Patient-derived Xenograft Tissue Research Articles

THER-12. PRECLINICAL EVALUATION OF NERATINIB PLUS T-DM1 IN ORTHOTOPIC PDX MODELS OF HER2-POSITIVE BREAST CANCER BRAIN METASTASES

Breast cancer brain metastases (BCBM) are a major cause of morbidity and mortality, despite multimodal management including surgery, radiotherapy, and systemic therapies. There is an urgent need to develop novel, efficacious alternatives. Neratinib is an orally bioavailable, irreversible pan-HER tyrosine kinase inhibitor that is FDA-approved in the extended adjuvant treatment setting for HER2-positive, early breast cancer (NCT00878709). Ado-trastuzumab emtansine (T-DM1) is an antibody-drug conjugate with reported single-agent activity against HER2-positive BCBM. Here, we used HER2-positive orthotopic patient derived xenograft (PDX) models of BCBM to test if combining neratinib with T-DM1 could improve tumor response. PDX cells are labelled with luciferase to allow tumor growth measurement in vivo. We found that neratinib is able to reduce phosphorylated HER2 in an orthotopic PDX tumor derived from HER2-positive BCBM, indicating that neratinib can cross the BBB and inhibit HER2 activation in BCBM PDX tissues. However, in both HER2-positive DF-BM354 and DF-BM355 PDX models, single agent neratinib did not block orthotopic tumor growth compared to vehicle control as monitored by bioluminescence measurements. In contrast, combined treatment of neratinib with T-DM1 significantly reduced tumor growth compared to single agent treatment with neratinib or T-DM1 at earlier time points in both models. At later time points, the combined treatment is comparable to T-DM1 alone in DF-BM354 model, but significantly prolong the survival of mice bearing DF-BM355 tumors. These data warrant further testing of neratinib alone and in combination with T-DM1 in additional BCBM PDX models to better understand drivers of resistance and susceptibility to HER2-inhibitors in HER2-positive BCBMs. Furthermore, they support the launch of a prospective clinical trial (NCT01494662) to test the efficacy and tolerability of T-DM1 in combination with neratinib in patients with progressive HER2-positive BCBM.

Abstract 3780: Spatially resolved transcriptional patterns of PARP inhibitor-resistant high grade serous ovarian cancer PDX tissue

Abstract Ovarian cancer is the leading cause of mortality among gynecological malignancies; 70% of patients who initially respond to therapy eventually relapse and die. Within highly heterogeneous tumors, resistance originates from a small number of cells that express beneficial genes and receive favorable cues from their microenvironment. Understanding the mechanisms of drug resistance therefore requires the knowledge of which cell types compose the tumor, which genes they express, and where they are located respective to other cells and tissue landmarks. To this end, we have explored cell type and transcriptional heterogeneity in homologous recombination deficient patient-derived xenograft (PDX) models of high grade serous ovarian cancer (HGSOC) treated with the PARP inhibitor talazoparib. We are using single-cell RNA sequencing (scRNA-seq) to identify the discrete cell types present in drug naïve and drug resistant PDX tissue and to characterize the transcriptional changes driving the resistant phenotype. We are using single-molecule hybridization chain reaction (smHCR) to fluorescently label mRNAs and diSPIM light sheet microscopy to visualize fluorescently labeled mRNAs within large volumes of PDX tissue. We are currently developing computational strategies to deconvolve the 3D spatial organization of PDX cell types and to understand where and when resistance genes are expressed. This integrated scRNA seq and smHCR microscopy framework has allowed us to identify rare cell populations within HGSOC PDXs and to visualize their 3D localization relative to other cell types and tissue features, and it has the potential to elucidate how transcription of key genes supports transition to a resistant state. The insights gained from single cell transcriptional and 3D spatial profiles has the potential to greatly advance our understanding of how the 3D tumor tissue microenvironment allows and encourages rare cells with pre-resistant transcriptional programs to escape PARP inhibitor treatment. Further, this research has the potential to be predictive of disease progression and to suggest additional therapeutic targets that could be exploited as part of novel and/or personalized combination therapy regimens. An additional advantage is that this approach could be generalizable to many types of cancer to characterize the interplay between tissue organization and diverse processes such as tumor evolution, metastasis, and drug resistance. Citation Format: Benjamin King, Elke Van Oudenhove, Selim Misirlioglu, Ernesto Arostegui Fernandez, Douglas Levine, Timothee Lionnet. Spatially resolved transcriptional patterns of PARP inhibitor-resistant high grade serous ovarian cancer PDX tissue [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 3780.

Abstract 4832: Preclinical evaluation of neratinib plus T-DM1 in orthotopic PDX models of HER2-positive breast cancer brain metastases

