\\"\\u003e43\\u003c/span\\u003e). The present study is the first to provide a risk model for OS with f5 lncRNAs with prognostic relevance linked to hypoxia.\\u003c/p\\u003e \\u003cp\\u003eAfter the prognostic RS model for OS was constructed, subsequent KM analysis and ROC curves further indicated the predictive value of the RS model. The high-risk group was associated with a poor prognosis. The AUC of the 5-year survival prediction ROC curve was 0.746, which was similar to other signatures of OS (\\u003cspan citationid=\\\"CR38\\\" class=\\\"CitationRef\\\"\\u003e38\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR44\\\" class=\\\"CitationRef\\\"\\u003e44\\u003c/span\\u003e). Moreover, RS was an independent prognostic factor associated with disease at diagnosis. The nomogram constructed based on the two independent prognostic factors could accurately predict the 1-, 3-, and 5-year survival rates of patients with OS. These results suggested that this prognostic model has the potential to evaluate the prognosis of patients with OS.\\u003c/p\\u003e \\u003cp\\u003eTo further investigate the molecular mechanisms between the two risk groups, GSEA was conducted to analyze the pathways between the two groups. The hypoxia pathway was among the upregulated pathways, and epithelial-mesenchymal transition was one of the downregulated pathways. Epithelial-mesenchymal transition is associated with tumor proliferation, vascularization, and metastasis (\\u003cspan citationid=\\\"CR45\\\" class=\\\"CitationRef\\\"\\u003e45\\u003c/span\\u003e). Both epithelial mesenchymal transition and hypoxia are regarded as separate events that promote the invasion and metastasis of many cancer types (\\u003cspan citationid=\\\"CR46\\\" class=\\\"CitationRef\\\"\\u003e46\\u003c/span\\u003e).\\u003c/p\\u003e \\u003cp\\u003eAdditionally, the correlation between the risk groups and immunity was analyzed. Recent studies have examined the mechanisms of tumor immune evasion (\\u003cspan citationid=\\\"CR47\\\" class=\\\"CitationRef\\\"\\u003e47\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR48\\\" class=\\\"CitationRef\\\"\\u003e48\\u003c/span\\u003e). Some cancers have high infiltration levels of cytotoxic T cells; however, these T cells are often in a dysfunctional state. In other cancers, immunosuppressive factors may exclude T cells from infiltrating tumors (\\u003cspan citationid=\\\"CR49\\\" class=\\\"CitationRef\\\"\\u003e49\\u003c/span\\u003e). TIDE was developed by Peng et al. (\\u003cspan citationid=\\\"CR30\\\" class=\\\"CitationRef\\\"\\u003e30\\u003c/span\\u003e) to identify the factors underlying tumor immune escape mechanisms. In this study, there was a significant negative correlation between the TIDE score and RS, indicating different responses to immunotherapy in the two risk groups. Moreover, the high-risk group contained more samples from cluster 1, which had a poor prognosis.\\u003c/p\\u003e\"},{\"header\":\"Conclusions\",\"content\":\"\\u003cp\\u003eIn conclusion, the present study divided OS into two subtypes based on 30 hypoxia-related lncRNAs. Further analyses identified five hypoxia-related lncRNAs and constructed a prognostic signature. The model lncRNAs may be involved in the progression of OS via the epithelial-mesenchymal transition and hypoxia pathways. Further experiments are required to confirm these findings.\\u003c/p\\u003e\"},{\"header\":\"Abbreviations\",\"content\":\"\\u003cdiv class=\\\"DefinitionList\\\"\\u003e \\u003cdiv class=\\\"DefinitionListEntry\\\"\\u003e \\u003cdiv class=\\\"Term\\\"\\u003eLncRNA\\u003c/div\\u003e \\u003cdiv class=\\\"Description\\\"\\u003e \\u003cp\\u003eLong-noncoding RNA\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv class=\\\"DefinitionListEntry\\\"\\u003e \\u003cdiv class=\\\"Term\\\"\\u003eOS\\u003c/div\\u003e \\u003cdiv class=\\\"Description\\\"\\u003e \\u003cp\\u003eOsteosarcoma\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv class=\\\"DefinitionListEntry\\\"\\u003e \\u003cdiv