Figure 3. Imprime binds to and licenses the DC to prime an9gen-specific CD8 T cells. Ex vivo effect on human Monocyte-derived dendri9c cells (MoDC) - MoDC were prepared by culturing Imprime-bound monocytes enriched from whole blood in XVivo15 media supplemented with 10% autologous serum and 50 ng/mL rhGM-CSF for 6 days and then maturing the cells with LPS and TNF-α (50 ng/ml) for 48 hrs. MoDC were subsequently evaluated for, (A) phenotype, (B) enhancing CD4/CD8 T cell proliferaSon by CFSE-diluSon assay and IFN-γ producSon in the supernatant. Shown here are representaSve results from 4 different experiments. In vivo effect on mouse DC - (C) Naïve OT-I TCR transgenic CD8 T cells specific to H-2K b /OVA257-264 pepSde were transferred into congenic hosts. The next day, mice were immunized with OVA pepSde with or without Imprime. OT-I expansion was analyzed in the spleen 7 days later. (D) For funcSonal evaluaSon, splenocytes were sSmulated in vitro with 1uM OVA pepSde for 5 hrs in the presence of brefeldin A. Cells were then stained intracellularly for producSon of cytokines. To evaluate degranulaSon, anS-CD107a anSbody was present during the 5hr sSmulaSon. CD107a (Lamp1) IFN-g no pepSde OVA pepSde OVAp sSmulaSon OVAp+ Imprime 0 10 20 30 40 50 0 10 20 30 40 0 20 40 60 80 *** % of OT-I *** *** IFN-γ TNF-α IL-2 PBS OVA peptide OVA peptide + Imprime 0.0 0.1 0.2 0.3 0.4 0.5 % OT-I of total CD8 Frequency of OT-I PBS OVA peptide OVA peptide + Imprime 0 1×10 4 2×10 4 3×10 4 4×10 4 5×10 4 total OT-I per spleen OT-I totals 2.3x ** CD90.1 + ~1x10 5 naïve OT-I OVA 257-264 pepSde +/- Imprime (1.2mg/mouse) CD90.2 + Abstract # Abstract Poster #A014 Innate immune modulaSon: the novel immunotherapeuSc Imprime PGG triggers the anS-cancer immunity cycle in concert with tumor-targeSng, anS-angiogenic and checkpoint inhibitor anSbodies Nandita Bose, Keith Gorden, Anissa Chan, Adria Bykowski Jonas, Nadine C Oioson, Richard M Walsh, Xiaohong Qiu, Ben Harrison, Takashi Kangas, Kathryn Fraser, Ross Fulton, Steven M Leonardo, Mark Uhlik, and Jeremy Graff. Biothera PharmaceuScals Inc. Eagan, MN. Cancer immunotherapeuScs largely focus on awakening T cell mediated recogniSon and eradicaSon of tumor cells. Indeed, checkpoint inhibitor anSbodies (e.g. pembrolizumab) unleash T cells already involved in anS-cancer responses and have shown remarkable clinical acSvity, though only in ~20-30% of solid tumor paSents. Numerous approaches are being explored to enhance the percent of paSents who benefit from checkpoint inhibitor therapies. Chief amongst these are the innate immune modulaSng therapies collecSvely designated as PAMPs- pathogen- associated molecular paierns. PAMPs operate as the criScal “non-self” signals that, in response to pathogen infecSon, ignite the funcSon of the innate immune system to trigger the immunity cycle. TLR and STING agonists act as PAMPs and reflect bacterial and viral danger signals that can drive dendriSc cell maturaSon, enhancing T cell funcSon. These agents are in development in combinaSon with other immunotherapies, including checkpoint inhibitors, but inspire intolerable cytokine storms and are thereby limited to direct intra-tumoral delivery approaches. We therefore sought to discover and develop a novel, systemically administered PAMP- Imprime PGG (Imprime). Ex vivo studies with whole blood from healthy human donors show that Imprime consistently elicits the acSvaSon of innate immune cells. M2 state macrophages repolarize, showing increased expression of M1 markers (CD86, PD-L1) with