Diffuse large B-cell lymphomas (DLBCLs) are a heterogeneous group of malignancies, the complexity of which still remains to be fully resolved. At least 30-40% of patients are not cured with current chemo-immunotherapy regimens. Improved treatments are urgently needed for these patients. Chemotherapy resistance and relapse are particularly high in patients with certain molecular features, such as those with an “Activated B-cell (ABC)” like gene expression signature, or those with translocations or overexpression of the MYC and BCL2 oncogenes (so-called “double-hit” DLBCLs). Even when cured our current best therapies are highly toxic and carry a significant risk of inducing secondary cancers. Our research seeks to identify fundamental biological mechanisms that drive the survival of DLBCLs including its resistant subtypes. Along these lines we find that DLBCLs are generally addicted to particular stress response proteins. A tumor-enriched form of Hsp90 (TEHsp90) plays an essential role in DLBCLs by supporting the actions of the BCL6, MYC and BCL2 oncoproteins, as well as maintaining B-cell receptor (BCR) signaling. Our team developed a small molecule that selectively inhibits TE-Hsp90 without affecting general Hsp90 housekeeping functions. This molecule, called PUH71 has an accordingly wide therapeutic window and potent activity against DLBCL cells. PUH71 is well tolerated in our phase I clinical trial. We developed a method to accurately measure tumor exposure through PET imaging of I124-labeled PUH71, which may serve as a companion biomarker to guide individualized dosing and interpret clinical responses. We hypothesize that PUH71 can serve as an effective therapeutic agent for DLBCL, including those with ABC- and double-hit molecular signatures. We predict that response can be predicted through PUH71 imaging and biologic biomarkers. We predict that PUH71 will serve as a platform drug for development of effective and well-tolerated combinatorial therapy regimens. Finally, we also developed YK198, a potent and specific inhibitor of specific allosteric states of Hsp70, with activity against DLBCL cells at least in part due to disruption of anti-apoptotic signaling pathways. We hypothesize that Hsp70 inhibitors will be useful therapeutic agents for DLBCL and may overcome possible PUH71 induced Hsp70 feedback resistance. Therefore we propose to perform a PUH71 phase II clinical trial in patients with DLBCL with concordant imaging and biomarker studies, to compare and contrast biological dependency of Hsp90 and Hsp70 inhibitors in DLBCL patient specimens, and to design and test rational combinatorial regimens with PUH71 and YK198.