class=\\\"Term\\\"\\u003eRS\\u003c/div\\u003e \\u003cdiv class=\\\"Description\\\"\\u003e \\u003cp\\u003eRisk score\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv class=\\\"DefinitionListEntry\\\"\\u003e \\u003cdiv class=\\\"Term\\\"\\u003eDE\\u003c/div\\u003e \\u003cdiv class=\\\"Description\\\"\\u003e \\u003cp\\u003edifferentially expressed\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv class=\\\"DefinitionListEntry\\\"\\u003e \\u003cdiv class=\\\"Term\\\"\\u003eKM\\u003c/div\\u003e \\u003cdiv class=\\\"Description\\\"\\u003e \\u003cp\\u003eKaplan\\u0026ndash;Meier\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv class=\\\"DefinitionListEntry\\\"\\u003e \\u003cdiv class=\\\"Term\\\"\\u003eTIDE\\u003c/div\\u003e \\u003cdiv class=\\\"Description\\\"\\u003e \\u003cp\\u003eTumor immune dysfunction and elimination\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv class=\\\"DefinitionListEntry\\\"\\u003e \\u003cdiv class=\\\"Term\\\"\\u003eCYT\\u003c/div\\u003e \\u003cdiv class=\\\"Description\\\"\\u003e \\u003cp\\u003eCytolytic activity\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv class=\\\"DefinitionListEntry\\\"\\u003e \\u003cdiv class=\\\"Term\\\"\\u003eTLS\\u003c/div\\u003e \\u003cdiv class=\\\"Description\\\"\\u003e \\u003cp\\u003eTertiary lymphoid structure\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003c/div\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eAcknowledgments\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eNot applicable.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eAuthors\\u0026rsquo; contributions\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eAll authors contributed to this paper. YZ\\u0026nbsp;and\\u0026nbsp;KZ\\u0026nbsp;devised the study. XW and QB performed\\u0026nbsp;experiments and conducted data analysis.\\u0026nbsp;XW, BX, YT and YC\\u0026nbsp;conducted data preparation.\\u0026nbsp;YZ, KZ, XW\\u0026nbsp;and\\u0026nbsp;QB\\u0026nbsp;participated in authoring the manuscript. All authors have read and approved the manuscript.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eFunding \\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThis study was supported by the National Natural Science Foundation of China (82173249).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eAvailability of data and materials\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eAll data generated or analysed during this study are included in this published article.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eEthics approval and consent to participate\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eNot applicable.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eConsent for publication\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eNot applicable.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eConflicts of Interest\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe authors declare that they have no competing interests.\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\n\\u003cli\\u003eHuang C-Y, Wei P-L, Wang J-W, Makondi PT, Huang M-T, Chen H-A, et al. Glucose-regulated protein 94 modulates the response of osteosarcoma to chemotherapy. Disease Markers. 2019;2019.\\u003c/li\\u003e\\n\\u003cli\\u003eLin J, Wang X, Wang X, Wang S, Shen R, Yang Y, et al. Hypoxia increases the expression of stem cell markers in human osteosarcoma cells. Oncology Letters. 2021;21(3):1.\\u003c/li\\u003e\\n\\u003cli\\u003eAnderson ME. Update on survival in osteosarcoma. Orthopedic Clinics. 2016;47(1):283-92.\\u003c/li\\u003e\\n\\u003cli\\u003eZhang Z, Luo G, Yu C, Yu G, Jiang R, Shi X. Retracted: MicroRNA‐493‐5p inhibits proliferation and metastasis of osteosarcoma cells by targeting Kruppel‐like factor 5. Journal of cellular physiology. 2019;234(8):13525-33.\\u003c/li\\u003e\\n\\u003cli\\u003eKansara M, Teng MW, Smyth MJ, Thomas DM. Translational biology of osteosarcoma. Nature Reviews Cancer. 2014;14(11):722-35.