TDERS, an exosome RNA-derived signature predicts prognosis and immunotherapeutic response in clear cell renal cell cancer: a multicohort study

Tumor-derived exosomes are involved in tumor progression and immune invasion and might function as promising noninvasive approaches for clinical management. However, there are few reports on exosom-based markers for predicting the progression and adjuvant therapy response rate among patients with clear cell renal cell carcinoma (ccRCC). The signatures differentially expressed in exosomes from tumor and normal tissues from ccRCC patients were correspondingly deregulated in ccRCC tissues. We adopted a two-step strategy, including Lasso and bootstrapping, to construct a novel risk stratification system termed the TDERS (Tumor-Derived Exosome-Related Risk Score). During the testing and validation phases, we leveraged multiple external datasets containing over 2000 RCC cases from eight cohorts and one inhouse cohort to evaluate the accuracy of the TDERS. In addition, enrichment analysis, immune infiltration signatures, mutation landscape and therapy sensitivity between the high and low TDERS groups were compared. Finally, the impact of TDERS on the tumor microenvironment (TME) was also analysed in our single-cell datasets. TDERS consisted of 12 mRNAs deregulated in both exosomes and tissues from patients with ccRCC. TDERS achieved satisfactory performance in both prognosis and immune checkpoint inhibitor (ICI) response across all ccRCC cohorts and other pathological types, since the average area under the curve (AUC) to predict 5-year overall survival (OS) was larger than 0.8 across the four cohorts. Patients in the TDERS high group were resistant to ICIs, while mercaptopurine might function as a promising agent for those patients. Patients with a high TDERS were characterized by coagulation and hypoxia, which induced hampered tumor antigen presentation and relative resistance to ICIs. In addition, single cells from 12 advanced samples validated this phenomenon since the interaction between dendritic cells and macrophages was limited. Finally, PLOD2, which is highly expressed in fibro- and epi‑tissue, could be a potential therapeutic target for ccRCC patients since inhibiting PLOD2 altered the malignant phenotype of ccRCC in vitro. As a novel, non-invasive, and repeatable monitoring tool, the TDERS could work as a robust risk stratification system for patients with ccRCC and precisely inform treatment decisions about ICI therapy.