Search

Preclinical in vivo models are used to profile or refine CAR T therapies before advancing to human studies. Establishing long-term CAR T persistence and efficacy continues to challenge progress in the CAR T space, thus development of robust platforms that can provide longitudinal assessments of CAR T persistence and functionality is paramount. To this end, we used the NALM6-Luc ALL model to develop a flow cytometry platform that provides quantitative analysis of CAR T cells over time as well as surface markers that are documented to correlate with sustained T cell persistence and activation in vivo. Tumored NSG mice were enrolled into treatment groups based on tumor burden calculated from bioluminescence imaging (BLI) data. T cells transduced to express a CD19 CAR (or left untransduced (UTD)) were injected intravenously at 1.0E+07 cells/mouse. Tumor burden was monitored by BLI and flow cytometry was performed weekly on survival bleeds to measure CAR T persistence and examine phenotype. Treatment with CD19 CAR T cells delayed tumor growth resulting in an increase in time to progression of 107.1% compared to UTD controls. However, no regressions were observed. On day 1 post-transfer, CD3+ T cells were detectable in mice that received both CAR T and UTD cells (140+/-26 and 162+/-56 cells/uL blood respectively). T cells declined to near undetectable levels by Day 20, a point when all animals in the UTD treatment group had reached euthanasia criteria. After Day 20, T cells expanded in circulation in the CD19 CAR T treatment group reaching 695+/-327 cells/uL blood by day 40. T cell expansion coincided with exponential tumor outgrowth in the treatment group. To assess T cell functionality, flow cytometry was used to measure the expression of biomarkers for T cell activation (CD25, 4-1BB, and ICOS) and exhaustion markers (TIM-3, PD-1, and LAG-3). CD25, 4-1BB, and ICOS expression did not exceed positivity on more than 15% of CD8+ T cells and peaked by day 30 before downregulation was observed. Notably, PD-1 and LAG-3 expression levels continued to increase throughout the study, suggesting T cells were taking on an exhausted phenotype. Similar trends were observed on CD4+ T cells. To investigate whether the late phase T cell expansion was a graft vs. host response, CAR T cell measurements were compared to non-tumored animals. Expansion as well as PD-1/LAG-3 expression was only observed in tumor-bearing mice indicating the responses were tumor-specific. Taken together, these data demonstrate that CD19 CAR T cells can inhibit NALM6-Luc tumor growth in vivo and expand in circulation in an antigen-specific manner. Furthermore, CAR T cell failure to control tumor growth may be due to onset of an exhausted phenotype. Finally, we demonstrate that flow cytometry can be used to characterize T cell persistence and functionality in murine xenograft tumor models. Citation Format: David W. Draper, Derrik Germain, Stacey Roys, Olivia Nelson, Scott Wise. Preclinical assessment of chimeric antigen receptor (CAR) T persistence and functionality in the disseminated NALM6-Luc human B cell acute lymphoblastic leukemia (ALL) model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2749.

The tumor microenvironment (TME) has several characteristics that distinguish it from normal tissue, including elevated interstitial fluid pressures and hypoxia. To study the effects of TME culture conditions, commonly used target cell lines were serially passaged under hyperbaric and hypoxic conditions for a period of eight weeks. Conventionally expanded parental cell lines (normoxia) were compared against TME-acclimated tumor cell lines (TACTLs; hyperbaric and hypoxic), using RNAseq, ATACseq, and flow cytometry analysis. Commonly used target cell lines for cell therapy development were selected for TME acclimation (SKOV3, A549, JEKO1, and NALM6). TACTLs were generated using an AVATAR system, with culture conditions set at 1% O2 and 2.0 PSI, and serially passaged twice a week for a minimum of 8 weeks. Cell doubling times were measured weekly and compared against parental lines maintained under normoxic culture conditions in a conventional CO2 incubator. In-depth characterization was performed on TME-Acclimated Tumor Cell Lines (TACTLs), utilizing a multi-omic approach. Differential gene expression analysis was performed using RNAseq datasets, and surface biomarkers/targets used for cell therapy development were assessed via nanostring and flow cytometry. Growth kinetics and cell doubling times of TACTLs were initially inhibited during the first two weeks of culture under TME conditions, but eventually reached parity with their normoxia maintained parental cell lines at 6 to 8 weeks, signaling a successful adaptation of the tumor lines to low oxygen and hyperbaric conditions. RNAseq analysis revealed upregulation in glycolytic pathways and epithelial-to-mesenchymal signaling, accompanied by altered metabolic profiles. Surface target expression showed increased expression of checkpoint ligands, such as PD-L1. Cell therapy targets, such as ROR1 were also upregulated in A549 and JEKO1 cell lines passaged under TME. Preliminary drug screening experiments were conducted using checkpoint inhibitors and CAR-T candidates on TACTLs, revealing significant changes in half maximal inhibitory concentration (IC50) when compared to parental tumor lines maintained under normoxia. In summary, acclimation of tumor cell lines to TME culture conditions can provide unique mechanistic insights that can facilitate drug development efforts. Future studies will incorporate TACTLs for CDX tumor models with the goal of identifying cell therapies that work effectively in the tumor microenvironment. Citation Format: James Lim, Albert Wong, Ann Lu, Candy Garcia, Evan Massi, Yewei Xing, Ningchun Liu, Scott Wise. Long-term acclimation of target cell lines to tumor microenvironment culture condition provides mechanistic insights into cell therapy effectiveness [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 589.