coincident reducSon in M2 markers (CD163, CD206). DendriSc cells (DCs) mature, showing enhanced surface expression of CD80, CD86 and MHC class II. FuncSonally, the anSgen presentaSon capability of these re-polarized macrophages and acSvated DCs is substanSally enhanced and drives the robust expansion of co-cultured CD8 T cells as well as the marked upregulaSon of the potent anS-tumor cytokine interferon gamma. In preclinical tumor studies, Imprime is administered IV and profoundly enhances the efficacy of numerous anSbody therapies. Using the B16 experimental metastasis model, we show that Imprime (administered IV) synergizes with the anS-TRP1 tumor-targeSng anSbody TA-99, nearly eradicaSng B16 metastases as measured by visual counts, TRP-1 RT-PCR and in situ immunofluorescence for TRP1. In the H441 and H1299 non-small cell lung cancer xenogrars, Imprime synergizes with the anS-VEGFR2 anSbody DC101 to flat-line tumor growth. In the MC-38 and CT-26 syngeneic tumor models, Imprime synergizes with both anS-PD-1 and PD-L1 checkpoint inhibitor anSbodies to repress tumor growth and/or to eradicate cancer lesions. In situ imaging of these preclinical tumor Sssues shows that Imprime insSgates a re-orientaSon of the immune microenvironment, promoSng an M1 state (e.g. increased iNOS2, decreased Arginase 1), as well as the influx of myeloid cells and, in the syngeneic models, CD8 T cells. In clinical trials in > 400 total paSents to date, Imprime has been safely administered by IV infusion (4mg/kg over 2 hours) and has repeatedly shown evidence for efficacy in combinaSon with tumor targeSng or anS- angiogenic anSbodies. Studies with checkpoint inhibitor anSbodies are slated to begin summer of 2016. We now provide the first evidence in healthy human volunteers that Imprime (IV- 4mg/kg, 2 hours) drives the same innate immune acSvaSon events evident in the preclinical studies (e.g. chemokine and cytokine release, PD-L1 and CD86 upregulaSon) verifying that the clinical dose acSvates the innate immune system. Together, these preclinical and clinical studies provide evidence that the novel PAMP, Imprime PGG, can be safely administered systemically and can drive the criScal innate immune acSvaSon necessary to spark the anS-cancer immunity cycle. 2016 AACR Tumor Immunlogy and Immunotherapy # A89 Background Figure 1. Imprime (PAMP) Ac9vates An9-Cancer Immunity Cycle • Imprime PGG, a yeast-derived pharmaceuScal-grade soluble 1,3/1,6 β-glucan is being developed for the treatment of cancer in conjuncSon with tumor targeSng and immunomodulatory anSbodies (Abs). - Imprime has shown promising results in mulSple Phase 2 clinical trials in non-small cell lung cancer (NSCLC), and chronic lymphocySc leukemia (CLL) with addiSonal studies ongoing. • β-glucans are conserved microbial structures found in the cell wall of unicellular and mulScellular pathogens. They are considered pathogen-associated molecular paierns (PAMPs) recognized by the paiern recogniSon receptors including DecSn-1 and Complement Receptor 3 (CR3). Imprime forms an immune complex with endogenous serum immunoglobulin IgG or IgM anS-beta-glucan anSbodies (ABA) before being recognized by CR3 and FcgRIIA on innate immune cells. Results 0 15 30 45 0 20 40 60 80 100 Time (Minutes) Vehicle None Imprime PGG None Vehicle Rituximab Imprime PGG Rituximab PMN treatment Tumor treatment ReacSve Oxygen Species (ROS) in RLU % ADCP 0 20 40 60 80 *** Abs: None Rituximab Raji Vehicle Imprime A B In vivo studies C D • Syngeneic B16 melanoma model - Imprime + An9-Tryp1, TA99 Figure 2. Imprime treatment enhances efficacy of an9-tumor an9bodies. Ex vivo studies – Neutrophils were isolated from WB by negaSve selecSon and co-incubated with Raji B cell lymphoma cells pre-treated with vehicle or coated with the anS-CD20 anSbody Rituximab (1 μg/ml). ROS generaSon was measured by the luminol method and is shown as relaSve light units (RLU). (B) M2c macrophages were prepared from enriched monocytes isolated from WB, and cultured in X-Vivo 10 medium (supplemented with 5% autologous serum and 50 ng/ml rhM-CSF) for 6 days with 10 ng/ml rhIL-10 addiSon for the last 24 hrs. M2C’s ability to phagocytose fluorescently- labeled Rituximab-opsonized Raji cells was measured by flow cytometry. Data are representaSve of > 3 independent experiments from different donors. In vivo studies – C57BL/6 mice were injected via the tail vein with 100,000 B16F10 melanoma cells. The anS-Tryp1 monoclonal anSbody TA99 and Imprime were administered intraperitoneally (50 ug/ mouse D1,3,5,7,10) and intravenously (1.2 mg/mouse D1,3,7,10,14), respecSvely. Mice were euthanized 10 days arer tumor challenge to count lung metastases. (C) Mean number of metastases per treatment group (± SEM). (D) RepresentaSve immunohistochemical staining of tumor Sssues showing complete repression of outgrowth of metastases in the TA99 + Imprime treatment group. Imprime enhances DC matura9on and an9gen presenta9on to T cells 0 5 10 15 20 25 % of Proliferating Cells **p = 0.0038 ***p = 0.0004 Allo-CD4T Allo-CD8T 0 50 100 150 IFNγ (pg/ml) **p = 0.0026 C B In Vivo Studies Imprime repolarizes tumor microenvironment Figure 7. ABA-dependent Clinical Responses. (A) Immunopharmacodynamic (IPD) responses elicited by IV administraSon of Imprime in healthy human subjects. Whole blood or serum was drawn from 24 healthy volunteers at various Sme points before and arer a single dose (cohort 1) or mulSple weekly doses (cohort 2) of Imprime infusion. Cell mobilizaSon was measured by complete blood cell counts, plus differenSals. Cytokine/chemokine measurement in serum was performed using Novex magneSc mulSplex assay (Life Technologies) the Luminex XMAP technology. (B) IgG ABA were determined by ELISA for all evaluable paSents in the Primus trial (third line CRC paSents treated with Cetuximab or Imprime+Cetuximab) . Upper ler panel- Hazard RaSo (HR) vs ABA level is graphically represented. The ABA level for each paSent is shown on the x axis. The solid line in the graph represents the HR while the doied lines that bracket this line represent the 95% confidence intervals. VerScal lines imposed on the graph simply represent the cut-points chosen for Kaplan- Meier analyses. The inset shows the % of paSents in this trial at the different ABA cutpoints. Other panels- each represents Kaplan-Meier analyses for OS at the indicated ABA cut-points (20, 35, or 45μg/ml). HR for each, and staSsScal significance are noted in the inset. CD163 CD86 PD-L1 Count 0 10 2 10 3 10 4 10 5 Median 86 2434 759 Median 103 439 790 Mean Fluorescence Intensity Median 129 1175 2108 Isotype ctrl Staining M2-Vehicle M2-Imprime A Week 1 Pre-Infusion Week 1 15min Week 1 30min Week 1 EOI Week 1 1 Hour Week 1 4 Hour Week 1 24 Hour Week 2 Pre-Infusion Week 2 15min Week 2 30min Week 2 EOI Week 2 1 Hour Week 2 4 Hour Week 2 24 Hour Week 2 48 Hour Week 3 Pre-Infusion 0 200 400 600 800 1000 Monocytes (cell/ul) IgG Positive Group (>35ug/ml) 4 10 22 34 Imprime Saline Week 1 Pre-Infusion Week 1 15min Week 1 30min Week 1 EOI Week 1 1 Hour Week 1 4 