Abstract 4999: Evaluation of anti-Gr1 antibody for depletion of MDSC in preclinical NSG mouse models of pediatric sarcoma

Preclinical in vivo studies of chimeric antigen receptor (CAR) T cells often rely on NOD-scid gamma (NSG) mouse models which lack T cells, B cells and NK cells and thereby allow for reliable engraftment of human tumor xenografts. But murine myeloid cells present in the NSG mouse strain can affect the tumor microenvironment as well as the function of adoptively transferred human immune effector cells. Long et al. (Cancer Immunol Res. 2016 Oct;4(10):869-880) reported that pediatric sarcoma xenografts in NSG mice induce in vivo expansion of murine CD11b+ myeloid-derived suppressor cells (MDSC), and that this cell population suppresses human CAR T cell proliferation in vitro. We investigated a strategy to avoid inhibitory effects of murine MDSCs on Ewing sarcoma xenografts in NSG mice by in vivo pretreatment with murine antibody against two cell surface antigens, Ly6C/Ly-6G (granulocyte-differentiation antigen-1, Gr-1), expressed on murine MDSC. Analysis of the CD11b+ myeloid cell populations in NSG mice 17-37 days after subcutaneous transplantation of the Ewing sarcoma cell line TC-71 showed a noticeable increase of CD11b+ cells in the peripheral blood compared to non-tumor bearing NSG mice (median 3.4 × 103, range 0.7-15.5 × 103 cells/ml, n=8 vs median 2.1 × 103, range 1.2-4.0 × 103 cells/ml, n=7). The increase was even more pronounced in the spleens, with a median absolute number of 2.6 × 106 CD11b+ cells (range 0.2-21 × 106 cells) per spleen in sarcoma-bearing mice versus 0.5 × 106 CD11b+ cells (range 0.2-0.8 × 106 cells) per spleen in mice without tumors. The granulocytic MDSC subset coexpressing Ly6Gpos was the most prominent subpopulation in spleen and blood. To eliminate murine MDSCs, we treated sarcoma-bearing mice twice weekly with 200 µg anti-Gr1 antibody RB6-8C5 over a period of 2 weeks, starting at tumor volumes of 100-200 mm3. While the combined percentage of the Ly6Gpos and Ly6Cpos cell populations in the spleens decreased compared to untreated mice (median 55%, range 36-75%, n=9 versus median 74%, range 72-86%, n=8), the total numbers of CD11b+ cells further increased (median 3.1 × 106, range 0.4 × 106-11 × 106 cells/spleen). Our data suggest that anti-Gr1 antibody pretreatment leads to blockade of the Ly6C/Ly6G receptors rather than eliminating MDSC subsets. To optimize preclinical pediatric sarcoma models, other methods for depletion or functional inactivation of mouse MDSCs will need to be evaluated.

Citation Format: Sareetha Kailayangiri, Bianca Altvater, Katja Urban, Jutta Meltzer, Lea Greune, Nicole Farwick, Silke Jamitzky, Claudia Rossig. Evaluation of anti-Gr1 antibody for depletion of MDSC in preclinical NSG mouse models of pediatric sarcoma [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4999.

Overcoming Heterogeneity of Antigen Expression for Effective CAR T Cell Targeting of Cancers

Chimeric antigen receptor (CAR) gene-modified T cells (CAR T cells) can eradicate B cell malignancies via recognition of surface-expressed B lineage antigens. Antigen escape remains a major mechanism of relapse and is a key barrier for expanding the use of CAR T cells towards solid cancers with their more diverse surface antigen repertoires. In this review we discuss strategies by which cancers become amenable to effective CAR T cell therapy despite heterogeneous phenotypes. Pharmaceutical approaches have been reported that selectively upregulate individual target antigens on the cancer cell surface to sensitize antigen-negative subclones for recognition by CARs. In addition, advanced T cell engineering strategies now enable CAR T cells to interact with more than a single antigen simultaneously. Still, the choice of adequate targets reliably and selectively expressed on the cell surface of tumor cells but not normal cells, ideally by driving tumor growth, is limited, and even dual or triple antigen targeting is unlikely to cure most solid tumors. Innovative receptor designs and combination strategies now aim to recruit bystander cells and alternative cytolytic mechanisms that broaden the activity of CAR-engineered T cells beyond CAR antigen-dependent tumor cell recognition.