Abstract Up to half of patients with metastatic HER2-positive breast cancer will develop brain metastases. Breast cancer brain metastases (BCBM) are a major cause of morbidity and mortality, despite multimodal management including surgery, radiotherapy, and systemic therapies. Therefore, there is an urgent need to develop novel, efficacious therapies. Neratinib is an orally bioavailable, irreversible pan-HER tyrosine kinase inhibitor that is FDA-approved in the extended adjuvant treatment setting for HER2-positive, early breast cancer. Neratinib has only modest activity as a single agent in clinical trials of patients with HER2-positive brain metastases. Though the combination of neratinib and capecitabine results in CNS responses in up to half of patients, patients eventually develop drug resistance, and toxicities have been a concern-thus exploration of alternative neratinib combinations is of significant clinical interest. Ado-trastuzumab emtansine (T-DM1) is an antibody-drug conjugate with reported single-agent activity against HER2-positive BCBM. Here, we used HER2-positive orthotopic patient derived xenograft (PDX) models of BCBM to test if combining neratinib with T-DM1 could improve tumor response. PDX cells are labelled with luciferase to allow tumor growth measurement in vivo. We found that neratinib is able to reduce phosphorylated HER2 in an orthotopic PDX tumor derived from HER2-positive BCBM, indicating that neratinib can cross the BBB and inhibit HER2 activation in BCBM PDX tissues. However, in the HER2-positive DF-BM354 PDX model, single agent neratinib did not block orthotopic tumor growth compared to vehicle control as monitored by bioluminescence measurements. In contrast, combined treatment of neratinib with T-DM1 significantly reduced tumor growth compared to single agent treatment of neratinib or T-DM1 alone at earlier time points. At later time points, the combined treatment is comparable to T-DM1 alone. These data warrant further testing of neratinib alone or in combination with T-DM1 in additional BCBM PDX models to better understand drivers of resistance and susceptibility to HER2-inhibitors in HER2-positive BCBMs. Furthermore, they support the launch of a prospective clinical trial (NCT01494662) to test the efficacy and tolerability of T-DM1 in combination with neratinib in patients with progressive HER2-positive BCBM. Citation Format: Jing Ni, Yanzhi Wang, Irmina Diala, Sheheryar Kabraji, Rachel Freedman, Nancy Lin, Jean Zhao. Preclinical evaluation of neratinib plus T-DM1 in orthotopic PDX models of HER2-positive breast cancer brain metastases [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 4832.

Abstract 4710: Surface protein marker and single cell gene expression profiling of individual tumor cells dissociated from small cell lung cancer pdx mouse models can be correlated with in vivo sensitivity to the p70S6K/AKT1/3 inhibitor M2698

Abstract Small Cell Lung Cancer (SCLC) is characterized by rapid tumor growth and currently, there are few therapeutic options or predictive biomarkers. Patient derived xenograft (PDX) models are capable of recapitulating solid tumor growth including the intra-tumor heterogeneity (ITH) observed in the original patient tumor. The present study aimed to correlate surface marker profiles of SCLC PDX models with previously observed drug responses to M2698, a potent, selective inhibitor of p70S6K and AKT 1/3, and investigate ITH through gene expression profiling of tumor cell subpopulations. Drug response data had been established in a preclinical screen of 45 PDX models of SCLC, in which two mice were implanted subcutaneously with tumors from each model; one mouse was treated with vehicle while the other was treated with M2698 25 mg/kg QD po until the tumor in the vehicle-treated mouse reached ~1200 mm3. Tumor control (stasis or regression) was seen in 12 (27%) of the models. We showed previously that cell surface marker profiles of PDX tumor tissue demonstrated high intra-model reproducibility for many surface markers and uniquely associate with each model. Also, distinctly heterogeneous markers were identified that allowed FACS sorting of tumor cell subpopulations with similarly distinct gene expression profiles. To build on these data, a subset of models that were the most and the least sensitive to M2698 (n=7) in the above described screen were selected for implantation into a new set of mice. Tumors were profiled once they reached ~800 mm3. We evaluated 80+ markers commonly used to identify tumor initiating cells (e.g. CD44, CD90, CD133, CD166, CD184), EMT or aggressiveness (e.g. CD166, EphB2, CD324, CD325), poor outcome in SCLC (CD49b, CD221, claudin3) or drug targets (e.g. CD184, EGFR, Her2) to establish extensive marker profiles. Our data reveal that surface marker profiles in these models allow a meaningful subclassification of SCLC PDX tissue. Correlation of these profiles with efficacy of M2698 (above) suggest that surface markers may have predictive value. Our results prequalify a number of these markers for validation. Furthermore, several models harbor cell subpopulations identifiable by various surface markers. Along with their distinct gene expression profiles this suggests an equilibrium between functionally different compartments within a lesion. In one case, the ratio of two populations shifted concurrently with growth differences, raising the possibility of a dynamic relationship between this equilibrium and the growth stage. Overall, our workflow may provide tools for sample characterization, quality control and elucidation of cellular response markers to varying selective pressures, such as drug challenges. Citation Format: Warren Porter, Eileen Snowden, Friedrich Hahn, Mitchell Ferguson, Frances Tong, William S. Dillmore, Anderson Clark, Hong Zhang, Rainer Blaesius. Surface protein marker and single cell gene expression profiling of individual tumor cells dissociated from small cell lung cancer pdx mouse models can be correlated with in vivo sensitivity to the p70S6K/AKT1/3 inhibitor M2698 [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 4710.

Abstract 1243: Effect of antigen retrieval on immuno-based RNA microdissection methods