Hour Week 1 24 Hour Week 2 Pre-Infusion Week 2 15min Week 2 30min Week 2 EOI Week 2 1 Hour Week 2 4 Hour Week 2 24 Hour Week 2 48 Hour Week 3 Pre-Infusion 0 200 400 600 800 1000 Monocytes (cell/ul) ABA Negative Group Imprime Saline 47 39 25 Week 1 Pre-Infusion Week 1 15min Week 1 30min Week 1 EOI Week 1 1 Hour Week 1 4 Hour Week 1 24 Hour Week 2 Pre-Infusion Week 2 15min Week 2 30min Week 2 EOI Week 2 1 Hour Week 2 4 Hour Week 2 24 Hour Week 2 48 Hour Week 3 Pre-Infusion 0 200 400 600 800 1000 Monocytes (cell/ul) IgM Positive Group (>50ug/ml) Imprime Saline 37 16 2 PRE-IMPRIME POST-IMPRIME 1 HOUR 24 HOURS 0 20 40 60 80 100 2000 4000 6000 8000 10000 IL-8 (pg/ml) IgG Positive Group (>35 ug/ml) IL-8 010 022 004 034 ABA Negativ PRE-IMPRIME POST-IMPRIME 1 HOUR 24 HOURS 0 20 40 60 80 100 2000 4000 6000 8000 10000 IL-8 (pg/ml) IgM Positive Group (>50 ug/ml) IL-8 002 016 037 oup Imprime induces monocyte mobilizaSon in high IgG ABA subjects Imprime induces cytokine/chemokine producSon in high ABA subjects Vehicle Imprime TA99 TA99 + Imprime 0 20 40 60 # Mets (D10) P value 0.0054 Count CD80 CD86 CD83 HLA-DR Mean Fluorescence Intensity (MFI) Median 151 4286 7049 Median 223 640 744 Median 162 168 466 Median 151 871 10759 Isotype ctrl staining Imprime A M1 polariza9on of macrophages: Ex vivo human studies B MDSC Differen9a9on – Ex vivo human studies A B Figure 5. Imprime-treated MDSC have APC-like characteris9cs. (A) Phenotype and funcSon of human MDSC – Human MDSC were prepared from human cord blood per a modified version of the published protocol (Wu et al., PNAS, 2014). Flow cytometry analyses showed increased CD86 expression on both monocySc MDSC (Mo-MDSC) and PMN-MDSC (not shown). FuncSonal assessment in a T-cell proliferaSon assay showed that Imprime-treated MDSC were less suppressive to T-cell proliferaSon. (B) H1299 lung cancer xenograr model: In vivo administraSon of Imprime also enhanced CD86 expression on PMN-MDSC in the tumor. Significant (p<0.05) tests: 36 out of 86 (41.9%) 2 1 1/2 1/4 1/8 1/16 HR with 95% CI 10 20 30 40 50 1.0 0.8 0.2 0.0 0.6 0.4 Survival Probability HR=0.24 (0.1-0.55) p=0.00031 ABA ≥ 45 ABA < 45 0 10 20 30 40 50 0 10 20 30 40 50 ABA < 20 ABA ≥ 20 1.0 0.8 0.2 0.0 0.6 0.4 HR=0.77 (0.49-1.22) p=0.27 Survival Probability % population at ABA cutpoints ≥ 20 μg/ml = 49% ≥ 35 μg/ml = 24% ≥ 45 μg/ml = 16% 57 25 5 1 0 55 28 9 2 1 1 94 40 10 1 0 18 13 4 2 1 1 1.0 0.8 0.2 0.0 0.6 0.4 0 10 20 30 40 50 ABA ≥ 35 ABA < 35 HR=0.40 (0.22-0.73) p=0.0018 Survival Probability 85 36 8 1 0 0 27 17 6 2 1 1 ABA levels (μg/ml) in evaluable patients Time (months) Time (months) Time (months) Conclusions Acknowledgements All experiments funded by Biothera Pharmaceuticals Inc. No external funding was received to support the work. ABA and OS in Imprime -treated colorectal cancer pa9ents Phase I healthy human volunteer trial Ex Vivo Human Studies Ex Vivo Human Studies A B Imprime enhances innate immune cell killing in concert with tumor-targe9ng an9bodies DAPI Tryp1 Veh Imp TA99 TA99 + Imp Imprime enhances the efficacy of an9- angiogenics and checkpoint inhibitors Clinical biomarker research In vitro studies In vivo studies Imprime Vehicle 2:1 PBMC:MDSC Division Index 1.0 0.5 0.0 1.5 *0.0118 Vehicle Imprime Mo-MDSC CD86 3500 3000 2500 2000 1500 MFI **0.0082 Figure 6. Imprime enhances in vivo an9-tumor efficacy of an9-angiogenics and checkpoint inhibitors. (A) Imprime in combinaSon with DC101 (murine surrogate for ramucirumab) in H441 and H1229 NSCLC xenograr models, respecSvely. Once tumors reached ~100mm 3 , mice were administered DC101 (10 mg/kg twice weekly IP for up to six weeks) and/or Imprime (1.2 mg/mouse i.V twice weekly for up to six weeks) . (B) Imprime in combinaSon with anS-PD-1 anSbody was tested in CT26 colon cancer-bearing BALB/c mice or with anS-PD-L1 anSbody in MC38 colon cancer –bearing C57Bl/6 mice. 3 days arer subcutaneous injecSon of tumor, the mice were administered Imprime and/or anS-PD-1/ PD-L1 (100 μg/mouse twice weekly IP for up to five weeks). 