The Pan Class-I PI3K Inhibitor Copanlisib Has Preclinical Activity in Mantle Cell Lymphoma, Marginal Zone Lymphoma and Chronic Lymphocytic Leukemia As Single Agent and in Combination with Other Targeted and Conventional Agents

Introduction. Copanlisib is a pan class-I PI3K inhibitor (i) with preferred activity towards PI3Kα and PI3Kδ that is currently in clinical development as single agent and in combination for patients with non-Hodgkin lymphomas. Here, we performed a screening of copanlisib as single agent and in combination with 15 other anti-cancer agents on a panel of 17 lymphoma cell lines derived from mantle cell lymphoma (MCL), marginal zone lymphoma (MZL) and chronic lymphocytic leukemia (CLL).

Methods. Cell lines derived from MCL (Jeko1, Rec1, JVM2, Granta519, Maver1, Mino1, SP-49, SP-53, UPN1, Z138), MZL (Karpas1718, VL51, SSK41, ESKOL, HAIR-M, HC-1) and CLL (MEC1) were exposed to increasing doses of copanlisib alone and in combination with increasing doses of other compounds using the fixed ratio set-up. Tested compounds were approved as well as experimental inhibitors of key regulatory pathways: e.g. anti-CD20 monoclonal antibody rituximab, BTK-i ibrutinib, CDK-i roniciclib, DNA damage agent bendamustine, HDAC-i panobinostat, HDAC1/2-i romidepsin, immunomodulatory lenalidomide, JAK1/2-i ruxolitinib, MALT-i MI2, proteasome-i bortezomib. Synergy was assessed with Chou-Talalay combination index (CI): synergism (<0.9), additive (0.9-1.1), antagonism/no benefit (> 1.1). Gene expression profiling (GEP) was done using the Illumina-HumanHT-12 Expression-BeadChips and GSEA (FDR<0.25).

Results. Copanlisib showed anti-tumor activity in the vast majority of cell lines with a median IC50 of 22nM (95%C.I.; 15-98) across all the cell lines, 36nM (6-175) in MZL, 22nM (95%C.I., 3-103) in MCL, 23nM in the CLL MEC1. Among all cell lines, the most active other compounds were bortezomib (IC50 5nM; 95%C.I., 5-7), romidepsin (34nM, 2-94), roniciclib (23nM, 18-29), panobinostat (161nM, 11-1263), MI2 (490nM; 224-1000) with remaining having median IC50s >500nM. Among these the only showing higher activity in one histotype was panobinostat, more active in MCL than MZL (IC50 17 vs 557nM, P 0.04). As pattern of activity across the cell lines, copanlisib was positively correlated with ibrutinib (R 0.58, P 0.001) and negatively with ruxolitinib (R -0.55, P 0.021).

When copanlisib-containing combinations were tested in the 17 cell lines, different ones frequently appeared synergistic or additive: for example copanlisib plus MI2, 88% (15); plus ibrutinib, 82% (14); plus AKT-i or panobinostat, 76% (13); plus lenalidomide or BET-i, 71% (12); plus rituximab, 65% (11); plus romidepsin, 59% (10); plus roniciclib, 53% (9); plus bortezomib, 47% (8); plus bendamustine, 35% (6); plus ruxolitinib, 12% (2). For combinations with the most promising effect, baseline GEP identified gene-sets associated with different sensitivity to these combinations. High expression of genes involved in interferon signaling, oxidative phosphorylation, fatty acid metabolism, apoptosis, PI3K/AKT/mTOR and IL6/JAK/STAT signaling and low expression of genes involved in cell cycle were associated with synergism to copanlisib combinations with several compounds. Interestingly, largely the opposite was observed for the combination with others, with better synergy in cells with high expression of E2F/MYC targets and genes involved in cell cycle and low expression of transcripts involved in interferon PI3K/AKT/mTOR and IL6/JAK/STAT signaling.

Conclusion. Copanlisib was active in cell lines derived from MCL, MZL and CLL. Combinations of copanlisib with a variety of additional anti-cancer compounds were synergistic and/or additive when combined in different cell lines. The combinations of the synergy profiles in the cell lines tested and the specific GEP features might predict lymphomas that could benefit from these regimens.