Abstract Introduction: Clinical applications and research studies routinely utilize immunohistochemistry (IHC) approaches on tissue sections. The aim is to selectively target antigens (proteins) in the cells of a tissue section by exploiting the principle of antibody binding, thus enabling target specificity. The antigen retrieval (AR) process employed at the beginning of most IHC protocols increases epitope availability and improves staining characteristics; however, the procedure can damage nucleic acids, in this case RNA, to an unknown extent. Experimental Procedures: To better understand the effects of AR on RNA quality and quantity, model system samples (lung cancer patient derived xenograft tissue) were subjected to a commonly employed AR method [heat induced epitope retrieval (HIER)] and the effects on RNA were assessed by Qubit, Fragment Analyzer, and digital droplet PCR (ddPCR). The ddPCR experiments were performed following reverse transcriptase (RT) product generation and construction of cDNA, which then served as input to the ddPCR assay. Absolute quantitation of gene expression levels and expressed mutations were evaluated by ddPCR. Data Summary: The results showed that HIER resulted in optimal staining characteristics, but induced significant damage to RNA, producing extensive fragmentation and decreased yields. In general, fragmentation showed an inflection point at ~200 nt transcript length. Variation in the HIER protocol mitigated but did not eliminate the negative effects. Apart from the AR procedure, the IHC process itself also resulted in a significant decreased yield of RNA. Of note, in these experiments DAB chromogen was used exclusively in the IHC process. However, in spite of the observed deleterious effects on RNA, none of the AR methods combined with IHC negatively affected RT product generation and PCR amplification of small amplicons. Regarding the detection of RT product under different conditions, the size of the amplicon was found to make a difference, with the detection of longer amplicons being more difficult. In addition to a number of successful human and mouse gene expression measurements, KRAS gene mutations were successfully identified in the clinical cases under all conditions. Conclusions: The data indicate that RNA recovered from histology slides after standard AR and IHC processing can be successfully employed for genomic applications utilizing RT product formation (cDNA) for gene expression and detection of expressed mutations, especially if the detection methods are based on relatively short length nucleic acids. Studies that require larger RNA fragments such as long-read sequencing should avoid the use of HIER. Citation Format: Donald J. Johann, Meeiyueh Liu, Adam Roberge, Sarah Laun, Michael Emmert-Buck, Michael Tangrea. Effect of antigen retrieval on immuno-based RNA microdissection methods [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 1243.

Abstract 2503: Predicting drug sensitivity to a novel BCL-2 and BCL-xL dual inhibitor in solid tumor PDX trials

Abstract Background: BCL-2, an anti-apoptotic protein and oncogene, has been considered one of the top 10 targets for cancer drug development. The inhibition of BCL-2 and other anti-apoptotic proteins will help to restore the apoptotic pathway. Aiming to target much diverse BCL-2 family protein dependency in solid tumors, we have developed a potent BCL-2 and BCL-xL dual inhibitor APG-1252, which is currently in phase 1 trials for solid tumors in USA, China and Australia. Mechanistically, BCL-2 and BCL-xL protect cells from cell death by sequestering pro-death BH3-only proteins like BIM. Thus, BCL-2 and BCL-xL inhibitors are designed to disrupt the protein complexes and release pro-death proteins to trigger downstream apoptotic cascade. In this study, we intended to further understand the mechanism of action of APG-1252 in vivo and factors regulating its sensitivity. To this end, we conducted trials in SCID mice carrying patient-derived xenograft (PDX) tumors, and studied target engagement. Methods and Experiments: We first selected eleven gastric and esophageal cancer PDX models that bore high levels of BCL-xL (BCL-xLhigh) but normal levels of MCL-1 (MCL-1nor) based on the gene expression profiles and the consideration of that BCL-xL may play a major role in tumorigenesis in solid tumors and MCL-1 a resistant factor for the dual inhibitor. Mice bearing PDX tumors were intravenously treated with 100 mg/kg of APG-1252 twice per week for three weeks. At the end of treatments, tumor samples were collected. An advanced ELISA system (MSD, Meco Scale Discovery) and Western blotting were performed to detect BCL-2 and BCL-xL protein complex and individual proteins. Results: 1. The MSD assays revealed dominant BCL-xL:BIM complex signals in the solid tumor PDX samples in comparison with hematological malignancy cell lines, in which BCL-2:BIM was the major complex. 2. The dual inhibitor treatment disrupted BCL-xL:BIM complex, thus BCL-xL:BIM complex may serves as a pharmacodynamic (PD) marker for target engagement. 3. BCL-2 family protein profiles showed high BCL-xL and MCL-1 levels in the solid tumor PDXs. The treatment with APG-1252 led to a significant increase in BCL-2 protein levels, while there was no major change on pro-apoptotic BH3-only proteins BIM and PUMA. 4. The more efficacious responders to the treatment exhibited a more substantial decrease in BCL-xL:BIM complex signals and low MCL-1 protein levels, whereas a much less significant decrease in BCL-xL:BIM complex and higher MCL-1 levels were found in the poor or non-responders. Conclusion: Both BCL-xL protein and BCL-xL:BIM complex are dominant in the solid tumor PDX tissues. The decrease of BCL-xL:BIM complex indicates the on-target activity after the treatment with APG-1252. Thus, the complex may be used to monitor the target engagement and, potentially, a predictive biomarker for APG-1252 efficacy. These assays may be applied to patient samples to guide the clinical trials. Citation Format: Jing Deng, Xiaojing Huang, Guoqin Zhai, Tingting Mao, Yan Yin, Douglas D. Fang, Dajun Yang, Yifan Zhai. Predicting drug sensitivity to a novel BCL-2 and BCL-xL dual inhibitor in solid tumor PDX trials [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 2503.

