Neuroblastoma (NB) is the most common extracranial solid malignancy in children and accounts for about 10 % of childhood cancer deaths [1], [2]. NB presents highly heterogeneous clinical manifestations, and its outcomes vary from spontaneous regression to rapid progression [3], [4]. Tumour progression is a multi-step process with genetic and epigenetic changes [5]. The biological characteristics of tumour cells have been considered important markers for prognosis; however, accumulating evidence has shown that the cellular and matrix components around tumour cells, which constitute the tumour microenvironment (TME) [6], play roles as well that cannot be ignored when predicting a patient’s clinical outcome and choosing personalized treatment. The TME affects the biological behaviour of tumour cells in a variety of ways such as stimulating proliferation, promoting metastasis and inhibiting apoptosis [6], [7]; On the other hand, tumour cells shape the TME for their own survival advantages such as promoting angiogenesis and expressing high levels of immune checkpoint molecules [8], [9].

Accordingly, in addition to traditional treatments such as surgery, chemotherapy and radiotherapy, emerging treatments targeting the TME are being adopted clinically with the immune checkpoint inhibitors (ICIs) being remarkable successful in several adult tumours. However, most paediatric cancer patients, including NB, do not benefit from ICIs therapy, yet [10], [11]. NB cells have many intrinsic immune escape mechanisms, including low tumour mutation burden (TMB) to decrease neoantigens that may induce anti-tumour adaptive immunity [9], [12]. Furthermore, NB cells are surrounded by an immunosuppressive TME characterized by high infiltration levels of tumour-associated macrophages, myeloid-derived suppressor cells and Treg cells [13], which leads to a low response to ICIs therapy. Therefore, besides the urgent need for more effective immunotherapeutic strategies, the most important challenge we face is still the evaluation and quantification of TME in NB patients and a reliable biomarker that predicts a NB patient’s response to immunotherapy.

Currently, the International Neuroblastoma Risk Group (INRG) proposes that INSS stage, age at diagnosis, NMYC amplification status, histologic category, grade of tumour differentiation and DNA ploidy to be used as the main factors for the prognosis of NB [14]. The INRG factors are widely used by the neuroblastoma community. However, the factors are mostly surrogates for the underlying tumour biology. Therefore, a prognosis biomarker that reflects more accurately the biological characteristics of NB is still lacking.

The accumulation of genomic profiling over the past decades provides us an opportunity to further understand the developmental origins of NB [15], the complexity and heterogeneity of the TME of NB, and to discover novel biomarkers that can be used to predict a patient’s prognosis and response to therapies [16]. To date, comprehensive analyses that include both NB tumour and the TME in the same studies are rare. In this study, we first identified four subtypes of NB based on the TME landscapes and the distinct biological behaviour characteristics of tumour cells, and constructed a double scoring system, an Immune Score and a Proliferation Score, to quantify both the immune infiltration levels of the TME and the malignancy degree of NB. The double scoring system was further validated in neuroblastoma as well as some other cancers such as non-small cell lung cancer and melanoma. We then employed the double scoring system to screen for drugs to treat NB, and it turned out that axitinib and BI-2536 were good candidates. Both in vivo and in vitro studies showed that axitinib induced pyroptosis of the NB cells and inhibited tumour growth on a syngeneic NB mouse model through Caspase-3/GSDME pathway, and pyroptosis has been shown to activate anti-tumor immunity in certain systems (colorectal cancer, triple negative breast cancer, etc) [17], [18], [19]. In addition, we found that BI-2536 induced cycle arrest of NB cells at S phase. In summary, the double scoring system constructed in this study, including the Immune Score and the Proliferation Score, is a powerful tool to predict prognosis and the responses to immunotherapy, and may help the clinical practitioners to choose personalized treatment for NB patients.