EZH2 Inhibition in Ewing Sarcoma Upregulates GD2 Expression for Targeting with Gene-Modified T Cells

Chimeric antigen receptor (CAR) engineering of T cells allows one to specifically target tumor cells via cell surface antigens. A candidate target in Ewing sarcoma is the ganglioside G_D2, but heterogeneic expression limits its value. Here we report that pharmacological inhibition of Enhancer of Zeste Homolog 2 (EZH2) at doses reducing H3K27 trimethylation, but not cell viability, selectively and reversibly induces G_D2 surface expression in Ewing sarcoma cells. EZH2 in Ewing sarcoma cells directly binds to the promoter regions of genes encoding for two key enzymes of G_D2 biosynthesis, and EZH2 inhibition enhances expression of these genes. G_D2 surface expression in Ewing sarcoma cells is not associated with distinct in vitro proliferation, colony formation, chemosensitivity, or in vivo tumorigenicity. Moreover, disruption of G_D2 synthesis by gene editing does not affect its in vitro behavior. EZH2 inhibitor treatment sensitizes Ewing sarcoma cells to effective cytolysis by G_D2-specific CAR gene-modified T cells. In conclusion, we report a clinically applicable pharmacological approach for enhancing efficacy of adoptively transferred G_D2-redirected T cells against Ewing sarcoma, by enabling recognition of tumor cells with low or negative target expression.

Carbohydrate Targets for CAR T Cells in Solid Childhood Cancers

Application of the CAR targeting strategy in solid tumors is challenged by the need for adequate target antigens. As a consequence of their tissue origin, embryonal cancers can aberrantly express membrane-anchored gangliosides. These are carbohydrate molecules consisting of a glycosphingolipid linked to sialic acids residues. The best-known example is the abundant expression of ganglioside G_D2 on the cell surface of neuroblastomas which derive from G_D2-positive neuroectoderm. Gangliosides are involved in various cellular functions, including signal transduction, cell proliferation, differentiation, adhesion and cell death. In addition, transformation of human cells to cancer cells can be associated with distinct glycosylation profiles which provide advantages for tumor growth and dissemination and can serve as immune targets. Both gangliosides and aberrant glycosylation of proteins escape the direct molecular and proteomic screening strategies currently applied to identify further immune targets in cancers. Due to their highly restricted expression and their functional roles in the malignant behavior, they are attractive targets for immune engineering strategies. G_D2-redirected CAR T cells have shown activity in clinical phase I/II trials in neuroblastoma and next-generation studies are ongoing. Further carbohydrate targets for CAR T cells in preclinical development are O-acetyl-G_D2, NeuGc-GM3 (N-glycolyl GM3), G_D3, SSEA-4, and oncofetal glycosylation variants. This review summarizes knowledge on the role and function of some membrane-expressed non-protein antigens, including gangliosides and abnormal protein glycosylation patterns, and discusses their potential to serve as a CAR targets in pediatric solid cancers.

Abstract 4631: Expression of ganglioside GD2 in Ewing sarcoma cells is not associated with specific functional characteristics or stem cell features

