Brain invasion, limited drug delivery due to the blood-brain barrier (BBB), molecular/cellular heterogeneity, and treatment resistance render glioblastoma (GBM) virtually incurable. Surgical resection, a cornerstone of GBM treatment, directly impacts tumor biology and leaves behind residual invasive disease (RID), which is also biologically distinct from the bulk tumor that is removed. The RID seeds tumor recurrence and drives poor outcomes in patients. RID and recurrent tumor biology have been challenging to study longitudinally in humans, due the limited number of patients who undergo repeat biopsy/resection. Similarly, existing pre-clinical tumor models rarely include surgery, and current resection methods are unable to perform precise, repeated, and controlled biopsy/resection with tissue collection and animal survival. This gap has left tumor evolution, treatment responses, and potential therapeutic targets within the GBM RID and recurrent tumor largely unexplored. We developed the Murine Intracranial Surgery (MIS) system, a scalable and reproducible approach that integrates stereotactic localization, controlled biopsy and resection, and sterile, high-viability tissue collection, all with animal survival. Using the MIS in five different murine orthotopic brain tumor models, we find that extent of resection is associated with survival, recurrent growth patterns vary by model, and transcriptional profiles of matched murine primary-recurrent patient-derived xenograft tumors closely mirror matched human primary-recurrent GBM biology. Finally, we integrate the MIS RID model with peri-resection microbubble-enhanced focused ultrasound for BBB opening. In this advanced drug-device combination model we performed safe, repeated, and volumetrically contoured MB-FUS treatments targeted to the peri-resectional invasive margins of the tumor. This new model mirrors the BBB opening, acoustic emissions-based dosing, imaging, and histologic features observed in human adjuvant MB-FUS treatments. Furthermore, we demonstrate enhanced localized delivery of therapeutic surrogates to the non-enhancing peri-resectional RID regions, including a small molecule dye and nanoparticles. Finally, we demonstrate a relationship between subspot level acoustic emissions dose and new T1c MRI signal in adjuvant peri-resectional MB-FUS treatments. The technological and conceptual advances presented in this Dissertation permit key insights into the evolution of murine GBM RID and establish a platform to explore longitudinal biology and therapeutic responses, assess neoadjuvant and adjuvant therapies, identify novel therapeutic targets relevant to the RID, and assess advanced device-drug combinations in the adjuvant setting of murine GBM.