176 Background: Despite androgen deprivation therapy, prostate cancer will develop resistance known as castrate resistant prostate cancer (CRPC). Androgen signaling is associated with pathways that influence cell proliferation and survival via cell cycle regulation. AR signaling results in inactivation of Rb tumor suppressor by way of AKT and reduction in intrinsic CDK4/6 inhibitors. Since PTEN loss is found in 40% of mCRPC and AKT pathway is a mechanism of CDK4/6 inhibitor resistance, we hypothesize adding an AKT inhibitor to a CDK4/6 inhibitor will have synergistic activity while delaying CDK4/6 inhibitor resistance. Preclinical xenografts have been shown to recapitulate patient response, therefore, can be used to guide clinical trials using novel combinations. Methods: We have established patient derived tumor xenograft models (GUR-017M, TMA-027 and PRJ-88T) from patients with mCRPC with each having a unique genomic alteration profile. In the clinical setting, GUR-017M was resistant to abiraterone, enzalutamide, cabazitaxel and had exposure to docetaxel. TMA-027 and PRJ-88T were resistant to enzalutamide. We combined a CDK4/6 inhibitor with an AKT inhibitor and treated each of these patient tumors in NOD-SCID mice for a total duration of 40 days. Mice were implanted and separated into 4 groups (placebo, CDK4/6 inhibitor, AKT inhibitor and CDK4/6 + AKT inhibitor) with 10-13 mice in each group. Tumor volumes were measured twice a week. For GUR-017M, abemaciclib (50 mg/kg) was combined with MK2206 (120 mg/kg) and was compared to each placebo and each single agent alone. For TMA-027, palbociclib (50 mg/kg) was combined with ipatasertib (50 mg/kg). After 40 days, tumors were harvested for immunohistochemistry analysis. PRJ-88T was grown as 3D spheroids in ultra-low attachment plates and treated with palbociclib + ipatasertib (10 uM-100uM) to evaluate for synergy. Results: GUR-017M and TMA-027 showed significant delayed tumor growth compared to placebo or either single agent alone. In the GUR-017M group, the final mean tumor volume for the combination group was 149 mm3, compared to 377 mm3 (abemaciclib), 358 mm3 (MK2206) and 527 mm3 placebo group, representing a percent decrease compared to control by 72%, 29%, and 32%, respectively. For TMA-027, the final mean tumor volume for the combination group was 102 mm3, compared to 259 mm3 (palbociclib), 416 mm3 (ipatasertib), and 402 mm3 placebo group, representing a percent decrease compared to control by 75%, 36%, and 4%, respectively. PRJ-88T 3D spheroids showed synergy with palbociclib + ipatasertib. In vivo experiments using PRJ-88T are in progress. Conclusions: The preclinical results show combining CDK4/6 inhibitors with AKT inhibitors is a promising treatment strategy and provides the basis for a clinical trial in patients with mCRPC.