EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Machine learning–based integrative analysis identifies CXCL13-driven tertiary lymphoid structures as favorable immune and prognostic features in osteosarcoma

Jie Jiang, Jiuhui Xu, Lu Xie, Yiyang Yu, Xin Sun, Xiaojiao Sun, Shen Yang, Huanmin Wang, Tingting Ren, Xiaodong Tang

Cellular Oncology · 16.05.2026 · 49 (4)

DOI: 10.1007/s13402-026-01226-1 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Cellular Oncology, online erschienen am 16.05.2026. Verfasst von Jie Jiang und 9 weiteren Autor:innen. Beteiligte Einrichtung: Department of Musculoskeletal Tumor, People's Hospital, Peking University, Beijing, 100044, China. Hauptthemen laut MEDLINE: Osteosarcoma, Chemokine CXCL13, Tertiary Lymphoid Structures, Machine Learning. Die Arbeit nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Capital's Funds for Health Improvement and Research (CFH), Noncommunicable Chronic Diseases-National Science and Technology Major Project.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Osteosarcoma, an aggressive bone malignancy with limited response to immunotherapy, remains a major clinical challenge. Tertiary lymphoid structures (TLS), as organized ectopic lymphoid aggregates, play a pivotal role in modulating antitumor immune responses. However, their landscape, clinical significance, and molecular determinants in osteosarcoma remain largely unexplored.

Methods

TLS were identified in osteosarcoma tissues by immunohistochemistry and multiplex immunofluorescence. Using transcriptomic data from the TARGET, GEO (GSE21257, GSE16091, GSE39055), and PKUPH cohorts, we developed a TLS-based prognostic index (TLSPI) through integrative machine-learning modeling. Functional enrichment, immune infiltration, and immunotherapy response prediction analyses were performed to characterize TLSPI-related biological features. The role of the key TLS-associated gene CXCL13 was validated through immunohistochemistry, single-cell transcriptomic profiling, and in vitro functional experiments.

Results

TLS were detected in osteosarcoma and were significantly associated with improved overall survival ( p = 0.024). Differential expression and enrichment analyses revealed that TLSPI-related genes were primarily involved in immune activation and antigen presentation pathways. Immune profiling revealed that the low-TLSPI group was characterized by greater immune infiltration and higher expression of immune checkpoint molecules, as well as lower TIDE scores, which indicated greater predicted sensitivity to immunotherapy. CXCL13 was identified as a key TLS-associated gene whose high expression correlated with improved survival. Immunohistochemical analysis confirmed the prognostic significance of CXCL13, while single-cell and functional assays demonstrated its association with B-cell activation within the osteosarcoma microenvironment.

Conclusion

This study provides a comprehensive characterization of TLS in osteosarcoma and establishes a TLS-based prognostic index with potential clinical applicability. CXCL13 may serve as a critical mediator linking TLS formation and antitumor immunity, providing potential therapeutic implications for osteosarcoma immunotherapy.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

OsteosarcomaChemokine CXCL13Tertiary Lymphoid StructuresMachine LearningBone NeoplasmsHumansPrognosisTumor MicroenvironmentGene Expression Regulation, NeoplasticFemaleGene Expression ProfilingCell Line, TumorMale

Stichwörter der Autor:innen

CXCL13Machine learningOsteosarcomaTertiary lymphoid structuresTumor immune microenvironment

Substanzen

Chemokine CXCL13CXCL13 protein, human

Autor:innen und Institutionen

  • Department of Musculoskeletal Tumor, People's Hospital, Peking University, Beijing, 100044, China.
  • Department of Surgical Oncology, Beijing Children's Hospital, Capital Medical University, National Center for Children's Health, Beijing, 100045, China.
  • MOE Key Laboratory of Major Diseases in Children, Beijing Children's Hospital, Capital Medical University, National Center for Children's Health, Beijing, 100045, China.
  • Department of Musculoskeletal Tumor, People's Hospital, Peking University, Beijing, 100044, China. tumorcenter@163.com.
  • Department of Musculoskeletal Tumor, People's Hospital, Peking University, Beijing, 100044, China. tang15877@163.com.

Förderung

  • Capital's Funds for Health Improvement and Research (CFH)
  • Noncommunicable Chronic Diseases-National Science and Technology Major Project

Zitieren

APA

Jiang, J., Xu, J., Xie, L., Yu, Y., Sun, X., Sun, X., Yang, S., Wang, H., Ren, T., & Tang, X. (2026). Machine learning–based integrative analysis identifies CXCL13-driven tertiary lymphoid structures as favorable immune and prognostic features in osteosarcoma. Cellular Oncology, 49(4). https://doi.org/10.1007/s13402-026-01226-1

Vancouver

Jiang J, Xu J, Xie L, Yu Y, Sun X, Sun X, et al. Machine learning–based integrative analysis identifies CXCL13-driven tertiary lymphoid structures as favorable immune and prognostic features in osteosarcoma. Cellular Oncology. 2026;49(4). doi:10.1007/s13402-026-01226-1

BibTeX
@article{Jiang2026,
  author = {Jiang, Jie and Xu, Jiuhui and Xie, Lu and Yu, Yiyang and Sun, Xin and Sun, Xiaojiao and Yang, Shen and Wang, Huanmin and Ren, Tingting and Tang, Xiaodong},
  title = {Machine learning–based integrative analysis identifies CXCL13-driven tertiary lymphoid structures as favorable immune and prognostic features in osteosarcoma},
  journal = {Cellular Oncology},
  year = {2026},
  volume = {49},
  number = {4},
  doi = {10.1007/s13402-026-01226-1}
}
Kostenlos starten →