\\u003c/li\\u003e\\n\\u003cli\\u003eRuan K, Song G, Ouyang G. Role of hypoxia in the hallmarks of human cancer. Journal of cellular biochemistry. 2009;107(6):1053-62.\\u003c/li\\u003e\\n\\u003cli\\u003eIyer NV, Leung SW, Semenza GL. The human hypoxia-inducible factor 1\\u0026alpha; gene: Hif1astructure and evolutionary conservation. Genomics. 1998;52(2):159-65.\\u003c/li\\u003e\\n\\u003cli\\u003eGuo Z, Wang X, Yang Y, Chen W, Zhang K, Teng B, et al. Hypoxic tumor-derived exosomal long noncoding RNA UCA1 promotes angiogenesis via miR-96-5p/AMOTL2 in pancreatic cancer. Molecular Therapy-Nucleic Acids. 2020;22:179-95.\\u003c/li\\u003e\\n\\u003cli\\u003eVaupel P, Kallinowski F, Okunieff P. Blood flow, oxygen and nutrient supply, and metabolic microenvironment of human tumors: a review. Cancer research. 1989;49(23):6449-65.\\u003c/li\\u003e\\n\\u003cli\\u003eTerry S, Buart S, Chouaib S. Hypoxic stress-induced tumor and immune plasticity, suppression, and impact on tumor heterogeneity. Frontiers in immunology. 2017;8:1625.\\u003c/li\\u003e\\n\\u003cli\\u003eWalsh JC, Lebedev A, Aten E, Madsen K, Marciano L, Kolb HC. The clinical importance of assessing tumor hypoxia: relationship of tumor hypoxia to prognosis and therapeutic opportunities. Antioxidants \\u0026amp; redox signaling. 2014;21(10):1516-54.\\u003c/li\\u003e\\n\\u003cli\\u003eFu Y, Bao Q, Liu Z, He G, Wen J, Liu Q, et al. Development and validation of a hypoxia-associated prognostic signature related to osteosarcoma metastasis and immune infiltration. Frontiers in Cell and Developmental Biology. 2021;9:633607.\\u003c/li\\u003e\\n\\u003cli\\u003eJiang F, Miao X-L, Zhang X-T, Yan F, Mao Y, Wu C-Y, et al. A hypoxia gene-based signature to predict the survival and affect the tumor immune microenvironment of osteosarcoma in children. Journal of Immunology Research. 2021;2021.\\u003c/li\\u003e\\n\\u003cli\\u003eDong H, Hu J, Zou K, Ye M, Chen Y, Wu C, et al. Activation of LncRNA TINCR by H3K27 acetylation promotes Trastuzumab resistance and epithelial-mesenchymal transition by targeting MicroRNA-125b in breast Cancer. Molecular cancer. 2019;18:1-18.\\u003c/li\\u003e\\n\\u003cli\\u003eHan Y, Wang X, Mao E, Shen B, Huang L. Analysis of differentially expressed lncRNAs and mRNAs for the identification of hypoxia-regulated angiogenic genes in colorectal cancer by RNA-seq. Medical Science Monitor: International Medical Journal of Experimental and Clinical Research. 2019;25:2009.\\u003c/li\\u003e\\n\\u003cli\\u003eGoldman M, Craft B, Hastie M, Repečka K, McDade F, Kamath A, et al. The UCSC Xena platform for public and private cancer genomics data visualization and interpretation. biorxiv. 2018:326470.\\u003c/li\\u003e\\n\\u003cli\\u003eSmyth GK. Limma: linear models for microarray data. Bioinformatics and computational biology solutions using R and Bioconductor. 2005:397-420.\\u003c/li\\u003e\\n\\u003cli\\u003eLiberzon A, Subramanian A, Pinchback R, Thorvaldsd\\u0026oacute;ttir H, Tamayo P, Mesirov JP. Molecular signatures database (MSigDB) 3.0. Bioinformatics. 2011;27(12):1739-40.\\u003c/li\\u003e\\n\\u003cli\\u003eShannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome research. 2003;13(11):2498-504.\\u003c/li\\u003e\\n\\u003cli\\u003eWilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26(12):1572-3.\\u003c/li\\u003e\\n\\u003cli\\u003eH\\u0026auml;nzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC bioinformatics. 2013;14:1-15.\\u003c/li\\u003e\\n\\u003cli\\u003eKawada J-i, Takeuchi S, Imai H, Okumura T, Horiba K, Suzuki T, et al. Immune cell infiltration landscapes in pediatric acute myocarditis analyzed by CIBERSORT. Journal of Cardiology. 2021;77(2):174-8.\\u003c/li\\u003e\\n\\u003cli\\u003eBecht E, Giraldo NA, Lacroix L, Buttard B, Elarouci N, Petitprez F, et al. Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome biology. 2016;17(1):1-20.\\u003c/li\\u003e\\n\\u003cli\\u003eYoshihara K, Shahmoradgoli M, Mart\\u0026iacute;nez E, Vegesna R, Kim H, Torres-Garcia W, et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nature communications. 