Abstract 1923: KRAS-mutant (mt) colorectal cancer (CRC) organoid models generated from patient-derived xenografts (PDX) show response to combination of trametinib (Tm), neratinib (N), and trastuzumab (Tz)

Abstract Background: KRAS-mt CRC has constitutively activated RAF-MEK-ERK pathways and resistance to anti-EGFR therapies. In pre-clinical models, we found that cells with inflammatory subtype tumors were sensitive to the MEK-inhibitor, MEK162, plus N, regardless of KRAS mutations. However, both KRAS-mt and -wt stem-like CRC cells were resistant to this combination. In C-07 and C-08 clinical studies, patients (pts) with stem-like subtype tumors were resistant to chemotherapy and had a very poor prognosis. We hypothesized that dual-HER2 targeting may provide more robust ERBB inhibition than N alone. We therefore tested Tz combined with Tm+N using PDX organoids (PDXOs) derived from KRAS-mt CRC tumors. PDOs have emerged as powerful preclinical models to predict clinical outcomes. The goal of this study was to identify more efficacious drug combinations for KRAS-mt tumors and stem-like subtype tumors using PDXOs. Methods: PDXO’s were generated using recently described methods from Hans Clevers' laboratory. Three KRAS-mt PDX tissues from Champions Oncology were used to generate four PDXOs for drug testing. The CTG-0406 PDX model was used to develop PDXOs from two different mice. Results: All three KRAS-mt PDXOs: CTG-0406 (inflammatory), CTG-1170 (inflammatory), and CTG-0079 (stem-like), were resistant to Tm, N, and Tz, as single agents. N+Tz was effective in inhibiting cell viability in all four models; however, more significant viability inhibition occurred when Tm was added to N+Tz in three of the PDXOs including the stem-like model. In all PDXOs, Tm+N+Tz inhibited 67-76% viability. Conclusion: We demonstrate that KRAS-mt PDXOs were inhibited to a greater extent with Tm+N+Tz compared to any of these drugs alone. PDXOs provide a rapid and cost-effective preclinical platform for screening of unique drug combinations for sensitivity or resistance. Table 1.% PDXO cell viabilityDrugsCTG-0406 (1) (KRAS mt) InflammatoryCTG-0406 (2) (KRAS mt) InflammatoryCTG-1170 (KRAS mt) InflammatoryCTG-0079 (KRAS mt) Stem-LikeTm (10nM)96%116%95%94%N (125nM)109%97%96%92%Tz (20 µg/ml)92%119%97%93%Tm (10nM) + N (125nM)64%59%31%56%Tm (10nM) + Tz (20 µg/ml)74%75%41%70%N (125nM) + Tz (20 µg/ml)34%70%23%43%Tm (10nM) + N (125nM)+ Tz (20 µg/ml)26%33%24%31% Support: NSABP Foundation, Inc. Citation Format: Rekha Pal, Ashok Srinivasan, Peter C. Lucas, Carmen J. Allegra, Angela M. Davies, Alshad S. Lalani, Samuel A. Jacobs, Katherine L. Pogue-Geile. KRAS-mutant (mt) colorectal cancer (CRC) organoid models generated from patient-derived xenografts (PDX) show response to combination of trametinib (Tm), neratinib (N), and trastuzumab (Tz) [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 1923.

SAT-330 Preclinical Screen of Patient CRPC Tumors Shows Sensitivity to Bromodomain Inhibitors Includes AR-null and Neuroendocrine Disease

Castration-resistant prostate cancer (CRPC) remains a lethal disease despite the introduction of second generation androgen receptor (AR) directed therapies. Resistance to AR directed therapies, such as abiraterone and enzalutamide, emerges through diverse alterations in the AR pathway or transition to aggressive AR-null phenotypes. Although bromodomain and extra-terminal (BET) inhibitor therapies are in current clinical trials for CRPC, understanding of the full spectrum of patient responders is limited by a lack of models for preclincial testing. Thus, the goal of this study was to examine the efficacy of BET inhibitors in a panel of patient-derived models of CRPC, including AR-null and neuroendocrine differentiated tumors that are inherently resistant to AR-directed therapy. To rapidly screen the sensitivity of tumors to BET inhibitors, we used ex vivo cultures of patient-derived xenograft tissue from the MURAL (Melbourne Urological Research Alliance) cohort of patient tumors and treated them with JQ1. Changes in proliferation and apoptosis were measured using histology for Ki67 and cleaved caspase 3. Consistent with previous studies, our results showed that all AR-positive tumors were sensitive to JQ1, with decreases in proliferation and increases in apoptosis. Importantly, this includes tumors from patients who failed abiraterone and/or enzalutamide and have diverse alterations of the AR, including mutations, genomic amplifications and expression of truncated AR variants. Surprisingly, 50% of AR null tumors were just as sensitive to BET inhibitors and AR-positive tumors. This includes tumors with large cell and small cell neuroendocrine histopathology, which are inherently resistant to AR-directed therapies. We also assessed intra-patient heterogeneity by examining tumors from different metastases of the same patient. We observed consistent responses in one patient, but divergent responses in another, likely due to differences in AR status. Collectively, our data show that BET inhibitors are promising treatments for tumors with diverse mechanisms of resistance to AR-directed therapies; however, the responses of AR-null tumors are more heterogeneous. Therefore, further studies to identify biomarkers of responsiveness to BET inhibitors in AR-null and neuroendocrine tumors are warranted.

Deciphering the complex role of thrombospondin-1 in glioblastoma development

We undertook a systematic study focused on the matricellular protein Thrombospondin-1 (THBS1) to uncover molecular mechanisms underlying the role of THBS1 in glioblastoma (GBM) development. THBS1 was found to be increased with glioma grades. Mechanistically, we show that the TGFβ canonical pathway transcriptionally regulates THBS1, through SMAD3 binding to the THBS1 gene promoter. THBS1 silencing inhibits tumour cell invasion and growth, alone and in combination with anti-angiogenic therapy. Specific inhibition of the THBS1/CD47 interaction using an antagonist peptide decreases cell invasion. This is confirmed by CD47 knock-down experiments. RNA sequencing of patient-derived xenograft tissue from laser capture micro-dissected peripheral and central tumour areas demonstrates that THBS1 is one of the gene with the highest connectivity at the tumour borders. All in all, these data show that TGFβ1 induces THBS1 expression via Smad3 which contributes to the invasive behaviour during GBM expansion. Furthermore, tumour cell-bound CD47 is implicated in this process.

