Metastatic neuroblastoma (NB) is one of the most challenging childhood cancers to treat, with a high rate of relapse despite aggressive therapies. Most NB cells express disialoganglioside GD2, a tumor-specific antigen with limited expression in normal tissues. While GD2-targeting monoclonal antibodies have shown promising clinical success, a significant number of patients still experience relapse. This is thought to be due, in part, to heterogeneous or low GD2 expression, which complicates the effectiveness of immunotherapies. However, studying GD2-targeted treatments and metastasis in NB is hindered by a lack of suitable in vitro models. Current 2D culture systems are limited in their ability to mimic the complex environment of metastatic tumors, and murine NB cells typically lose GD2 expression when cultured. To overcome these challenges, I have developed a novel 3D cell culture system known as "tumorspheres." This model allows NB cells to grow in a more physiologically relevant environment and better replicates the conditions of tumor metastasis. Preliminary results demonstrate that tumorspheres retain high GD2 expression and show significant growth and motility within a matrix scaffold, mimicking in vivo tumor behavior more accurately than traditional 2D models. Importantly, these tumorspheres can be formed from a mix of GD2-positive and GD2-negative cells, enabling the study of antigenic heterogeneity and its impact on therapeutic resistance. Additionally, tumorspheres can be implanted subcutaneously into mice to promote in vivo tumor growth, further enhancing their utility as a model for studying neuroblastoma metastasis and immunotherapy. This new 3D model offers a more representative platform for investigating GD2-targeted therapies and tumor metastasis, with the potential to advance our understanding of treatment resistance and lead to more effective strategies for combating metastatic neuroblastoma.