Background
The inadequacy of chronological age as an indicator of biological aging has spurred the development of biomarkers that quantify hallmarks of the aging process. Stem cell exhaustion (SCE) is a recognized hallmark of cellular senescence; however, its clinical significance in nasopharyngeal carcinoma (NPC) remains unclear, and scalable tools for its assessment are lacking.
Methods
We computed an SCE score from transcriptomic data of 181 non-metastatic NPC patients. Its relationship with the tumor immune microenvironment was characterized using bioinformatic and spatial transcriptomic analyses. To develop a clinically applicable tool, we derived a Circulating Stemness Exhaustion Signature (CSES) from routine blood parameters via an artificial neural network. The CSES was validated in three independent cohorts: an internal cohort (n = 492), an external cohort (n = 122), and an immunotherapy-treated cohort (n = 42).
Results
A high SCE score independently predicted improved overall survival (HR = 0.03, 95% CI: 0.00-0.65, P = 0.025) and delineated an immunologically active TIME characterized by enhanced T-cell cytotoxicity and elevated MHC-I expression. The CSES strongly correlated with the SCE score (r = 0.844, P < 0.001) and consistently stratified patient survival in validation cohorts. In the immunotherapy cohort, a high CSES predicted a significant survival benefit (difference in restricted mean survival time at 36 months: 7.47 months, P = 0.039) and was independently associated with superior treatment response.
Conclusion
Our study establishes SCE as a critical determinant of the immune niche and prognosis in NPC. Furthermore, we introduce the CSES as a practical, non-invasive biomarker capable of guiding prognostic stratification and informing immunotherapy decisions.