Despite a variety of novel therapeutic options in Hodgkin lymphoma (HL),1 treatment of primary refractory and early-relapsed patients still represents an unmet medical need requiring the development of more effective strategies.2 Brentuximab vedotin (BV) is a first-in-class antibody-drug conjugate that is approved in the United States and the European Union for the treatment of relapsed or refractory HL following the failure of two chemotherapy regimens or autologous stem cell transplant.3 However, the capacity of this drug to induce a long-term disease control is still under investigation, suggesting that combination therapies using molecularly targeted agents with higher efficacy and reduced toxicity are needed.4 The PI3K/Akt/mTOR pathway has a pathogenetic role in HL, providing the rational to target this pathway.5 Of the several PI3K isoforms, PI3K-δ has a significant role in supporting the growth and survival of B-cell malignancies.6 TGR-1202 is a novel, oral, PI3K-δ-specific inhibitor that is currently in clin. development for a variety of hematol. malignancies as a single agent and for all types of B-cell malignancies in combination with other agents.7 In the present study, we characterized the activity of TGR-1202 in combination with BV in non-clin. models of HL to provide translational support for this combination as a novel therapeutic strategy.The rationale for using TGR-1202 in combination with BV is supported by three lines of evidence.First, PI3K-δ inhibitors and BV act by affecting cell cycle progression;8, 9, 10 thus, we can expect enhanced effects by combining these mols.Second, PI3K inhibitors can sensitize tumor cells to chemotherapeutics.11 Third, BV is a clin. drug, and TGR-1202 is currently in advanced clin. development and has a different safety profile than other PI3K-δ inhibitors, with limited liver and gastrointestinal toxicity.12 We first assessed the effects of TGR-1202 and BV as single agents on the proliferation and survival of HL cells.In L-540, KM-H2 and L-428 cells, the 50% proliferation inhibition (IC50) of TGR-1202 was 11, 16 and 46 μM, resp., at 48 h (Supplementary Figure 1a).Exposure to 10 μM TGR-1202 for 72 h significantly increased cell death in L-540 cells (L-540: control: 13±1%, TGR-1202: 43±7%, P≤0.0001) (Supplementary Figure 1b), whereas a modest cytotoxic effect was observed in KM-H2 and L-428 cells (Supplementary Figure 1b).Notwithstanding the antiproliferative and cytotoxic effects on HL cells, TGR-1202 (5, 10, 20 μM) failed to induce any cytotoxicity in hematopoietic progenitor cells (Supplementary Figures 2a and b) or PBMCs from healthy donors (Supplementary Figure 2c).Consistently with CD30 expression in HL cells, treatment with BV induced significant proliferation inhibition (P≤0.001 at least) of all HL cells (Supplementary Figure 1c, insert).PI3K and the downstream kinases Akt, GSK, Aurora and Stathmin have recently been implicated in regulating both microtubule dynamics and organization.13 We analyzed the effects of TGR-1202 on the PI3K/Akt signaling pathway and microtubule-associated proteins.All HL cell lines showed constitutive phosphorylation of Akt, which was markedly inhibited by TGR-1202 (<1 h) (Supplementary Figure 3a).With longer exposures (8, 24 h) to TGR-1202, Akt, GSK-3β, Aurora and Stathmin were also dephosphorylated (Supplementary Figures 3a-c).Consistent with the activity of BV as a microtubule-disrupting agent, we detected a marked decrease in GSK-3β, Aurora and Stathmin phosphorylation associated with the deregulation of the mitotic spindle, as shown by tubulin staining inhibition, in all BV-treated HL cells (Supplementary Figure 3c).To investigate the potential synergy of the two drugs, dose-response anal. of cell death following TGR-1202/BV treatment of L-540 cells for 48 h yielded confidence interval values below 0.7, indicating a synergistic interaction (Supplementary Figure 4a).When compared with single-agent exposure, pharmacol. achievable concentrations of TGR-1202 (10 μM) combined with BV (10 ng/mL) significantly reduced the proliferation of HL cells (Figure 1a), and increased apoptotic cell death (Figure 1a) resulting in a confidence interval value of 0.05 (Supplementary Figure 4a).Western blot anal. revealed that the two-drug exposure markedly increased caspase activation and PARP cleavage (Figure 1b), whereas the pan-caspase inhibitor Z-VAD-fmk abrogated apoptosis (Supplementary Figure 4b).Furthermore, the anti-apoptotic proteins Mcl-1 and Bcl-2 were downregulated after BV treatment and remained downregulated even in the combined treatment (Supplementary Figure 4c).Compared with single agents, the combined treatment of HL cells resulted in an almost complete inhibition of Akt and S6 phosphorylation (Figure 1b).As the blockade of the G2/M transition is a prominent feature of microtubule-disrupting agents that induce pro-apoptotic signaling pathways,14 the effect of TGR-1202/BV on cell cycle progression was assessed.Compared with single agents, a 24-h exposure to TGR-1202/BV induced a marked G2/M accumulation, which was paralleled by G0/G1 reduction (Figure 1c).Such an accumulation was followed at 48 h by an increase