Osteosarcoma is the most frequent primary bone tumor in the world, as well as the third most common malignancy in children and adolescents.[29] Despite the emergence of multiple treatment strategies with modern medical advances, 5-year survival rates for patients with osteosarcoma have remained unimproved over the past decades, making the development of effective risk stratification methods and individualized targeted treatment strategies essential.[30, 31] Fig. 1 shows the flow chart of our experiment. The interaction between malignancy and the coagulation system. In cancer patients, there are many coagulation abnormalities, therefore in this experiment we identified two molecular subtypes that exhibit significantly different coagulation system landscapes.[8] Individuals with fair prognosis had better immune status, higher ESTIMATE scores and immune scores, and lower tumor purity compared to individuals with poor prognosis. This also demonstrates that the coagulation system not only affects the prognosis of osteosarcoma, but also the immune microenvironment of osteosarcoma. In addition, we created a predictive risk model based on CRGs that effectively foreseen the prognosis of osteosarcoma patients. Our findings may help develop drugs for osteosarcoma and assist physicians in making more informed treatment choices.
Consensus clustering is a dependable method for categorizing data into various subgroups using gene expression matrices and CRG.[32] We initially used consensus clustering to identify two molecular subgroups that varied considerably in terms of overall survival. Immunological and functional studies were used to investigate the involvement of the coagulation system in osteosarcoma.
It is well known that tumor development is often accompanied by changes in the surrounding stroma, and thus TIME plays a key role in patient prognosis, while immune cells constitute the tumor stroma as key components.[33] In addition, abnormalities in the metabolic state of tumor cells can also affect the metabolism of TIME. ESTIMATE is a computerized algorithm that can be used to infer the level of infiltration of immune and stromal cells in tumor tissue based on expression profiles.[34] In the results of the ESTIMATE algorithm we obtained the immune score, which shows the immune fraction in the tumor sample and gives a picture of the tumor immune microenvironment; in addition we obtained the tumor purity score, defined as the percentage of malignant cells in the tumor tissue, which is strongly correlated with the prognosis. In recent years, there has been a consistent indication that high immune scores and little tumor purity are related with a bad prognosis in osteosarcoma. As a result, we used ESTIMATE to assess the tumor immune microenvironment in both groups. According to our findings, patients with a better prognosis had higher immunological ratings and lower purity, which is consistent with earlier research.[35] We also used TIMER, ssGSEA, CIBERSORT, QUANTISEQ, XCELL, EPIC, and MCPCOUNTER to examine the immunological state of two molecular subpopulations. TIMER is a method that provides a quantitative assessment of six immune subpopulations that infiltrate tumors. [36]The quantity of five of the six immune cells was considerably greater in five of them, consistent with the ESTIMATE results, and showed a downregulation of the immunological landscape in cluster 2. The ssGSEA examination revealed the richness of 28 immune-related cells, and the results showed that patients in cluster 2 had a comparatively poor immunological state, validating the ESTIMATE and TIMER findings. Not only that, but we found the same results using a variety of additional analysis approaches. Immune checkpoints are a group of molecules produced on immune cells that limit the degree of immune activation and play a vital role in avoiding the start of autoimmunity (abnormalities in immune function that cause assaults on normal cells).[37] We simultaneously obtained the same results by performing immunomodulatory point gene analysis on both groups and concluded that cluster 1 had better expression of immune checkpoint, regulatory point genes than cluster 2.Human leukocyte antigen (HLA) plays a crucial role in tumorigenesis, development. [38] Consequently, we analyzed the expression of HLA family genes for both groups of samples, and there is no doubt that cluester1 was stronger than cluster2 in the expression of almost all HLA family genes. The immune cytolytic activity (CYT) was associated with better survival[39], and we concluded that the immune profile of cluster 1 was better than that of cluster 2 by comparing the CYT of the two groups. In conclusion, it is plausible to believe that poor prognosis is connected with low immunological scores and immune state.
Next, differential analysis between the two subgroups was performed and functional analysis using differential genes was performed to explore the underlying biological mechanisms. Based on the identified DEGs, GO analysis, KEGG analysis and PPI network construction suggest a possible role of the coagulation system in tumorigenesis and progression of osteosarcoma by affecting the extracellular matrix as well as the immune response.
Immediately after, we constructed a prognostic model for osteosarcoma based on the results of Lasso regression analysis. To more accurately assess the accuracy of the prognostic model, Kaplan-Meier survival curves and ROC curves of the predictive model at 1, 3, and 5 years were plotted in this study, and the corresponding AUC values were calculated. The results showed significant differences in survival between the high- and low-risk subgroups, and the AUC indicated that the prognostic model was efficient in predicting patient survival. With the help of COX regression analysis, risk score and the occurrence of distant metastases were independent factors affecting patient prognosis. Finally, a Nomogram was drawn using risk score, age, gender, and metastatic status, through which we could visually predict the 0ne-year, three-year, and five-year survival rates of different patients based on their risk scores. Finally, in order to obtain whether the genes affecting the prognostic profile of osteosarcoma play the same role in other cancers, we selected the gene with the highest HR, ITIH1, for a pan-cancer analysis. We discovered that this gene was highly expressed in some of the tumors and not only that it also had a significant effect on the prognosis of some of the genes.