Tumor microenvironments (TME) are rich in cells that potentially inhibit T cell function, such as regulatory T cells (Tregs) and myeloid derived suppressor cells (MDSCs). Many TMEs are dominated by myeloid cells, making targeting these cells an area of interest. The purpose of this study was to develop an in vitro assay system to assess the ability of novel drugs to overcome MDSC-mediated immune suppression by monitoring changes to T cell function. Arginase1 (ARG1), an enzyme that catalyzes L-arginine into L-ornithine and urea, is highly expressed in MDSCs. An ARG1 target is currently in clinical trials (INCB001158, CB-1158 from Calithera Biosciences) and was used as a target of interest in this study. MDSC generation from bone marrow cells (BM) was compared using GM-CSF and IL-6 or GM-CSF, IFNγ, and LPS. α-Difluormethylornithine (DFMO), a 2nd generation ARG1 inhibitor that irreversibly binds ARG1, which leads to increased availability of L-arginine, known to modulate T cell activity, was tested during MDSC generation. BM were harvested and characterized with a myeloid flow cytometry panel. GM-CSF, IFNγ, and LPS generated more reproducible MDSCs that express ARG1. M-MDSC-derived BM expressed 2 times more ARG1 as compared to DFMO cultured MDSC-derived BM. The cytokines and other growth factors produced by MDSC-derived BM have the potential to alter the activation state of CD8+ T cells; therefore, the effect of a suppressive TME on CD8+ T cells was monitored by co-culturing CD8+ T cells with MDSCs in vitro. MDSC-derived BM were co-cultured for 4 days with CD8+ T cells (purified from C57BL/6 splenocytes) and activated with anti-CD3/anti-CD28 beads. CD8+ T cells were characterized for proliferation using CellTraceTM violet, surface activation markers (CD69 and PD-1), and intracellular cytokines (IFNγ, TNFα, and IL-2). CD8+T cells co-cultured with DFMO-inhibited MDSC-derived BM were more proliferative, had greater surface activation, and greater intracellular cytokine of CD8+ T cells as compared to when co-cultured with MDSC-derived BM. CD8+ T cell proliferation was measured at multiple ratios of CD8+ T cells and MDSC-derived BM with a proliferation range of 88-7% when CD8+ T cells were co-cultured with MDSC-derived BM and 91-30% when co-cultured with DFMO-inhibited MDSC-derived BM. Similar trends were observed with CD69, PD-1, IFN-y, TNF-a and IL-2 expression in CD8+ T cells. Overall, the ability to monitor changes in T cell activation following MDSC has broad applications in in vitro screening of novel drug compounds that alter the TME as well as sets the experimental framework for ex vivo analysis of cells isolated from TME following treatment. Citation Format: Anita J. Zaitouna, Philip Lapinski, Amber Rowse, David Draper, Scott Wise. Inhibition of murine myeloid suppressor cells increases CD8+ T cell activation in Vitro [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2762.

