Background: Skull base chordoma is a rare and aggressive tumour of the bone that has a high likelihood of recurrence. The fundamental differences in single cells between primary and recurrent lesions remain poorly understood, impeding development of effective treatment approaches.Methods: To obtain an understanding of the differences in single cells between primary and recurrent chordomas, we performed single‐cell RNA sequencing and T‐cell/B‐cell receptor (BCR) sequencing. This allowed us to delineate the differences between the two types of tumour cells, tumour‐infiltrating lymphocytes, myeloid cells, fibroblasts and B cells. Copy number variants (CNVs) were detected and compared between the tumour types to assess heterogeneity. Selected samples were subjected to immunohistochemistry to validate protein expression. Fluorescence in situ hybridisation experiments, Transwell assays and xenograft mouse models helped verify the role of fibronectin 1 (FN1) in chordoma.Results: Promoting natural killer (NK) cell and CD8_GZMK T‐cell function or inhibiting the transformation of CD8_GZMK T cells to CD8_ZNF683 T cells and promoting the transformation of natural killer T (NKT) cells to NK cells are promising strategies for preventing chordoma recurrence. Additionally, inhibiting the M2‐like activity of tumour‐associated macrophages (TAMs) could be an effective approach. Antigen‐presenting cancer‐associated fibroblasts (apCAFs) and dendritic cells (DCs) with high enrichment of the antigen‐presenting signature were enriched in primary chordomas. There were fewer plasma cells and BCR clonotypes in recurrent chordomas. Remarkably, FN1 was upregulated, had more CNVs, and was more highly secreted by tumours, macrophages, CD4 T cells, CD8 T cells and fibroblasts in recurrent chordoma than in primary chordoma. Finally, FN1 enhanced the invasion and proliferation of chordomas in vivo and in vitro.Conclusion: Our comprehensive picture of the microenvironment of primary and recurrent chordomas provides deep insights into the mechanisms of chordoma recurrence. FN1 is an important target for chordoma therapy.

BACKGROUND: Lower-grade gliomas (LGGs) vary widely in terms of the patient’s overall survival (OS). There is no current, valid method that could exactly predict the survival. The effects of intratumoral immune infiltration on clinical outcome have been widely reported. Thus, we aim to develop an immune infiltration signature to predict the survival of LGG patients. METHODS: We analyzed 1216 LGGs from 5 public data sets, including 2 RNA sequencing data sets and 3 microarray data sets. Least absolute shrinkage and selection operator (LASSO) Cox regression was used to select an immune infiltration signature and build a risk score. The performance of the risk score was assessed in the training set (329 patients), internal validation set (140 patients), and 4 external validation sets (405, 118, 88, and 136 patients). RESULTS: An immune infiltration signature consisting of 20 immune metagenes was used to generate a risk score. The performance of the risk score was thoroughly verified in the training and validation sets. Additionally, we found that the risk score was positively correlated with the expression levels of TGF-β and PD-L1, which were important targets of combination immunotherapy. Furthermore, a nomogram incorporating the risk score, patient’s age, and tumor grade was developed to predict the OS, and it performed well in all the training and validation sets (C-index: 0.873, 0.881, 0.781, 0.765, 0.721, and 0.753). CONCLUSION: The risk score based on the immune infiltration signature has reliable prognostic and predictive value for patients with LGGs and is a potential biomarker for the cotargeting immunotherapy. FUNDING: This work was supported by The National Natural Science Foundation of China (grant nos. 81472370 and 81672506), the Natural Science Foundation of Beijing (grant no. J180005), the National High Technology Research and Development Program of China (863 Program, grant no. 2014AA020610), and the National Basic Research Program of China (973 Program, grant no. 2014CB542006).