A Simple Three-dimensional Hydrogel Platform Enables Ex Vivo Cell Culture of Patient and PDX Tumors for Assaying Their Response to Clinically Relevant Therapies.

A cell culture platform that enables ex vivo tissue growth from patients or patient-derived xenograft (PDX) models and assesses sensitivity to approved therapies (e.g., temozolomide) in a clinically relevant time frame would be very useful in translational research and personalized medicine. Here, we present a novel three-dimensional (3D) ECM hydrogel system, VersaGel, for assaying ex vivo growth and therapeutic response with standard image microscopy. Specifically, multicellular spheroids deriving from either 5 patients with glioblastoma (GBM) or a renal cell carcinoma (RCC) PDX model were incorporated into VersaGel and treated with temozolomide and several other therapies, guided by the most recent advances in GBM treatment. RCC ex vivo tissue displayed invasive phenotypes in conditioned media. For the GBM patient tumor testing, all five clinical responses were predicted by the results of our 3D-temozolomide assay. In contrast, the MTT assay found no response to temozolomide regardless of the clinical outcome, and moreover, basement membrane extract failed to predict the 2 patient responders. Finally, 1 patient was tested with repurposed drugs currently being administered in GBM clinical trials. Interestingly, IC50s were lower than C max for crizotinib and chloroquine, but higher for sorafenib. In conclusion, a novel hydrogel platform, VersaGel, enables ex vivo tumor growth of patient and PDX tissue and offers insight into patient response to clinically relevant therapies. We propose a novel 3D hydrogel platform, VersaGel, to grow ex vivo tissue (patient and PDX) and assay therapeutic response using time-course image analysis.

A Novel Patient-Derived Xenograft Model of Secondary CD20- CNS Lymphoma for Mechanistic and Therapeutic Exploration

IntroductionThe majority of secondary central nervous system (CNS) lymphoma cases are derived from histologically high-grade systemic lymphomas such as diffuse-large B cell lymphoma (DLBCL) and Burkitt's lymphoma. These CNS lymphomas are associated with poor prognosis and a low likelihood of long-term survival, especially rituximab-ineffective CD20- lymphoma. To investigate the biological features and potential therapeutic targets of CNS-involved CD20- lymphoma, we established the first known patient-derived xenograft (PDX) model using cerebrospinal fluid (CSF).MethodsSix- to 8-week-old male NSG mice (Jackson Laboratory) were housed and monitored in our animal research facility. All experimental procedures and protocols were approved by the Institutional Animal Care and Use Committee of The University of Texas MD Anderson Cancer Center. A total of 0.08 million cells were isolated from 2 mL of CSF obtained from a CD20- CD79A- DLBCL patient with Bcl-2 and Myc rearrangements (double-hit). The isolated cells were directly injected into a human fetal bone chip implanted into NSG (NSG-hu) mice. After passage, the mice were administered several small molecules, and tumor burden was measured via tumor volume and human β2M levels over the course of treatment. Kaplan-Meier survival analysis was also performed. PDX tissue was collected for histological analysis, immunophenotyping, genotyping, and proteomics analysis.ResultsThe PDX tumors possessed the same histological, genetic and immunophenotypical features as the primary patient tumor cells. The PDXs had Bcl-2 and Myc rearrangements and were positive for CD10, CD19, CD22, and CD45 and were negative for CD5, CD20, and CD79A. After tumor passage, the mice were treated with vehicle, 50 mg/kg ibrutinib (Bruton's tyrosine kinase (BTK) inhibitor) oral gavage daily, 50 mg/kg ABT199 (Bcl-2 inhibitor) oral gavage daily, 1 mg/kg carfilzomib (proteasome inhibitor) i.v. twice per week, or 7.5 mg/kg IACS-010759 (oxidative phosphorylation (OXPHOS) inhibitor) oral gavage daily. As shown in Figure 1, the PDX tumor was resistant to the BTK inhibitor ibrutinib and moderately responsive to the proteasome inhibitor carfilzomib. Although the Bcl-2 gene was rearranged, the PDX model only had a partial response to the Bcl-2 inhibitor ABT-199 (ABT-199 vs control, p = 0.0004). In contrast, inhibition of the OXPHOS pathway with IACS-010759 significantly inhibited 96% tumor growth (IACS-010759 vs control, p < 0.00001) and prolonged mouse survival by approximately 25 days compared with the vehicle control and the other drugs (IACS-010759 vs control, ABT-199, or ibrutinib, p < 0.0001). Moreover, proteomics analysis demonstrated that IACS-010759 not only inhibited OXPHOS but also upregulated the pro-apoptotic protein Bim.ConclusionTo our knowledge, this is the first PDX model generated from the CSF of a CD20- double-hit DLBCL patient. With this PDX model, we determined that the OXPHOS inhibitor IACS-010759 is a promising therapeutic agent for this aggressive cancer. [Display omitted] DisclosuresWang:Acerta Pharma: Honoraria, Research Funding; Kite Pharma: Research Funding; Novartis: Research Funding; Pharmacyclics: Honoraria, Research Funding; Dava Oncology: Honoraria; Juno: Research Funding; MoreHealth: Consultancy; Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; AstraZeneca: Consultancy, Research Funding; Janssen: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding.