in total hypodiploid cells (sub-G1 population) (Supplementary Table 1), suggesting that TGR-1202/BV induces an early G2/M arrest followed by apoptotic cell death.Interestingly, the G2/M accumulation was still evident and even enhanced when apoptosis was blocked by Z-VAD-fmk (Supplementary Figure 5).To elucidate the mechanisms by which TGR-1202/BV induces G2/M phase arrest, the levels of G2/M phase regulatory proteins were assessed.After TGR-1202/BV treatment, the levels of cyclin A2 and CDK1 increased compared with single agents (Supplementary Figure 6a).The activated form of CDK1, phospho-CDK1 (at Thr161), and the two mitosis markers, MPM-2 and phospho-histone H3, also markedly increased (Supplementary Figure 6a), indicating that TGR-1202/BV led to mitotic arrest.Cyclin B1 is a G2/M checkpoint regulator, whereas p21 protein has been involved in the control of both the G2/M and G1/S checkpoint.15 As shown in Supplementary Figures 6a and b, the levels of these two regulatory proteins were low in vehicle-treated cells, whereas markedly increased in TGR-1202/BV-treated cells.A remarkable nuclear accumulation of cyclin B1 protein was concomitantly observed in TGR-1202/BV-treated HL cells compared with vehicle controls (Supplementary Figure 6b).We hypothesized that TGR-1202/BV might induce mitotic arrest by suppressing microtubule dynamics.As expected, a normal microtubule distribution was observed in vehicle-treated HL cells, whereas, exposure to TGR-1202/BV was associated with significant microtubule disruption, as indicated by the appearance of diffuse staining, irregular microtubule fragments throughout the cytosols (Supplementary Figure 7a), resulting in a marked α-tubulin inhibition (P≤0.0001) (Supplementary Figure 7b).At last, the combination treatment led to disorganized mitotic spindle structures and delocalization of the microtubules, which were distributed in the membrane and perinuclear cytoplasmic regions, suggesting microtubule condensation in the mitotic cells (Supplementary Figure 7a).To determine whether TGR-1202/BV directly affects microtubule formation in vitro, we conducted a tubulin polymerization assay.Highly purified α- and β-tubulin and GTP were incubated with either TGR-1202 and/or BV.Similarly, to the CaCl2 treatment (pos. control 'inhibitor'), the combined TGR-1202/BV treatment significantly inhibited tubulin polymerization in a time-dependent manner (Supplementary Figure 7c).In contrast, TGR-1202 and BV as single agents only minimally affected the tubulin polymerization (Supplementary Figure 7c).These results suggest that TGR-1202/BV disrupts microtubules by impeding microtubule assembly.Under our exptl. conditions, in vivo treatment with single-agent TGR-1202 or BV slightly reduced the growth of L-540 nodules by 19% and 28% as compared with vehicle-treated controls, resp. (Figure 2a).Notwithstanding the limited antitumor activity of single agents, the TGR-1202/BV combination induced 66% (P≤0.0001), 57% (P≤0.0001) and 53% (P≤0.001) tumor growth inhibition, as compared with vehicle-treated controls, TGR-1202 or BV alone, resp. (Figure 2a).Similarly, the combination treatment significantly reduced KM-H2 tumor nodules by 75, 68 and 48%, as compared with vehicle controls, TGR-1202 or BV alone, resp. (Figure 2a).No significant changes in weight or other signs of potential toxicity were observed during the treatment with TGR-1202, BV or the combined treatment (Supplementary Figure 8).To elucidate the in vivo mechanism(s) of TGR-1202/BV-mediated anti-lymphoma activity, we determined the extent of necrotic areas as a marker of tumor cell death by hematoxylin and eosin staining (data not shown) and DNA fragmentation detection by Terminal deoxynucleotidyl transferase dUTP Nick-End Labeling staining (Figure 2b).TGR-1202/BV combination triggered significantly greater cell death than single-agents in HL tumor nodules (Figure 2b).Indeed, we detected a two- to four-fold increase in tumor necrosis in L-540 nodules (combination: 50±5%; TGR-1202: 13±1%; BV: 23±5%, P≤0.0001) and a three- to fourfold in KM-H2 nodules (combination: 47±8%; TGR-1202: 12±2%; BV: 18±6%, P≤0.001) (Figure 2b).We finally investigated whether the dual TGR-1202/BV treatment could inhibit microtubule assembly in vivo by studying α-tubulin expression in tumor sections derived from L-540 and KM-H2 nodules (Figure 2c).Treatment with TGR-1202 or BV modestly inhibited microtubule assembly, but the combination caused severe tubulin inhibition with abnormal mitotic spindle formation and the occurrence of mono-spindles (Figure 2c).Collectively, these results suggest that the in vivo antitumor efficacy of the TGR-1202/BV treatment is associated with tumor cell death resulting from increased microtubule disruption in tumor cells and areas of tumor necrosis.To the best of our knowledge, data reported herein represent the first demonstration of the capacity of a PI3K-δ inhibitor (TGR-1202) to enhance the anti-lymphoma activity of BV leading to marked anti-mitotic and cytotoxic effects in HL.These findings provide a strong preclin. rationale to future clin. trials using the TGR-1202/BV combination in relapsed/refractory HL.