Osteosarcoma is a common malignancy that can affect children and adolescents. Even though advancements have been made in treatment options and surgical procedures to improve local control rates and quality of life for patients with osteosarcoma, survival rates are as low as 20% or lower due to metastasis and recurrence(Martins-Neves et al., 2022, Bielack et al., 2009, Zeng et al., 2022a). Identification of high-risk patients in a timely manner is critical for improving their prognosis. Personalized/precision treatment and more comprehensive follow-up plans are also required. Consequently, the discovery of specific and novel predictive biomarkers for osteosarcoma is of utmost clinical significance. Recent studies indicate that NRG dysregulation is crucial in osteosarcoma occurrence and progression as it interferes with the cell death– cell division balance(Peterson et al., 2021, Subramanian et al., 2015). Clinical and NRG expression data of patients with osteosarcoma were retrieved from the GEO and TARGET databases. Two NET-related molecular subtypes were identified using consensus clustering analysis. PCA and tSNE plots of NRG expression revealed Clusters 1 and 2. The K-M plot revealed that Cluster 1 was associated with a poor outcome, whereas Cluster 2 was associated with a favorable outcome. The DEGs between both clusters were considerably enriched in immune-associated functions and pathways. Following GSEA, Cluster 2 was shown to be enriched in several immune-associated pathways. Increased immunogenicity contributes to a favorable prognosis among patients with cancer via the mechanism of activating the immune response. The ESTIMATE algorithm was employed to evaluate the TME, which is made up of tumor and stromal cells(Kessenbrock et al., 2010) and is crucial for tumorigenesis(Thakur et al., 2020), progression(Hinshaw and Shevde, 2019), and metastasis(Huang et al., 2018). Cluster 2 had a substantially higher stromal content and infiltration of immune cells, with a lower tumor purity compared to Cluster 1. Additionally, Cluster 2 exhibited elevated levels of B, NK, and CD8 + T cells, suggesting they may trigger an immune response. A higher CD8 + T cell infiltration is associated with a favorable prognosis(Galon et al., 2006, Koelzer et al., 2014). According to Chen et al.(Chen et al., 2021), cholangiocarcinoma characterized by high B-cell infiltration levels are associated with favorable prognoses. In addition, NK cells contribute significantly to antitumor immunity and are associated with better prognoses in several types of cancer(Shen et al., 2022, Li et al., 2022a, Vivier et al., 2011). Consequently, Cluster 2 osteosarcoma patients had a favorable prognosis compared to Cluster 1 patients due to an elevated immune effector cell infiltration level in the TME. Even though immunotherapy is increasingly being recognized as an effective and promising therapeutic option for osteosarcoma(Qu et al., 2022, Peng et al., 2022), no biomarker is currently available. Patients can be stratified according to biomarkers to receive different treatments. In light of the different immune landscapes between both clusters, we compared the immune checkpoint- and HLA-associated gene expression levels. A surprising result of these experiments was that immune checkpoints were associated with most of these genes' expression levels. Furthermore, patients with osteosarcoma in Cluster 2 stood a higher chance of benefiting from treatment as a result of a significant difference in the expression of HLA between Clusters 2 and 1, particularly in PD1 and PDL1. Further studies are necessary to determine the effect of NETs-related molecular subtypes on immunotherapy efficacy in osteosarcoma.
The prognostic NRG signature was established using LASSO. High risk scores were associated with unfavorable OS rates in both GSE21257 and TARGET cohorts. The NRG signature had excellent predictive power in both cohorts, as per the AUCs and time-dependent ROC curves. Because of the rapid advancement in bioinformatics, the most recent study highlighted various prognostic gene signatures for osteosarcoma from datasets that are open to the public. Yang et al.(Yang et al., 2021) found that osteosarcoma patients with immune-related gene signatures have favorable prognoses. According to Shi et al.(Shi et al., 2020), metastasis-related genes differentiate primary osteosarcoma from metastatic osteosarcoma and were used to establish a new osteosarcoma metastasis signature. Our NRG signature had a superior prognosis prediction performance than these other signatures. Additionally, the risk score was found to predict prognosis for osteosarcoma patients in an independent manner. We accurately predicted the prognoses of osteosarcoma patients using a nomogram established by integrating clinical parameters with risk scores. ROC curves and calibrations of the nomogram also demonstrated its potential as a prediction model.
Patients who have advanced osteosarcoma are at a higher risk of developing distant metastases, especially in the lungs, resulting in poor clinical outcomes(Lak et al., 2022). When stratifying risk and deciding on a course of treatment, physicians continue to place the greatest emphasis on osteosarcoma's metastatic behavior(Zeng et al., 2022b). We studied the relationship between clinical parameters and signatures in the present study. Higher metastasis rates were associated with higher risk scores. Our NRG signature may be used to predict osteosarcoma metastasis. There are a few limitations to the present research. The sample size employed was too small due to the low morbidity of osteosarcoma, and larger sample sizes should be used to validate our findings. In addition, our study was limited to bioinformatics, and further experiments should be performed to confirm our results. Thirdly, large-scale clinical trials and prospective studies should be used to validate the NRG signature. Additionally, further investigations are necessary to determine the biological function of NRGs in osteosarcoma.
Osteosarcoma patients were assigned to two groups based on their expression level of NRGs. Patients in Cluster 2 demonstrated greater immunogenicity, immune cell infiltration degree, and a better response to immunotherapy than Cluster 1 patients. Our NRG signature also accurately predicts the OS time for patients with osteosarcoma. Our research also has limitations. The sample size of this study was small, and although the prognosis of osteosarcoma was predicted by extracting NRG, laboratory validation was not performed, so further studies are needed to consolidate our results.