Triple immunodeficient mouse models, that lack functional T cells, B cells, and natural killer cells, are necessary when evaluating human cancer cell growth and cell-based therapy activity, allowing for assessment without innate mouse immune system involvement. Due to limited colony sizes at any given vendor, study timelines can sometimes become compromised unless a sufficient alternative is validated. Site location can also be a factor, as licensing restrictions may be an issue for international corporations. In an effort to prove overall similarities across four strains of triple immunodeficient mice, the growth characteristics of BT-474 were compared in female NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice, NOD.CB17-Prkdcscid IL2rgtm1/BcgenHsd (B-NDG) mice, NOD-Prkdcem26Cd52Il2rgem26Cd22/NjuCrl (NCG) mice, and NOD.Cg-Prkdcscid Il2rgtm1Sug/JicTac (CIEA NOG) mice. An immune cell profile was also performed across these four mouse strains and evaluated via flow cytometry to observe any immune cell subset differences in blood, spleen, and tumor samples. Samples analyzed post implant were collected when subcutaneous tumors were ~500 mm3. Model development studies were completed previously to optimize tumor growth kinetics. Historic growth conditions were BT-474 cells implanted at 1.0E+07 cells/implant in the presence of an extracellular matrix into the high right axilla of female NSG mice. Using the same optimized growth conditions, NSG, B-NDG, NCG, and CIEA NOG mice were implanted with BT-474 cells. BT-474 grew well in all the tested mouse strains, producing 100% take rate. Compared to NSG mice, median tumor volume doubling times differed by 1-5 days, median times to an evaluation size of 150 mm3 (typical study enrollment tumor volume) differed by +/- ~4 days and median time to an evaluation size of 1250 mm3 (tumor volume at the end of life) differed by +/- ~10 days. BT-474 did not induce body weight loss in any of the tested mouse strains and all clinical observations were similar. Flow cytometry was performed to quantify lymphocyte and myeloid immune subsets in blood, spleen, and tumor. Overall, compared to NSG mice, the absolute counts were similar for all immune subsets measured except for regulatory T cells (Tregs). B-NDG, NCG, and CIEA NOG mice had reduced Treg numbers in the spleen of tumor-bearing mice. In the tumor, Treg numbers were lower in the B-NDG and CIEA NOG mice. No differences were observed for any immune subset in the blood. Overall, the tumor growth kinetics of BT-474 in B-NDG mice was most similar to historical data using NSG mice. The flow data also confirmed that each strain’s immune cell population generally looked very similar, with limited differences between the four strains of triple immunodeficient mice. Citation Format: Justin Snider, Derrik Germain, Lauren Kucharczyk, Anita Zaitouna, Erin Trachet, Scott Wise. Subcutaneous growth and immune cell profiling of BT-474 human breast carcinoma in four strains of triple immunodeficient mice [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 41.

Chimeric antigen receptor (CAR) T cell therapy holds great promise for the treatment of various cancers, including solid tumors. However, attempts to model the behavior and effectiveness of CAR-T cell therapies for blood cancers and solid tumors have been challenging due to the unique tumor microenvironments in which these cancer cells are found. The tumor microenvironment (TME) is often characterized by hypoxia, increased acidity, and high interstitial fluid pressures, allowing cancer cells to effectively evade immune surveillance. This immunosuppressive TME also contributes to CAR-T cell exhaustion, thereby limiting its antitumor activity and function. To address these concerns, we have developed a proprietary cell-based assay to measure CAR-T cell potency and cytotoxic function in three-dimensional (3D) in vitro cell culture system, human acute B cell lymphoblastic leukemia mouse model, and immunosuppressive tumor microenvironments. Utilizing the AVATAR system, we replicated the oxygen and interstitial fluid pressures found in the vasculature, the bone marrow and solid tumor microenvironments. Tumor cytolysis assays were conducted in these environments to measure cell exhaustion as analyzed by flow cytometry and electrical impedance. In addition, serial tumor challenge assays were performed to examine CAR-T potency and effectiveness in TME. Proof-of-concept experiments were performed using ROR1 CAR-T cells targeting the ovarian adenocarcinoma cell line, SKOV3. CD19 CAR-T were also used targeting the acute lymphoblastic leukemia cell line, NALM6. Defined ratios of effector T cells to tumor cells was assessed to model CAR-T potency in vitro and elevated CD19 CAR-T mediated cytotoxicity was confirmed with the increased ratio of effector T cells to tumor cells in both 2D and 3D culture system. CD19 CAR-T cells also exhibited in vivo dose dependent efficacy against the systemic NALM6-Luc acute lymphoblastic leukemia (ALL) mouse model, quantified by bioluminescence (BLI) image monitoring method. Multiple pressure and oxygen settings were examined to model the cross-section of the bone marrow and solid tumor microenvironments (0 PSI to 5 PSI, 1% to 10% O2). Initial results from these screening experiments show significant decline in ROR1 CAR-T mediated cytotoxicity when performed under TME conditions. However, CD19 CAR-T showed effective cell killing under TME conditions. Interestingly, acclimating and expanding ROR1 CAR-T cells to high pressure and decreased oxygen culture conditions improved potency levels and warrants further investigation. In summary, we observe CAR-T cells comprise the tumor cell killing ability in both in vitro 3D and in vivo animal models. We also describe a physiologically relevant potency assay that incorporates hyperbaric and hypoxic incubation technology to predict the behavior of cell therapies in immunosuppressive tumor microenvironments. Citation Format: Yelena Bronevetsky, Evan Massi, Candy Garcia, Ningchun Liu, Yewei Xing, Natalie Czeryba, Scott Wise, James Lim. Functional potency assay predicts CAR-T effectiveness in tumor microenvironment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1782.