Medicine Cabinet Oncology: Disulfiram and Copper Gluconate Show Similar Efficacy in Pancreatic Cancer Cell Lines and Patient-Derived Xenograft Tissue

Medicine Cabinet Oncology: Disulfiram and Copper Gluconate Show Similar Efficacy in Pancreatic Cancer Cell Lines and Patient-Derived Xenograft Tissue

Heterogeneous cancer-associated fibroblast population potentiates neuroendocrine differentiation and castrate resistance in a CD105-dependent manner.

Heterogeneous prostatic carcinoma associated fibroblasts (CAF) contribute to tumor progression and resistance to androgen signaling deprivation therapy (ADT). CAF subjected to extended passaging, compared to low passage CAF, were found to lose tumor expansion potential and heterogeneity. Cell surface endoglin (CD105), known to be expressed on proliferative endothelia and mesenchymal stem cells, was diminished in high passage CAF. RNA-sequencing revealed SFRP1 to be distinctly expressed by tumor-inductive CAF, which was further demonstrated to occur in a CD105-dependent manner. Moreover, ADT resulted in further expansion of the CD105+ fibroblastic population and downstream SFRP1 in 3-dimensional cultures and patient derived xenograft tissues. In patients, CD105+ fibroblasts were found to circumscribe epithelia with neuroendocrine differentiation. CAF-derived SFRP1, driven by CD105 signaling, was necessary and sufficient to induce prostate cancer neuroendocrine differentiation in a paracrine manner. A partially humanized CD105 neutralizing antibody, TRC105, inhibited fibroblastic SFRP1 expression and epithelial neuroendocrine differentiation. In a novel synthetic lethality paradigm, we found that simultaneously targeting the epithelia and its microenvironment with ADT and TRC105, respectively, reduced castrate resistant tumor progression, in a model where either ADT or TRC105 alone had little effect.

Patient-derived Models of Abiraterone- and Enzalutamide-resistant Prostate Cancer Reveal Sensitivity to Ribosome-directed Therapy

BackgroundThe intractability of castration-resistant prostate cancer (CRPC) is exacerbated by tumour heterogeneity, including diverse alterations to the androgen receptor (AR) axis and AR-independent phenotypes. The availability of additional models encompassing this heterogeneity would facilitate the identification of more effective therapies for CRPC. ObjectiveTo discover therapeutic strategies by exploiting patient-derived models that exemplify the heterogeneity of CRPC. Design, setting, and participantsFour new patient-derived xenografts (PDXs) were established from independent metastases of two patients and characterised using integrative genomics. A panel of rationally selected drugs was tested using an innovative ex vivo PDX culture system. InterventionThe following drugs were evaluated: AR signalling inhibitors (enzalutamide and galeterone), a PARP inhibitor (talazoparib), a chemotherapeutic (cisplatin), a CDK4/6 inhibitor (ribociclib), bromodomain and extraterminal (BET) protein inhibitors (iBET151 and JQ1), and inhibitors of ribosome biogenesis/function (RNA polymerase I inhibitor CX-5461 and pan-PIM kinase inhibitor CX-6258). Outcome measurements and statistical analysisDrug efficacy in ex vivo cultures of PDX tissues was evaluated using immunohistochemistry for Ki67 and cleaved caspase-3 levels. Candidate drugs were also tested for antitumour efficacy in vivo, with tumour volume being the primary endpoint. Two-tailed t tests were used to compare drug and control treatments. Results and limitationsIntegrative genomics revealed that the new PDXs exhibited heterogeneous mechanisms of resistance, including known and novel AR mutations, genomic structural rearrangements of the AR gene, and a neuroendocrine-like AR-null phenotype. Despite their heterogeneity, all models were sensitive to the combination of ribosome-targeting agents CX-5461 and CX-6258. ConclusionsThis study demonstrates that ribosome-targeting drugs may be effective against diverse CRPC subtypes including AR-null disease, and highlights the potential of contemporary patient-derived models to prioritise treatment strategies for clinical translation. Patient summaryDiverse types of therapy-resistant prostate cancers are sensitive to a new combination of drugs that inhibit protein synthesis pathways in cancer cells.

Abstract 2197: Surface marker and gene expression profiling of tumors from PDX models of small cell lung cancer with varying sensitivity to growth inhibition by the p70S6K/AKT1/3 inhibitor M2698