Safe and effective targeting of cancer with chimeric antigen receptor (CAR) engineered T cells relies on the presence of adequate tumor-associated surface antigens. We have found that Ewing sarcomas (EwS) express ganglioside GD2 and are recognized by T cells engineered with GD2-specific CARs. In an extended series of 97 EwS patient samples, 52 were GD2pos by immunohistochemistry. Based on reports that GD2 in breast cancer defines a malignant population with stem cell characteristics, we hypothesized that GD2 expression in EwS is associated with high capacity to self-renew and reinitiate tumor growth. Among 15 individual EwS cell lines with variable surface expression of GD2 by flow cytometry, GD2 expression levels were not associated with the capacities to proliferate and expand in vitro, form colonies in semi-solid medium, nor with their chemosensitivity, assessed by cell viabilities in the presence of increasing concentrations of the cytotoxic drug doxorubicin. Subpopulations with GD2hi phenotype selected from GD2pos and from GD2low EwS cell lines by cell sorting maintained their GD2hi and GD2low expression status during subsequent cultures over several weeks. The two subpopulations did not have different in vitro growth, colony-formation capacity, or chemosensitivity. In xenografting experiments, GD2hi and GD2low subpopulations of two EwS cell lines initiated tumors with comparable efficacies. To obtain direct evidence that GD2 surface expression is irrelevant for the biology of EwS cells, we performed genetic knockdown of the GD3S gene which drives biosynthesis of GD2 by CRISPR/Cas9 gene editing. GD3S gene editing resulted in effective elimination of GD2 surface expression in the GD2hi EwS cell lines TC-71, VH-64 and A4573. The knockdown did not affect the capacity of the cells to proliferate, form colonies in soft agar in vitro, nor their chemosensitivity compared to wild-type EwS cells from the individual cell lines. We conclude that GD2 expression in EwS cells, other than reported in breast cancer, is not associated with distinct functional features. Specifically, GD2 does not affect the growth characteristics, clono- and tumorigenicity and chemosensitivity of the tumor cells. Elimination of GD2pos subpopulations from heterogeneous tumors by CAR T cell targeting is therefore unlikely to eradicate the disease and will have to be combined with additional targeting strategies.

Citation Format: Sareetha Kailayangiri, Bianca Altvater, Silke Jamitzky, Stefanie Lesch, Jan-Henrik Mikesch, Sonja Schelhaas, Schaefers Michael, Jutta Meltzer, Nicole Farwick, Wolfgang Hartmann, Eva Wardelmann, Petra Fischer, Jendrik Hardes, Claudia Rossig. Expression of ganglioside GD2 in Ewing sarcoma cells is not associated with specific functional characteristics or stem cell features [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 4631.

Programmed cell death ligand 1 (PD-L1) expression is not a predominant feature in Ewing sarcomas

BACKGROUND

Programmed cell death 1 (PD-1) receptor engagement on T cells by its ligand programmed cell death ligand 1 (PD-L1) is a key mechanism of immune escape, and antibody blockade of the interaction has emerged as an effective immunotherapeutic strategy in some cancers. The role and relevance of the PD-1 checkpoint in Ewing sarcoma (EwS) is not yet understood.

PROCEDURE

Here, we investigated expression of PD-L1 and PD-1 in EwS by immunohistochemistry analysis of pretherapeutic tumor biopsies and in tumor xenografts following treatment with human T cells engineered to express a chimeric antigen receptor (CAR) against the tumor-associated antigen G_D2 . PD-L1 surface expression in EwS cell lines was assessed by flow cytometry.

RESULTS

PD-L1 expression was not detectable on tumor cells in any of the 60 EwS biopsies. Infiltrating PD-L1 positive T cells were found in one tumor, and four biopsies contained PD-1-positive T cells. Of 13 EwS cell lines, none constitutively expressed PD-L1 on the cell surface. Interferon-γ cytokine stimulation induced upregulation of the ligand on all cell lines. Adoptive therapy with CAR gene-modified T cells in a mouse model did not induce PD-L1 expression in EwS xenografts despite tumor infiltration with PD-1+ CD3+ T cells.

CONCLUSIONS

EwS cells can upregulate PD-L1 under inflammatory conditions, but do not express the ligand in the pretherapeutic tumor microenvironment or postexposure to CAR T cells. PD-1 checkpoint blockade alone is thus unlikely to evoke potent immune responses against EwS. Identification of the relevant immune evasion strategies in EwS will be vital for the development of effective immune targeting strategies.

