Abstract Acute T-cell lymphoblastic leukemia (T-ALL) is a pediatric malignancy in which the MYC oncogene is frequently overexpressed. Although, T-ALL responds to standard therapies, in 20% of cases, treatment fails to induce sustained remission, resulting in reduced patient survival. Inhibition or restoration of the mutant gene products that drive tumorigenesis can reverse the aberrant cell growth and result in elimination of tumor cells through the phenomenon of oncogene addiction. In order to recapitulate tumor regression and recurrence, we have developed conditional mouse models of MYC-dependent T-ALL, thereby allowing for the investigation of therapeutic resistance. Recently, we have shown oncogene addiction in our MYC-driven murine model of T-ALL occurs through both tumor cell-intrinsic mechanisms, such as cellular senescence and apoptosis, and tumor cell-extrinsic, host-dependent mechanisms, including inhibition of angiogenesis, immune system-dependent tumor clearance, and autocrine/paracrine survival signaling. We now present evidence that loss of either p53 or p19ARF facilitates tumor recurrence through distinct mechanisms. p53 utilizes host-dependent mechanisms by allowing tumors to maintain angiogenesis, while p19ARF employs tumor-cell intrinsic mechanisms to circumvent senescence upon oncogene inactivation in vivo. To further interrogate the mechanisms that underlie p53- and p19ARF-dependent tumor recurrence, we performed gene expression analysis on MYC wild-type, MYC/p53-/-, and MYC/p19-/- tumors before and after MYC inactivation. Utilizing novel computational techniques to compare gene expression from our T-ALL cells to gene expression of normal hematopoietic differentiation from the Immunologic Genome Project, we have delineated the regulatory pathways underlying the effects of tumor suppressor loss in human MYC-driven lymphoma cells in the presence of targeted MYC-inactivation. We have also utilized expression analysis to establish gene signatures from unique to our MYC/p53-/- and MYC/p19-/- murine T-ALL models and shown them to be prognostic for specific human T-ALL subtypes, as well as patient relapse-free survival. Overall, we have utilized whole transcriptomic analysis to further define the underlying cell signaling by which loss of p53 and p19 facilitate tumor recurrence and demonstrated that loss of these tumor suppressors can predict outcome in human patients. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 4879. doi:1538-7445.AM2012-4879