2013;4(1):2612.\\u003c/li\\u003e\\n\\u003cli\\u003eWang P, Wang Y, Hang B, Zou X, Mao J-H. A novel gene expression-based prognostic scoring system to predict survival in gastric cancer. Oncotarget. 2016;7(34):55343.\\u003c/li\\u003e\\n\\u003cli\\u003eTibshirani R. The lasso method for variable selection in the Cox model. Statistics in medicine. 1997;16(4):385-95.\\u003c/li\\u003e\\n\\u003cli\\u003eFriedman J, Hastie T, Tibshirani R, Narasimhan B, Tay K, Simon N. glmnet: Lasso and elastic-net regularized generalized linear models. R package version. 2009;1(4):1-24.\\u003c/li\\u003e\\n\\u003cli\\u003eYu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics: a journal of integrative biology. 2012;16(5):284-7.\\u003c/li\\u003e\\n\\u003cli\\u003eGeeleher P, Cox N, Huang RS. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PloS one. 2014;9(9):e107468.\\u003c/li\\u003e\\n\\u003cli\\u003eJiang P, Gu S, Pan D, Fu J, Sahu A, Hu X, et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nature medicine. 2018;24(10):1550-8.\\u003c/li\\u003e\\n\\u003cli\\u003eRooney MS, Shukla SA, Wu CJ, Getz G, Hacohen N. Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell. 2015;160(1-2):48-61.\\u003c/li\\u003e\\n\\u003cli\\u003eCabrita R, Lauss M, Sanna A, Donia M, Skaarup Larsen M, Mitra S, et al. Tertiary lymphoid structures improve immunotherapy and survival in melanoma. Nature. 2020;577(7791):561-5.\\u003c/li\\u003e\\n\\u003cli\\u003eChang Y-N, Zhang K, Hu Z-M, Qi H-X, Shi Z-M, Han X-H, et al. Hypoxia-regulated lncRNAs in cancer. Gene. 2016;575(1):1-8.\\u003c/li\\u003e\\n\\u003cli\\u003eWang X, Wang L, Xu W, Wang X, Ke D, Lin J, et al. Classification of osteosarcoma based on immunogenomic profiling. Frontiers in Cell and Developmental Biology. 2021;9:696878.\\u003c/li\\u003e\\n\\u003cli\\u003eKubista B, Klinglmueller F, Bilban M, Pfeiffer M, Lass R, Giurea A, et al. Microarray analysis identifies distinct gene expression profiles associated with histological subtype in human osteosarcoma. International orthopaedics. 2011;35:401-11.\\u003c/li\\u003e\\n\\u003cli\\u003eQian B-Z, Pollard JW. Macrophage diversity enhances tumor progression and metastasis. Cell. 2010;141(1):39-51.\\u003c/li\\u003e\\n\\u003cli\\u003eCersosimo F, Lonardi S, Bernardini G, Telfer B, Mandelli GE, Santucci A, et al. Tumor-associated macrophages in osteosarcoma: from mechanisms to therapy. International Journal of Molecular Sciences. 2020;21(15):5207.\\u003c/li\\u003e\\n\\u003cli\\u003eLiu H, Chen C, Liu L, Wang Z. A four-lncRNA risk signature for prognostic prediction of osteosarcoma. Frontiers in genetics. 2023;13:1081478.\\u003c/li\\u003e\\n\\u003cli\\u003eWang A-H, Jin C-H, Cui G-Y, Li H-Y, Wang Y, Yu J-J, et al. MIR210HG promotes cell proliferation and invasion by regulating miR-503-5p/TRAF4 axis in cervical cancer. Aging (Albany NY). 2020;12(4):3205.\\u003c/li\\u003e\\n\\u003cli\\u003eLi Z-Y, Xie Y, Deng M, Zhu L, Wu X, Li G, et al. c-Myc-activated intronic miR-210 and lncRNA MIR210HG synergistically promote the metastasis of gastric cancer. Cancer Letters. 2022;526:322-34.\\u003c/li\\u003e\\n\\u003cli\\u003eLi J, Wu Q-M, Wang X-Q, Zhang C-Q. Long noncoding RNA miR210HG sponges miR-503 to facilitate osteosarcoma cell invasion and metastasis. DNA and cell biology. 2017;36(12):1117-25.\\u003c/li\\u003e\\n\\u003cli\\u003eYing T, Dong J-l, Yuan C, Li P, Guo Q. The lncRNAs RP1-261G23. 7, RP11-69E11. 4 and SATB2-AS1 are a novel clinical signature for predicting recurrent osteosarcoma. Bioscience Reports. 2020;40(1).\\u003c/li\\u003e\\n\\u003cli\\u003eBu X, Liu J, Ding R, Li Z. Prognostic value of a pyroptosis-related long noncoding RNA signature associated with osteosarcoma microenvironment. Journal of Oncology. 2021;2021.\\u003c/li\\u003e\\n\\u003cli\\u003eYu S, Shao F, Liu H, Liu Q. A five metastasis-related long noncoding RNA risk signature for osteosarcoma survival prediction. BMC Medical Genomics. 2021;14(1):124.