Improving responses to dual PI3K/BCL2 inhibition in lymphomas: results of a pharmacological screen with over 1,400 compounds

Background: Copanlisib, a pan-PI3K inhibitor with stronger activity against the alpha and delta isoforms, is Food and Drug Administration (FDA) approved for the treatment of patients with relapsed or refractory follicular lymphoma, and it is has shown clinical activity in other indolent lymphomas including marginal zone lymphoma (MZL). Preclinical studies have demonstrated strong synergism of copanlisib and BCL2 inhibitor venetoclax in different lymphomas (Tarantelli Blood Adv 2020). A phase 1 study has explored this specific combination (NCT03886649), while others have explored similar schemes based on dual PI3K and BCL2 targeting. We have previously reported a model of resistance to copanlisib/venetoclax developed by prolonged exposure to copanlisib (Arribas ENA 2020). Here, we present the data from a large pharmacological screen with over 1400 compounds in this MZL model. Methods: Copanlisib/venetoclax-resistant cells were exposed to DMSO or a library of 1443 FDA-approved compounds at 5 μM dose, as single or in combination with 1 nM copanlisib +50 nM venetoclax for 72 hours (hr). Spatial and intra- and inter-plates effects were corrected. Values were then normalized to control (DMSO). Either highly active compounds as single agent (cell viability <30%) or inhibitors with improved activity upon copanlisib/ venetoclax combination (cell viability <50%, ratio combo/single <0.7) were selected for further validation upon copanlisib/venetoclax in resistant and parental cells (MTT assay, 72 hr). Synergy of combinations was evaluated according to the Chou-Talalay combination index (CI), as well as potency and efficacy, estimated according to the MuSyC algorithm. Results: The screen identified 82 highly active compounds and 99 drugs that improved the response to copanlisib/venetoclax. These inhibitors targeted WNT, CDK, HDAC, HSP, PLK, ALDH1, AURKA, proton pump and microtubule polymerization, among others. A selection underwent further validation experiments in parental and resistant upon copanlisib/venetoclax combination. Addition of the ALDH1 inhibitor disulfiram was beneficial in both parental and resistant cells, but especially in the latter condition with improved synergy and enhanced efficacy of copanlisib/ venetoclax, overcoming the resistance. Combinations with ganetespib (HSP90), rigosertib (PLK), panobinostat (HDAC), alisertib (AURKA) and AZ1602 (WNT) were beneficial either in parental and resistant cells and improved the sensitivity to copanlisib/venetoclax in resistant cells. Conclusions: In conclusion, a large pharmacological screen has identified a series of compounds that could be added to copanlisib/ venetoclax in a triple combination or used in patients developing resistance to copanlisib/venetoclax. Further experiments are on-going to extend the findings to additional PI3K/BCL2 inhibitors. Work partially supported by Swiss Cancer Research. No conflict of interest.