Abstract Small Cell Lung Cancer (SCLC) is characterized by rapid tumor growth and currently, there are few therapeutic options. As part of the development of M2698, a potent, selective inhibitor of p70S6K and AKT 1/3, the compound was tested in a preclinical screen of 45 patient derived xenograft (PDX) models of SCLC. In this screen, two mice were implanted subcutaneously with tumors for each model; one mouse was treated with vehicle while the other was treated with M2698 25 mg/kg QD po and tumors were measured 2x/week until the tumor in the vehicle-treated mouse reached ~1200 mm3. Tumor control (tumor stasis or regression) was seen in 12 (27%) of the models. From these data, a subset of models that were the most and the least sensitive to M2698 (n=7) were selected for implantation into a new set of mice. Tumors were profiled once they reached ~800 mm3 for tumor cell surface markers potentially correlating with sensitivity to M2698 seen in the previous study. PDX model systems are capable of recapitulating the intra-tumor heterogeneity (ITH) observed in the original patient tumor. The relationship between ITH and drug response/acquired drug resistance has been clearly demonstrated. Thus, a more detailed examination of this complexity in PDX tissue enables one to elucidate the underlying mechanisms in a reproducible, biologically relevant system. We have previously shown that cell surface marker profiles of PDX derived tumor tissue demonstrate high intra-model reproducibility whereas each model has a unique profile. In addition, some markers have a distinct heterogeneity. Gene expression profiling of tumor cell subpopulations identified by some of these specific markers suggest different biological roles for these subpopulations with relevance to their drug responses. Here we show surface marker and gene expression profiles of multiple PDX models for SCLC. We have evaluated 50+ markers commonly used to identify tumor initiating cells (e.g. CD44, CD90, CD133, CD166, CD184), EMT or aggressiveness (e.g. CD166, EphB2, CD324, CD325) or drug targets (e.g. CD184, EGFR, Her2) to establish extensive marker profiles. Our data reveal that surface marker profiles in SCLC PDX models are consistent among tumors from mice bearing the same model, show major differences from other PDX (e.g. breast) cancer types and heterogeneity exists for several markers of interest. By correlating gene expression and surface marker profiles in the current preclinical study, we aim to illuminate the interactions of major cell subpopulations within SCLC PDX tumor tissue and also investigate potential markers of sensitivity to growth inhibition by treatment with M2698. Overall, our workflow of deep phenotyping may provide tools for sample characterization, quality control and possibly study of cellular response to varying selective pressures such as drug challenges. Citation Format: Friedrich Hahn, Warren Porter, Eileen Snowden, Mitchell Ferguson, Frances Tong, William S. Dillmore, Anderson Clark, Hong Zhang, Rainer Blaesius. Surface marker and gene expression profiling of tumors from PDX models of small cell lung cancer with varying sensitivity to growth inhibition by the p70S6K/AKT1/3 inhibitor M2698 [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 2197.

Abstract 1121: Profiling patient-specific glioblastoma drug response in vitro using complex 3D microtumors

Abstract Glioblastoma (GBM) has a median survival of less than 2 years due to intra-tumoral heterogeneity, diffuse infiltrations of adjacent brain tissue, and a lack of effective therapies. Development of more efficacious therapies will require better GBM models for the testing and identification of novel agents. Towards this end, we have successfully developed a GBM 3D tissue model that can provide in vitro, patient-specific compound screening. Stable populations of glioma stem cells (GSC) from 24 of 41 patient samples have been successfully established and cultured long-term with minimal changes. To confirm stemness of the GSC population, we have successfully established a limiting-dilution series within SCID/Bg mice and characterized the resultant tumors. 4 of these lines have been used to establish patient-derived xenograft (PDX) models in mice. The original, primary patient tissue established GSC populations, and the resultant PDX tissues have been characterized by flow cytometry, IHC, RNA expression, NGS, and MGMT methylation status. With the goal of better modeling the patient tumor tissue in vitro, our GSC populations have also been used to establish complex microtumors within the KIYATEC 3DKUBE™ perfusion system, consisting of monoculture GSCs, GSCs co-cultured with human brain endothelial cells (HBECs), and GSCs co-cultured with HBECs and CD14+ peripheral blood mononuclear cells. Our monoculture microtumors consisting of only GSCs show a maintenance of GSC markers Nestin and Sox2 by both IHC and mRNA. Interestingly, when these cells are used to produce PDX, they up-regulate GFAP as a marker of differentiation that is not observed in the neurosphere or monoculture microtumor cultures. We have shown these 3D models to be viable for more than 1 month in perfusion and to be effective models for drug compound screening by dosing the microtumors on a weekly basis with temozolomide (TMZ). We have correlated TMZ response to MGMT methylation as reported both clinically and measured in vitro. Finally, In vitro drug response has been compared to both matched PDX in vivo drug response and the patient's clinical response to TMZ and MGMT methylation. Our data supports that this complex, 3D, patient-derived GBM model can be used to effectively screen, identify and characterize novel treatments of GBM. Citation Format: Ashley M. Smith, Melissa Millard, Lillia Holmes, Michael T. Lewis, Lacey E. Dobrolecki, Charles Kanos, Stephen Gardner, Philip Hodge, Fred Nelson, Michael Lynn, Jeff Edenfield, Christopher Corless, David Schammel, Howland E. Crosswell, Teresa M. DesRochers. Profiling patient-specific glioblastoma drug response in vitro using complex 3D microtumors [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 1121.

Comparison of the Genetic Alterations between Primary Colorectal Cancers and Their Corresponding Patient-Derived Xenograft Tissues

Patient-derived xenograft (PDX) models are useful tools for tumor biology research and testing the efficacy of candidate anticancer drugs targeting the druggable mutations identified in tumor tissue. However, it is still unknown how much of the genetic alterations identified in primary tumors are consistently detected in tumor tissues in the PDX model. In this study, we analyzed the genetic alterations of three primary colorectal cancers (CRCs) and matched xenograft tissues in PDX models using a next-generation sequencing cancer panel. Of the 17 somatic mutations identified from the three CRCs, 14 (82.4%) were consistently identified in both primary and xenograft tumors. The other three mutations identified in the primary tumor were not detected in the xenograft tumor tissue. There was no newly identified mutation in the xenograft tumor tissues. In addition to the somatic mutations, the copy number alteration profiles were also largely consistent between the primary tumor and xenograft tissue. All of these data suggest that the PDX tumor model preserves the majority of the key mutations detected in the primary tumor site. This study provides evidence that the PDX model is useful for testing targeted therapies in the clinical field and research on precision medicine.

The identification of a novel antibody for CD133 using human antibody phage display.