Lung cancer remains a devastating cancer diagnosis around the world, resulting in 18% of all cancer-related deaths. Lung cancer often metastasizes to the brain, significantly reducing life expectancy. Up to 7% of patients with non-small cell lung cancer (NSCLC) already have brain metastasis when they are first diagnosed, and 20%-40% of patients with NSCLC will develop this complication during disease progression. Currently there are no targeted therapies specific for brain metastases, and the blood-brain barrier can pose a physiologic impediment to many cytotoxic drugs and antibody-based therapies. Here, we will discuss the development and use of xenograft models to address metastatic brain disease via a direct intracranial implant coupled with a subcutaneous “primary” tumor to effectively allow evaluation of the response to treatment at both locations. Treatment with targeted therapies could have significantly different penetration and absorption rates, which will alter pharmacokinetic and/or pharmacodynamic (PK/PD) assessment across either tissue. The dual implant technique can be a powerful tool to assess the simultaneous impact of treatment on established metastatic disease and primary tumor. We have characterized the dual disease induction parameters for two human NSCLC cell lines: NCI-H1975-Luc and PC-9-Luc. Both cell lines have been transfected with luciferase to allow for bioluminescence imaging (BLI) to monitor the intracranial disease progression. Both NCI-H1975-Luc and PC-9-Luc are of interest to the research community due to their unique mutational EGFR T790M status. Consistent with the historical and published data, the H1975-luc and PC-9-luc models behaved as expected, producing reliable tumor progression with minimal intragroup variability. In each case, subcutaneous tumors (primary site) reached an evaluation size of ~1000 mm3 in ~20 days, and intracranial tumors (metastatic site) exhibited a total flux doubling time of ~2 days by BLI. Treatment with Osimertinib, an approved first-line treatment for EGFR+ NSCLC patients, was effective against both the subcutaneous and intracranial tumors. Treatment with Osimertinib at 25 mg/kg, given orally, produced 100% complete regressions (CR) against the subcutaneous tumors, which never re-grew prior to study termination on Day 26 post implant and 100% partial regressions (PR) against the established intracranial tumors, which re-grew following the conclusion of treatment at the same rate as the control brain tumors. These data show a difference between the responsive nature of subcutaneous and intracranial NSCLC tumors representing the difference in responsiveness of primary tumors and brain metastases. Utilizing this in vivo dual implant technique can be effective in comparing and assessing the ability of targeted therapies to inhibit both primary and metastatic diseases. Citation Format: Erin Trachet, Scott Wise. A novel technique to preclinically assess the ability of targeted therapies to inhibit both primary and metastatic non small cell lung carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 37.

As it is rarely discovered early, the prognosis for people with pancreatic cancer is poor, with an average 5-year survival rate of 10% and only 3% for those with metastatic illness. Pancreatic cancer patients have few therapeutic options and innovative therapies are sorely needed to improve treatment of the disease. This is mainly due to the late diagnosis and partially due to the biology of the disease. Capturing the components of the tumor microenvironment, which serve as both a physical barrier and a source of stromal-driven resistance to therapeutics, is one of the issues faced by drug developers searching for agents to combat pancreatic cancer. We have created a 3D model of the pancreatic tumor microenvironment, the Reconstructed Pancreas (r-Pancreas), using Mia PaCa-2, PANC-1, and BxPC-3 pancreatic tumor cell lines embedded in an extracellular matrix (ECM) that was formulated to recapitulate the typical ECM of pancreatic tumors. We have shown that gemcitabine was effective against pancreatic tumor cells cultured in r-Pancreas (IC50=0.4-0.8µM), while 5-fluorouracil was ineffective (IC50 not reached), mimicking clinical response. In nude mice, subcutaneous PANC-1 tumors exhibited similar responses to gemcitabine and 5-fluorocurcil. Treatment with gemcitabine resulted in a Day 63 median ΔT/ΔC of 37% while 5-fluorouracil resulted in a Day 63 median delta ΔT/ΔC of 102%. Incorporating a collagen-rich capsule increases the physiological relevance of the r-Pancreas model because it has been shown that human pancreatic tumors have an outer layer of stiff ECM that functions as a physical barrier and prevents drugs from penetrating the tumor. Recently, we showed that, at least in the Mia PaCa-2 and BxPC-3 tumor cell lines, introducing such a collagen capsule dramatically reduced sensitivity to gemcitabine. Disrupting the collagen capsule with MMP-9 restored the sensitivity of the tumor cells to gemcitabine demonstrating that the capsule provides a physical barrier to drug entry. Additionally, creating a co-culture of primary activated pancreatic stromal cells and pancreatic tumor cell lines produces a more complete 3D model that permits the testing of potential therapeutic agents inside the pancreatic tumor microenvironment. Hence, we have created a 3D model where pancreatic tumor cells were co-cultured with human pancreatic fibroblasts. According to our results, the IC50 values for pancreatic tumor cell lines treated with gemcitabine increased by twofold when co-cultured with activated pancreatic fibroblasts compared to control sets of tumor cells alone (IC50=1-3µM). Together, these data imply that, to be clinically relevant, in vitro models of pancreatic cancer must integrate tumor-specific components of the microenvironment, including the collagen capsule and activated fibroblasts. Citation Format: Arnat Balabyev, Justin D. Phillips, Michael Steffey, Eleanore J. Kirshner, Aayushi Ahlawat, Arlette H. Uihlein, Daniel Saims, Scott Wise, Julia Kirshner. 3D Reconstructed Pancreas: A model capturing the unique tumor microenvironment and stromal architecture of pancreatic cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4561.

