Using a Gene Network of Pyroptosis to Quantify the Responses to Immunotherapy and Prognosis for Neuroblastoma Patients

BACKGROUND: Pyroptosis, as an inflammatory form of cell death, is involved in many physiological and pathological processes. Neuroblastoma is the most common extra-cranial solid tumor in children. In this study, the relationship between pyroptosis and tumor microenvironment in neuroblastoma was systematically studied. METHODS: We integrated four datasets of neuroblastomas. Through robust clustering of the mRNA expression profiles of 24 pyroptosis-related genes, a total of three pyroptosis patterns were identified. We then constructed a novel scoring method named as pyroscore to quantify the level of pyroptosis in neuroblastoma. Multi-omics data and single-cell RNA sequencing were used to accurately and comprehensively evaluate the effectiveness of pyroscore. Clinical data sets were used to evaluate the use of pyroscore to predict the responsiveness of immune checkpoint treatment. RESULTS: High pyroscore was associated with good prognosis, immune activation, and increased response to checkpoint blockade immunotherapy. Multivariate Cox analysis revealed that the pyroscore was an independent prognostic biomarker and could increase the accuracy of clinical prediction models. Etoposide, a drug picked up by our analysis, could increase the sensitivity of neuroblastoma cells to pyroptosis. External verification using four cohorts of patients who had received immunotherapy showed that high pyroscore was significantly associated with immunotherapy treatment benefit. CONCLUSIONS: Taken together, this study revealed that pyroptosis-related gene network could quantify the response of neuroblastoma to immune checkpoint blockade therapy and prognosis, and it may be helpful for clinical practitioners to choose treatment strategies for neuroblastoma patients.