0.0 0.2 0.4 0.6 0.8 1.0 ****p <0.0001 ns 0 200 400 600 800 0 20 40 100 200 300 400 500 600 IFNγ (pg/ml) IL-4 (pg/ml) **p = 0.0015 ns IL-4 IFN-γ CD4 T cell expansion (Division Index) An9- angiogenics – immunomodulatory effects of an9-VEGFR2 0 5 10 15 20 25 0 400 800 1200 1600 2000 Study Days Tumor Volume (mm 3 ) Vehicle Imprime DC101-10mg/kg Imprime + DC-101 Vehicle Imprime PD-L1 Imprime +PD-L1 %tumor free mice (day 29) 5.6% 11% 33% 83% Checkpoint inhibitors – An9-PD-1/PD-L1 an9bodies A B 0 10 20 30 40 0 400 800 1200 1600 2000 Study Days Tumor Volume (mm3) Vehicle Imprime PD-1 Imprime + PD-1 * **** H441 MC38 CT26 H1299 D PRE-IMPRIME POST-IMPRIME 1 HOUR 24 HOURS 0 20 40 60 80 100 2000 4000 6000 8000 10000 IL-8 (pg/ml) ABA Negative Group IL-8 039 025 Bev Bev+Imprime Vehicle Count CD86 An9-Tumor T cell Immunity- T cell killing and immune memory Enhances an9gen presenta9on to T cells Drives DC maturaSon & acSvaSon Imprime PGG- Binds to innate immune cells & triggers a coordinated response Repolarizes tumor immune microenvironment M2 to M1 polarizaSon MDSC differenSaSon Enhanced T cell priming Enhanced T cell effector funcSons Neutrophil Macrophage NK Cell Enables innate immune cell killing Possible Immunogenic cell death 1 2 3 Imprime PGG acts as a PAMP to enlist the broad funcSonality of the innate immune system to enhance the anS-tumor efficacy of tumor-targeSng, anS- angiogenic, and immune checkpoint inhibitor monoclonal anSbody therapies. Imprime PGG acSvates anS-cancer innate immune effector funcSons by: 1) triggering direct tumor killing by innate effector cells 2) acSvaSng the maturaSon of anSgen presenSng cells 3) overriding tumor-induced immunosuppression and These innate immune funcSons of Imprime PGG are criScal for triggering the immunity cycle that ulSmately drives the anS-tumor T cell-based immunity. 0 1 10 100 iNOS CXCL11 CXCL9 IL-12b CXCL10 IL-6 PD-L1 TNF IFNr CCR7 CD86 CCL2 CCL3 Arg 1 CCL17 TGFb Fizz-1 CD206 YM-1 Mouse BMM-Imprime Figure 4. Imprime treatment results in M1 polariza9on of macrophages. Ex vivo studies - M2 macrophages were prepared by culturing Imprime-bound monocytes enriched from human whole blood in the presence of M2-polarizing polarizing condiSon as described in Figure 2B for 6 days. Macrophages were subsequently evaluated for, (A) phenotype, (B) CD4 T cell proliferaSon by CFSE-diluSon assay, and IFN-γ and IL-4 producSon in the supernatant by ELISA. In vivo studies - (C) Bone marrow-derived macrophages (BMM) was prepared by culturing bone marrow cells harvested from Imprime-treated mice in RPMI medium supplemented with 10% fetal calf serum and 20 ng/ml rmM-CSF fpr 7 days, and subsequently measured by qRT-PCR. (D) qRT-PCR assay show up-regulaSon of M1 markers and down-regulaSons of M2 markers. M1 polariza9on of macrophages: In vivo studies C D Mouse BMM-Vehicle Vehicle 0.1 1 10 100 CXCL11 CXCL10 IL-12b TNFa Arg 1 TGFb IL-10 CD206 Mouse tumor-Vehicle Mouse tumor-Imprime 0 10 20 30 40 0 400 800 1200 Study Days Tumor Volume (mm 3) PBS Imprime DC-101 DC-101 + Imprime *