Cancer immunotherapies reprogram the patient’s immune system to mount a coordinated response against a malignant target. T cells engineered to express Chimeric Antigen Receptors (CARs) through transduction with a lentiviral vector represent an effective strategy to specifically eliminate cancerous cells from a patient. Currently, five CAR T cell therapies are approved by the FDA for the treatment of hematological malignancies. With the recent clinical and regulatory success of CAR T cell therapies, the next generation of CAR T cells are undergoing preclinical development. Labcorp Drug Development produces CAR T cells to support the development of new preclinical strategies. Here, the CAR T Cell Generation Service is demonstrated using an anti-CD19 CAR T cell as an example. Using peripheral blood mononuclear cells from healthy, human donors as a T-cell source, CAR T cells were produced by lentivirus transduction. Flow cytometry featuring a CAR-specific monoclonal antibody was used to determine transduction efficiency. To assess in vitro activity, co-cultures of T cells and CD19-expressing NALM6 cells were used to measure cytotoxicity. Proinflammatory cytokine production was analyzed using a Meso Scale Discovery V-PLEX assay. For evaluation of in vivo efficacy, a disseminated NALM6-Luc-mCh-Puro human acute B cell lymphoblastic leukemia model in female NSG mice was conducted. After T cells were treated with lentivirus, approximately 25% of total cells were CD3+/CAR+. Greater than 95% of all target NALM6 cells were killed by a high dose of anti-CD19 CAR T cells. In contrast, only 33% of target cells were killed by untransduced (UTD) T cells at the highest concentration tested. Cytotoxicity against the negative control MV-4-11 cells was equal for a high dose of UTD and a high dose of anti-CD19 CAR T cells. When co-cultured with CD19-expressing NALM6 cells, anti-CD19 CAR T cells produced proinflammatory cytokines including IFN-γ. Relative to mock-treated control mice, treatment with anti-CD19 CAR T cells at high, medium, and low doses increased the time of in vivo disease progression by 175%, 125%, and 16%, respectively. Anti-CD19 CAR T cells generated in our lab were specifically active against CD19-expressing target cells. As a premier contract research organization, Labcorp Drug Development is experienced in producing, handling and culturing T cells and CAR T cells for in vitro and in vivo early discovery studies. Preclinical screening of multiple CAR T cell candidates, alone or in combination with other agents, would facilitate the identification and selection of CARs with the most favorable activity profile for progression through the drug development pipeline. By providing a reliable source of CAR T cells, we are supporting early discovery studies and the development of therapeutic strategies in oncology. Citation Format: Andrew Karalewitz, Natalie Czeryba, Yewei Xing, Derrik Germain, Philip Lapinski, Sheri Barnes, Mark Cameron, Daniel Saims, Amber Rowse, Heidi Nielsen, Scott Wise. Chimeric antigen receptor (CAR) T-cell generation for therapeutic testing in the disseminated NALM6 human B-cell acute lymphoblastic leukemia mouse model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2835.

In preclinical drug development, there is an urgency to create models that are predictive and relevant at the same time. To support this effort, early drug development emphasizes establishing multiple strategies that address treatment response, resistance, and combination effects, all with rapid turnover. Given the ever-increasing relevance of 3D in vitro models, Labcorp employs 3D model systems co-developed with zPREDICTA. These organ-specific models can support tumor cell growth and mimic the composition and architecture of both solid tumors and hematological malignancies. While 2D in vitro assays are preferred for screening efforts of chemical libraries, mainly because of cost, they do not recapitulate the tumor architecture and the microenvironment that is more comparable to in vivo models. This can result in differences in pharmacological potencies of standard of care (SoC) agents when compared to the 3D and in vivo models. zPREDICTA 3D assays serve as a tool for in vitro screening of tumor cell lines, which in long term cultures develop similar characteristics to in vivo disease states. Specifically, this in vitro system correlates well with breast and lung carcinoma in vivo models when matching rank order of potencies of known SoC agents. zPREDICTA tumor-specific 3D systems as well as 2D cultures and in vivo models were used to compare and contrast the activity of a set of anticancer agents. Human MDA-MB-231 cells were sensitive to paclitaxel both in 2D and in the 3D r-Breast (IC50 of 7.9nM and 6.1nM, respectively) as well as in vivo, with complete growth inhibition at the dose of 15mg/kg. Paclitaxel was cytotoxic in 2D cultures of murine 4T1 breast tumor cells (IC50=59nM), but not in the 3D reconstructed mouse breast model (r-mBreast) or in vivo, where the syngeneic 4T1 model was refractory to paclitaxel up to 10mg/kg. Similarly, NCI-H460 cells were sensitive to paclitaxel in 2D (IC50=2.3nM), while minimal response was seen in the 3D r-Lung model, and in vivo there was minimal response seen at a 20mg/kg treatment. These same cells were refractory to erlotinib in 2D, 3D, and in vivo. Additional SoC agents were profiled in these models and will also be discussed. Taken together these data demonstrate that zPREDICTA’s organ-specific 3D models mimic in vivo response with a higher fidelity than the standard 2D cultures. Additional in vitro reconstructed 3D histotype models are being developed and validated with the goal of providing faster and more predictive in vivo results. Citation Format: Sumithra Urs, Michael Steffey, Ying Qu, Matthew Ryou, Julia Kirshner, Scott Wise. Reconstructed plate-based 3D screening assays that predict in vivo responses [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 283.