\\u003c/li\\u003e\\n\\u003cli\\u003eNantajit D, Lin D, Li JJ. The network of epithelial\\u0026ndash;mesenchymal transition: potential new targets for tumor resistance. Journal of cancer research and clinical oncology. 2015;141:1697-713.\\u003c/li\\u003e\\n\\u003cli\\u003eTam SY, Wu VW, Law HK. Hypoxia-induced epithelial-mesenchymal transition in cancers: HIF-1\\u0026alpha; and beyond. Frontiers in oncology. 2020;10:486.\\u003c/li\\u003e\\n\\u003cli\\u003eGajewski TF, Schreiber H, Fu Y-X. Innate and adaptive immune cells in the tumor microenvironment. Nature immunology. 2013;14(10):1014-22.\\u003c/li\\u003e\\n\\u003cli\\u003eJoyce JA, Fearon DT. T cell exclusion, immune privilege, and the tumor microenvironment. Science. 2015;348(6230):74-80.\\u003c/li\\u003e\\n\\u003cli\\u003eSpranger S, Gajewski TF. Tumor-intrinsic oncogene pathways mediating immune avoidance. Oncoimmunology. 2016;5(3):e1086862.\\u003c/li\\u003e\\n\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"Osteosarcoma, Long-noncoding RNA, Hypoxia, Prognosis, Biomarker\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-2945434/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-2945434/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003ch2\\u003eBackground\\u003c/h2\\u003e \\u003cp\\u003eRecently, several long-noncoding RNAs (lncRNAs) have been identified in hypoxia-associated cancer process including osteosarcoma (OS), enabling an adaptive survival under hypoxic stress conditions. However, hypoxia-related lncRNA signatures have rarely been reported. This study aimed to screen hypoxia-associated lncRNA signatures and assess their prognostic value in OS.\\u003c/p\\u003e\\u003ch2\\u003eMethods\\u003c/h2\\u003e \\u003cp\\u003eOS-related expression data were downloaded from the GEO and TARGET databases. Hypoxia-associated mRNAs were obtained from the HALLMARKHYPOXIA database. Hypoxia-associated lncRNAs were identified by correlation analysis with hypoxia-associated mRNAs. The tumor samples were clustered into different subtypes based on these lncRNAs, followed by immune microenvironment comparison. Prognostic hypoxia-associated lncRNAs were selected via univariate Cox regression analysis, and a prognostic signature was established using LASSO regression analysis. A risk score (RS) model was constructed, followed by pathway analysis, immunocorrelation analysis, and drug susceptibility prediction.\\u003c/p\\u003e\\u003ch2\\u003eResults\\u003c/h2\\u003e \\u003cp\\u003eThirty hypoxia-related lncRNAs were selected. The OS samples were classified into two subtypes based on lncRNAs. Nine immune cell types showed significantly different levels of infiltration between the two subtypes. Furthermore, five prognostic hypoxia-related lncRNAs were screened out through LASSO regression analyses, and an RS model was constructed. The high- and low-risk groups showed differences in prognosis, pathway, and drug susceptibility. The present study divided OS into two subtypes. A prognostic signature was constructed based on five hypoxia-related lncRNAs.\\u003c/p\\u003e\\u003ch2\\u003eConclusions\\u003c/h2\\u003e \\u003cp\\u003eThe study sucessfully identifies five hypoxia-related lncRNAs and this lncRNA signature may have significant prognostic value in OS.\\u003c/p\\u003e\",\"manuscriptTitle\":\"A hypoxia-related five-lncRNA signature predicts osteosarcoma prognosis\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2023-05-19 14:58:45\",\"doi\":\"10.21203/rs.3.rs-2945434/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"9c1ab936-e8a3-4f57-8d4c-efb54a88fffd\",\"owner\":[],\"postedDate\":\"May 19th, 2023\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2023-06-16T06:29:19+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2023-05-19 14:58:45\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-2945434\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-2945434\",\"identity\":\"rs-2945434\",\"version\":[\"v1\"]},\"buildId\":\"-HB7Z8yhvgn0wM9Nzuekk\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}