With the widespread adoption of mammograms for cancer detection, modern research has principally pivoted towards a focus on how to predict recurrence and ultimately to prevent it. Approximately 30% of breast cancer patients lacking any clinical or pathological signs of metastasis have disseminated tumor cells present at the time of their diagnosis. Thus, residual disease at the primary tumor site after systemic therapy could be viewed as a surrogate for micro-metastasis elsewhere in the body that may have evaded the drug and could potentially return as recurrent disease. Given that residual disease is an important determinant of clinical outcome, it is important to predict the potential for and mitigate distant recurrence. Through the emergence of complex and robust animal models, like genetically engineered mouse models, we can study common oncogenes and key features of dormancy, like regression to a non-proliferative state following therapy and ultimately recurrence, similar to human cancers to better inform future therapy options for these resilient diseases. We are creating a combination of metabolic and vascular imaging strategies and algorithms to provide an unprecedented window into metabolic reprogramming of residual disease and recurrence and facilitate understanding of how these tumors evade therapeutic stress. This can inform the development and application of new therapies to mitigate tumor recurrence or eliminate residual disease altogether.