Background: Five checkpoint immunotherapies that target the PD-1/PD-L1 axis are currently FDA approved. Novel approaches are helping to identify new combination treatment strategies for therapeutic intervention of bladder cancer, which is the thirteenth leading cause of cancer-related deaths. Using anti-mPD-1 treatment in a murine bladder cancer model MB49, we show that non-targeted immune gene expression profiling combined with flow cytometry provides a gene and cell-specific signature for the tumor microenvironment, which aids in the identification of targets for novel treatment approaches. Methods: C57BL/6 mice with established MB49 tumors were treated with anti-mPD-1 or isotype control antibodies. Tumors were collected 3 days after the last treatment. Treatment-induced immunophenotypic changes were examined in tumor-infiltrating immune subsets using T cell- and myeloid-focused flow cytometry panels. We used the mouse PanCancer IO 360™ Nanostring panel for transcriptomic analysis of 770 genes and the ROSALIND™ platform (OnRamp BioInformatics) to identify differentially regulated genes between treatment groups. Results: Treatment of tumors with anti-mPD-1 showed moderate anti-tumor activity, with a 58% tumor growth inhibition at day 18 post-implant. Immunophenotyping by flow cytometry revealed that anti-mPD-1 triggered an increase in tumor-infiltrating CD8+ T cells compared to control animals. Additionally, the CD8+ T cell phenotype was altered by treatment, with increased frequency of ICOS and LAG-3 in CD8+ T cells in tumors from treated animals. Changes in the T cell compartment also included reduction in the proportion of central memory CD8+ and CD4+ T cells compared to controls. In the myeloid compartment, iNOS expression increased in tumor-associated macrophages from treated animals. NanoString analysis revealed 62 genes were differentially regulated in tumors from treated animals compared to controls. ROSALIND analysis classified 30 of the genes as regulators of interferon, cytotoxicity, antigen presentation, and cytokine signaling. Among the genes upregulated by anti-mPD-1 were IDO, TIM-3, and CSFR1, which can promote tumor growth and are clinical targets being investigated for new immunotherapies. Conclusions: NanoString analysis complemented immunophenotyping to provide a comprehensive profile of the MB49 tumor model. Together, these data demonstrate that anti-mPD1 increases T cell recruitment into the tumor and upregulates the expression of genes known to enhance T cell recruitment and anti-tumor activity. iNOS protein upregulation suggests that anti-mPD-1 treatment may also exert effects by driving M2 macrophages towards an M1 phenotype. Further investigation may elucidate clinical implications for inhibitors of these gene products as treatment options in combination with anti-mPD-1. Citation Format: Philip E. Lapinski, David W. Draper, Scott Wise. Comprehensive cell- and gene-based tumor profiling using flow cytometry and Nanostring in a murine bladder cancer model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1633.