T cell infiltration into Ewing sarcomas is associated with local expression of immune-inhibitory HLA-G

Ewing sarcoma (EwS) is an aggressive mesenchymal cancer of bones or soft tissues. The mechanisms by which this cancer interacts with the host immune system to induce tolerance are not well understood. We hypothesized that the non-classical, immune-inhibitory HLA-molecule HLA-G contributes to immune escape of EwS. While HLA-G^pos suppressor T cells were not increased in the peripheral blood of EwS patients, HLA-G was locally expressed on the tumor cells and/or on infiltrating lymphocytes in 16 of 47 pretherapeutic tumor biopsies and in 4 of 12 relapse tumors. HLA-G expression was not associated with risk-related patient variables or response to standard chemotherapy, but with significantly increased numbers of tumor-infiltrating CD3+ T cells compared to HLA-G^neg EwS biopsies. In a mouse model, EwS xenografts after adoptive therapy with tumor antigen-specific CAR T cells strongly expressed HLA-G whereas untreated control tumors were HLA-G^neg. IFN-γ stimulation of EwS cell lines in vitro induced expression of HLA-G protein. We conclude that EwS cells respond to tumor-infiltrating T cells by upregulation of HLA-G, a candidate mediator of local immune escape. Strategies that modulate HLA-G expression in the tumor microenvironment may enhance the efficacy of cellular immunotherapeutics in this cancer.

Targeting Ewing sarcoma with activated and GD2-specific chimeric antigen receptor-engineered human NK cells induces upregulation of immune-inhibitory HLA-G

Activated and in vitro expanded natural killer (NK) cells have substantial cytotoxicity against many tumor cells, but their in vivo efficacy to eliminate solid cancers is limited. Here, we used chimeric antigen receptors (CARs) to enhance the activity of NK cells against Ewing sarcomas (EwS) in a tumor antigen-specific manner. Expression of CARs directed against the ganglioside antigen G_D2 in activated NK cells increased their responses to G_D2+ allogeneic EwS cells in vitro and overcame resistance of individual cell lines to NK cell lysis. Second-generation CARs with 4-1BB and 2B4 co-stimulatory signaling and third-generation CARs combining both co-stimulatory domains were all equally effective. By contrast, adoptive transfer of G_D2-specific CAR gene-modified NK cells both by intratumoral and intraperitoneal delivery failed to eliminate G_D2-expressing EwS xenografts. Histopathology review revealed upregulation of the immunosuppressive ligand HLA-G in tumor autopsies from mice treated with NK cells compared to untreated control mice. Supporting the relevance of this finding, in vitro co-incubation of NK cells with allogeneic EwS cells induced upregulation of the HLA-G receptor CD85j, and HLA-G1 expressed by EwS cells suppressed the activity of NK cells from three of five allogeneic donors against the tumor cells in vitro. We conclude that HLA-G is a candidate immune checkpoint in EwS where it can contribute to resistance to NK cell therapy. HLA-G deserves evaluation as a potential target for more effective immunotherapeutic combination regimens in this and other cancers.

Abstract 2314: Different stimulation conditions affect the immune phenotype of GD2-specific chimeric antigen receptor-expressing T cells

