尤因肉瘤自发转移PDX模型的建立与分析

Although improvements in local therapy have increased the 5-year survival rate for localized Ewing sarcoma (EWS) from less than 20% to 70%, little progress has been made in the treatment of metastatic disease, which has a 5-year survival rate of less than 30%. A major challenge in the field is a lack of preclinical models that can recapitulate spontaneous metastatic disease and can be used to better understand the disease and identify new vulnerabilities. In this study, we present and characterize a murine model of spontaneous distant EWS metastasis derived from human patient-derived xenografts (PDX) in two different mouse backgrounds that mimics the clinical progression of the human disease. A panel of seven molecularly diverse patient-derived xenograft (PDX) models (SJ18, SJ17, S049, NCH1, NCH4, PDMR-098, PDMR-077) were injected orthotopically into the gastrocnemius muscle in the left hind leg of athymic nude and NOD SCID gamma (NSG) mice. Once primary tumors reached 1500 mm 3 , hind limb amputation survival surgery was performed, and we observed animals for the development of distant spontaneous metastases. EWS PDXs formed spontaneous macrometastases in multiple sites including lymph nodes, lung, liver, and kidney in both NSG and nude mice. Each PDX model exhibited a distinct pattern of macrometastasis formation, based on site, metastasis frequency, and mouse strain. Immunohistochemical analysis for CD99 positivity revealed the presence of micrometastases in some locations where no macrometases were evident. The highest frequency of distant metastases (macro- plus micro-) was seen in PDMR-098 and SJ18 models in NSG mice (75% and 73.3% respectively); SJ18 most frequently metastasized to lung, whereas PDMR-098 most frequently metastasized to liver and mesenteric lymph nodes. Comparing results in NSG versus nude mice, we observed variations in metastasis frequency and site preference, with NSG mice demonstrating a higher rate of metastasis overall. Here we describe a preclinical model of spontaneous distant EWS metastasis that recapitulates the characteristics of human disease. The site and frequency of metastases vary based on the specific PDX model as well as the mouse background, highlighting that modeling metastasis formation is a multifactorial process driven by complex interactions between the tumor and host. Metabolomic profiling conducted to compare primary and macrometastatic tumors between models and metastatic sites revealed differences in multiple metabolic pathways. Future studies will focus on identifying potentially targetable metabolic vulnerabilities that could be used in the development of treatments for metastatic EWS.