In order to predict efficacy and toxicity of compounds advancing in the oncology drug discovery pipeline, and to minimize requirements for multiple animal studies, the development of in vitro assays that recapitulate in vivo conditions is increasingly important. While high throughput screening favors 2D assays, both for cost and relative speed in culturing and plating, they do not recapitulate the tumor architecture and the microenvironment that is more comparable to in vivo models. We now report this has been accomplished using zPREDICTA's organ-specific three-dimensional (3D) culture models which support long-term culture of carcinoma cells. Physiologic extracellular matrix (ECM) components are critical in maintaining the physical and spatial microenvironment as well as tumor phenotypes in reconstructed bone marrow (r-Bone) and mouse breast (r-mBreast) models. The r-Bone ECM allows for co-culture of tumor cells with mesenchymal stem cells, fibroblasts, lymphoid, myeloid, and CAR-T cells. The data here describes clear differences in the in vitro pharmacology of several standard of care agents in 2D, in adherent or suspension cultures, compared to the recapitulated 3D tumor microenvironment of the human hematopoietic compartment or the mouse breast ductal epithelium. For example, we find that cisplatin has a 10-30 fold higher IC50 in 4 different murine cell lines (MMTV-PyMT, 4T1, EMT6, and E0771) when cultured in the r-mBreast model than when those cells are adhered to the polystyrene in a 2D culture. Cytotoxicity assays such as CellTiter-Glo® can also be performed in a high throughput format in both 96 and 384-well 3D assay formats and we demonstrate equivalent bortezomib IC50 values in RPMI 8226 cell r-Bone cultures in both plate types. Additionally, cells can be isolated from the matrix after test agent exposure and analyzed by an appropriate downstream application, such as flow cytometry, where we show anti-CD19 CAR-T cell killing of CD19+ NALM6 cells in r-Bone cultures. These 3D models represent new early opportunities to screen test agents in a platform that is designed to more closely mimic in vivo tissue microenvironments and serve as a more relevant tool for multi drug screening and as a prerequisite to animal studies. Citation Format: Michael E. Steffey, Ying Qu, Andrew Karalewitz, Sumithra Urs, Julia Kirshner, Scott Wise. Human and mouse reconstructed plate-based 3-dimensional culture assays that mimic the tumor microenvironment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2643.

Radiotherapy is a highly utilized clinical treatment modality. More than 50% of all cancer patients receive some type of radiation therapy during the course of their illness. In mouse models, radiation treatment has been shown to increase the level of tumor antigen presentation and the variety of peptides available for cross-presentation. Current work in the field focuses on using radiation as a tool to bridge the gap from tumor equilibrium to tumor elimination, which could improve the response rate of immuno-oncology agents. 4T1 is a murine breast cancer model known to have a large percentage of myeloid derived suppressor cells (MDSC) making the model resistant to many immunotherapies and is considered an immunologically cold tumor. We hypothesized that treatment with focal radiation (RT, Xstrahl) would sensitize 4T1 tumors to anti-CTLA-4 treatment. To determine an appropriate dose of RT, mice with 4T1-Lu2 tumors were treated with a single, focal dose of 5, 10 or 20Gy RT. No response was seen at 5Gy, 10Gy resulted in 10 days growth delay, and 20Gy resulted in 16 days growth. In subsequent work, mice were placed into groups and treated with either isotype control (10mg/kg, MPC-11), anti-CTLA-4 (10mg/kg, 9D9), 10Gy RT, or the combination. Mice were monitored over time for changes in primary tumor volume by caliper measurements and for metastatic disease by bioluminescence imaging (BLI) of the thoracic region. Satellite groups were included for immunohistochemistry and the evaluation of changes in tumor infiltrating lymphocytes. Median tumor growth delay with anti-CTLA-4, RT, or the combination, was 2.4, 3.5 and 9.4 days, respectively. Time to progression, to 1,200mm3 tumors, was increased by 25% in the combination group but not significantly in the monotherapy groups. While all mice showed evidence of metastatic disease, mice in the combination group displayed a lower level of BLI signal on day 30 when compared to other groups. To examine the effects on immune cell infiltration, 11 subsets were profiled by flow cytometry. B cells and regulatory T cells were significantly reduced (>95% and 90%, respectively) in the combination group when compared to an untreated group. B cell reductions that exceeded 90% were also observed in some animals within the isotype control group. Examination of the CD8+ T cell phenotype demonstrated that combination therapy, but not monotherapies, increased expression of both CD69 and PD-1 on CD8+ T cells suggesting an enhancement of anti-tumor cytotoxicity in these cells. Interestingly, we found that the combination selectively reduced the levels of phosphorylated STAT3 in MDSC subsets. Collectively, this intimates that anti-CTLA-4 and RT selectively disrupt STAT3 signaling in MDSCs and combine to enhance CD8+ T cell activity, which may play a role in the 4T1 tumor growth delay observed. Citation Format: Maryland Rosenfeld Franklin, David Draper, Sumithra Urs, Scott Wise. Preclinical use of focal radiation and immune checkpoint blockade to improve therapeutic response in an immunologically cold tumor [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 535.