The in vivo persistence of chimeric antigen receptor (CAR) modified T cells is a major prerequisite for their antitumor activity and was found to be associated with a less differentiated immune phenotype. Here, we compared two in vitro T cell stimulation conditions: coated anti-CD3/CD28 antibodies [3/28], and Dynabead stimulation of enriched CD3+ T cells [DB]. Peripheral blood T cells from three healthy donors were stimulated with either of the two methods, retrovirally transduced with the GD2-specific CAR GD2-BBz on day 2 or 3, and expanded in RPMI/AIMV medium with 50 IU/ml recombinant human interleukin-2 for 13 days. T cell expansion rates were comparable between the two stimulation conditions and independent of CAR gene expression. Transduction efficiencies, determined by staining with the GD2-CAR-specific antibody Ganglidiomab, were also comparable. The immune phenotype by expression of CD3, CD4, CD8, CD45RO and CD197 was determined by flow cytometry analysis on day 13 or 14 after initial stimulation. The proportions of central memory (TCM), effector memory (TEM) or naïve T cells (TN) within the two types of cultures were noticeably different (Table 1), with a higher proportion of non-transduced CD8+ T cells with a TCM phenotype after DB compared to CD3/CD28 stimulation (p = 0.005 for DB d2, p = 0.01 for DB d3). Compared to non-transduced T cells, CAR-expressing cells of all types of cultures had higher proportions of TCM cells (p = 0.02 for CD4+ T cells, p<0.01 for CD8+ T cells). In conclusion, we found that the stimulation conditions have a strong impact on the T cell phenotype and that retroviral CAR gene transduction can also affect T cell differentiation. The optimal T cell culture conditions for a product with sustained persistence in vivo will ultimately emerge from clinical trials. Table 1:Proportions of T cell subpopulations on day 13 or 14 (medians and ranges)xyTN: CD45RO-/CD197+TCM: CD45RO+/CD197+TEM: CD45RO+/CD197-3/28NT37.9% CD4+ (31.8-41.3) 53.0% CD8+ (20.9-72.9)20.2% CD4+ (20.2-27.0) 6.1% CD8+ (3.7-6.9)35.2% CD4+ (32.8-45.4) 35.5% CD8+ (17.0-66.3)CAR d224.4% CD4+ (15.1-28.8) 32.9% CD8+ (23.9-48.9)33.6% CD4+ (29.9-49.3) 27.7% CD8+ (17.7-38.8)37.3% CD4+ (22.4-47.2) 27.1% CD8+ (25.4-32.7)DBNT33.1% CD4+ (15.0-45.6) 53.8% CD8+ (46.8-53.8)32.6% CD4+ (29.9-45.2) 19.6% CD8+ (17.8-26.3)23.3% CD4+ (16.6-51.7) 12.1% CD8+ (10.6-15.2)CAR d211.5% CD4+ (10.9-27.7) 37.1% CD8+ (24.7-41.860.7% CD4+ (53.0-69.7) 48.0% CD8+ (34.4-65.2)16.5% CD4+ (16.1-26.6) 8.7% CD8+ (5.7-10.3)DBNT33.2% CD4+ (26.5-55.2) 51.6% CD8+ (34.2-62.5)23.0% CD4+ (21.9-40.2) 14.2% CD8+ (11.2-16.9)19.1% CD4+ (17.6-48.6) 15.2% CD8+ (13.0-19.6)CAR d317.5% CD4+ (13.4-34.5) 36.6% CD8+ (28.5-46.5)45.7% CD4+ (36.6-57.5) 38.0% CD8+ (22.2-55.1)22.0% CD4+ (21.4-32.6) 10.8% CD8+ (8.5-19.6)

Citation Format: Laurin Ochs, Bianca Altvater, Sareetha Kailayangiri, Christian Spurny, Claudia Rossig, Silke Jamitzky. Different stimulation conditions affect the immune phenotype of GD2-specific chimeric antigen receptor-expressing T cells. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2314.

Insulin-like growth factor-1 receptor (IGF-1R) inhibition promotes expansion of human NK cells which maintain their potent antitumor activity against Ewing sarcoma cells

BACKGROUND

Patients with primary metastatic or relapsed Ewing sarcomas (EwS) have a poor prognosis. While inhibitory insulin-like growth factor 1 receptor (IGF-1R)-specific antibodies have shown single agent activity in some patients with refractory disease, effective therapeutic targeting will rely on optimal combinations with conventional or innovative therapies. Specifically, combination of inhibitory IGF-1R antibodies with adoptive transfer of activated natural killer (NK) cells may have therapeutic benefit in EwS without adding toxicity.

PROCEDURE

We investigated the in vitro effects of IGF-1R targeting on the immunological profile of EwS cells and on the survival and tumor targeting capacity of K-562-activated NK cells.

RESULTS

IGF-1R inhibition reliably reduced EwS cell viability without affecting expression of immune-modulatory and MHC molecules. In NK cells, we observed a significant superior expansion following in vitro activation in the presence of IGF-1R-specific antibodies, while expression of differentiation markers and activating receptors remained unaffected. Activated NK cells coincubated with EwS cells showed potent degranulation responses unaffected by IGF-1R inhibition. These findings were reproducible in a stimulator cell-free NK cell expansion system, suggesting that direct effects of IGF-R1 antibodies on the IGF-R1 pathway in NK cells induce their activation and expansion.

CONCLUSIONS

Activated human NK cells respond to IGF-1R inhibition with superior expansion kinetics while maintaining potent antitumor responses against EwS. Combination of adoptive NK cell transfer with IGF-1R targeting may be an efficient means to eliminate minimal residual disease after conventional therapy and thereby rescue patients at the highest risk of relapse.