Glioblastoma (GBM) is a fast-growing, aggressive type of central nervous system tumor that forms on the supportive tissue of the brain. Primary treatment options for patients with GBMs have not changed much in many years, and still include surgery, radiation and/or chemotherapy treatment with temozolomide (TMZ). GBM is an incurable disease where there is a desperate need for continued development of novel therapeutic options. A number of preclinical models exist to support the early work needed to develop these new drugs. Here we describe three different preclinical models of brain cancer, U87MG-Luc, GL261-Luc and BT142. Each model was evaluated following orthotopic implantation into the brain. As radiation therapy remains a steadfast approach to GBM, we evaluated each model for its response to radiation and tracked disease progression and response to treatment with either bioluminescence imaging (BLI) or magnetic resonance imaging (MRI) along with traditional survival (morbidity/mortality) analysis. U87MG-Luc is a human glioblastoma cell line that was developed from an astrocytoma. We found that orthotopic implant of U87MG-luc results in reliable growth with a tumor doubling time of 4 days and a median survival time of 45 days. We tested this model with radiation therapy (2.5Gy for 5 days on, 2 days off, for 2 cycles) and found a 70% partial response rate as measured by BLI. We also tested this model with TMZ therapy (33.3mg/kg; orally every day for 5 days) and found a significant delay in disease progression, resulting in a 93% increase in life span. GL261-Luc is a murine glioblastoma model that grows quite aggressively in the mouse brain. We find a tumor volume doubling time of 2 days and a median survival time of 14 days. As this model grows in a mouse strain with an intact immune system it allowed us to evaluate the tumor infiltrating immune cell populations. While we found a low number of CD45+ cells infiltrating into GL261-Luc tumors, immune population changes over time will be described. In this model we tested increasing doses of radiation and found that a single dose of 15Gy was curative whereas the tumors were more moderately responsive to a single dose of 7.5Gy (30% tumor regressions). We then combined 7.5Gy with the immune checkpoint inhibitor anti-PD-1 and saw an improved outcome with a 100% response rate. BT142 is a human anaplastic oligoastroytoma that is relatively unique due to its isocitrate dehydrogenase (IDH1/2) mutational status which results in an accumulation of the oncometabolite 2-hydroxygluarate (2-HG) in the tumor. This tumor line grows in the NOD SCID mouse strain and we found a tumor doubling time of 5 days, as measured by MRI. The BT142 model has proven to be sensitive to radiation following a single dose of 10Gy but less responsive when lower dosage levels were administered on a fractionated schedule. Thus, a variety of human and mouse GBM models exist to enable further drug discovery and development efforts for this intractable disease. Citation Format: Erin E. Trachet, Sumithra Urs, Alden Wong, Scott Wise, Maryland Rosenfeld Franklin. Radiation response in preclinical orthotopic models of brain cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1928.

To advance immuno-oncology drug development, there is a need for well-characterized syngeneic tumor models. In addition to subcutaneous modeling, orthotopic models are equally important for preclinical development as they are posited to have improved clinical translatability compared to subcutaneous implantation. To this end, we have generated both subcutaneous and systemic A20 murine B cell lymphoma models. Specific to orthotopic modeling, a luciferase-enabled A20 model (A20-luc) was created to monitor tumor progression by bioluminescence imaging. Systemic parental A20 or A20-luc models show slight differences in median survival (34.5 days and 39 days, respectively) but similar take rates and clinical manifestations, including abdominal distension and metastatic foci in the liver. These data suggest that parental A20 and A20-luc behave similarly in systemic disease. To understand how checkpoint blockade and costimulatory agonists act on A20, anti-mPD-1, anti-mPD-L1, anti-mCD137, and anti-mGITR were evaluated for efficacy against both established subcutaneous and systemic A20 models. Our data indicate that neither anti-mPD-1 nor anti-mGITR demonstrate antitumor activity in either model. Anti-mPD-L1 is inactive in systemic disease but shows some efficacy on A20 subcutaneous tumors with delayed growth observed in 40% of mice. Contrastingly, the costimulatory agonist anti-mCD137 elicits a stronger response, with 50% of established subcutaneous A20 tumors having delayed growth including two complete responses and one tumor free survivor. Similarly, in systemic A20 disease, 50% of animals treated with anti-mCD137 exhibited a profound decrease in luciferase signal. While modestly responsive, both A20 and A20-luc show promise for use in combination strategies. To determine whether this modest activity is due to an infiltration of immunosuppressive cells into the tumor, the composition of immune cells in established subcutaneous A20 tumors was evaluated. Despite modest activity, 50% of total CD45+ infiltrate were T cells and NK cells, 7% were MDSCs and macrophages, and the remaining were dendritic in nature. As high T cell infiltrate is more common in tumors responsive to immunotherapy, these infiltrated T cells are hypothesized to lack signals required for activation or are reactive to non-tumor-associated antigen. Similar assays with metastatic foci resulting from systemic disease are planned. Both subcutaneous and orthotopic A20 models exhibit favorable characteristics for evaluation of investigational agents. Response to immunotherapy is minimal to moderate in both models but are generally aligned. The high infiltrate of T cells in the subcutaneous model have potential for exploitation of this model in development of agents targeting T cells. Taken together, our results indicate that A20 is a favorable syngeneic tumor model for both subcutaneous and orthotopic applications. Citation Format: Sheri R. Barnes, Sumithra Urs, David Draper, Scott Wise, Maryland Rosenfeld Franklin. Systemic and subcutaneous modeling of A20 murine B cell lymphoma in mice: A comparative assessment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3719.