The transmembrane glycoprotein CD133 is believed to be a marker of adult prostate stem cells and cancer stem/initiating cells. Investigating the role of CD133 in the normal biology of the prostate and in cancer is complicated by the lack of a sensitive and accurate antibody for its detection. Here, we describe the characterization of a unique antibody identified using human antibody phage display that can recognize CD133 in both formalin-fixed tissues and cell lines. A human single-chain variable fragment (scFv) antibody phage display library possessing a diversity of 8 × 109 was screened against fully glycosylated recombinant CD133. A counter screen was performed against deglycosylated CD133 to select for clones that preferentially recognized a glycosylation-independent epitope. The lead scFv was analyzed by flow cytometry and cloned into a rabbit immunoglobulin scaffold for immunohistochemistry (IHC). The antibody designated HA10 was found to bind a glycosylation-independent epitope on the peptide backbone of CD133 with high affinity. As a reagent for flow cytometry, HA10 detected CD133 more accurately than a commonly used commercially available antibody. IHC analysis with HA10 documented the staining of basal cells and luminal cells in healthy prostate sections. Weak staining of luminal cells was observed in adenocarcinoma sections at a very low frequency. Examination of a LuCaP patient-derived xenograft tissue microarray found that only three of the LuCaP models were positive for CD133. The three CD133pos LuCaP models all originated from non-AR driven metastatic prostate cancer with neuroendocrine differentiation. Subsequent interrogation of liver biopsies from a patient who failed second-generation anti-androgen therapy found high levels of CD133 staining. The original transurethral resection of the prostate from that patient was, however, absent of CD133. We have developed a novel antibody that was able to detect CD133 by both IHC and flow cytometry. Using HA10 as an IHC reagent, we found that CD133 is a marker for a very rare cell type in both healthy prostate and adenocarcinoma sections. Our preliminary investigation also suggests that there may be an association between CD133 and non-AR driven prostate cancer with neuroendocrine differentiation.

Abstract B36: Novel immunodeficient rat models capable of supporting the growth of human tumor xenografts

Abstract Mouse models of human cancer have paved the way for studying cancer biology and genomics and their effects on cancer growth kinetics, propensity for metastasis, and treatment response. In addition, human cancer xenografts provide the opportunity to study cancer cell interactions with host stroma and tumor morphology. A plethora of genetically immunodeficient mouse models exist with different immune phenotypes, resulting in significant variability in tumor take rates and growth kinetics for a wide range of human cancer cell lines and patient-derived xenografts (PDX). Inconsistent or poor growth in these immunodeficient models have made downstream analysis and drug efficacy testing difficult. As a result, a significant number of mice are needed for drug efficacy screening to achieve a cohort of animals with tumors of similar size with similar tumor growth kinetics for treatment. It is possible that these cell lines might grow more consistently in a different immunodeficient model, such as an immunodeficient rat. Until recently, the only immunodeficient rat that existed was the nude (NIH-Foxn1rnu; RNU) rat. This rat lacks T cells, but maintains a normal repertoire of all other immune cells, including B and NK cells. As such, there are a limited number of human cancer cell lines that can survive in the nude rat. We have created a genetically modified rat with a functional mutation of the Rag2 gene (Sprague Dawley – Rag2 null; SDR), resulting in a loss of mature B and T cells. In addition, we have created a Rag2/Il2rg double knockout rat (Sprague Dawley – Rag2; Il2rg null; SRG) that lacks mature B cells, T cells, and has fewer NK cells than wild-type Sprague Dawley rats. We have shown that the SDR rat is permissive for solid tumor growth of the human acute lymphocytic leukemia REH cell line, human glioblastoma U87MG cell line, human non-small cell lung cancer H358 cell line, and cell lines derived from ovarian and endometrial PDX samples. In some cases, tumor growth kinetics are superior in the SDR rat compared with immunodeficient mouse models. We have demonstrated that the human prostate cancer cell line, VCaP, which has poor engraftment efficiency and growth kinetics in the mouse, grows well in the SRG rat. The SRG rat is currently being validated for growth kinetics of other human cancer cell lines and PDX tissues. In addition, we are developing several disseminated human leukemia models in the SRG rat and creating immune-humanized mice and rats to be used in conjunction with human tumor xenografts for immunotherapy efficacy studies. Citation Format: Fallon K. Noto, Kamesh Ravi, Angela Arey, Christopher McClain, Wei Zhang, Goutham Narla, Jack Crawford, Tseten Yeshi. Novel immunodeficient rat models capable of supporting the growth of human tumor xenografts [abstract]. In: Proceedings of the AACR Special Conference: Advances in Modeling Cancer in Mice: Technology, Biology, and Beyond; 2017 Sep 24-27; Orlando, Florida. Philadelphia (PA): AACR; Cancer Res 2018;78(10 Suppl):Abstract nr B36.

Murine Endogenous Retroviruses Are Detectable in Patient-Derived Xenografts but Not in Patient-Individual Cell Lines of Human Colorectal Cancer.

Endogenous retroviruses are remnants of retroviral infections. In contrast to their human counterparts, murine endogenous retroviruses (mERV) still can synthesize infectious particles and retrotranspose. Xenotransplanted human cells have occasionally been described to be mERV infected. With genetic engineered mice and patient-derived xenografts (PDXs) on the rise as eminent research tools, we here systematically investigated, if different tumor models harbor mERV infections. Relevant mERV candidates were first preselected by next generation sequencing (NGS) analysis of spontaneous lymphomas triggered by colorectal cancer (CRC) PDX tissue. Two primer systems were designed for each of these candidates (AblMLV, EcoMLV, EndoPP, MLV, and preXMRV) and implemented in an quantitative real-time (RT-qPCR) screen using murine tissues (n = 11), PDX-tissues (n = 22), PDX-derived cell lines (n = 13), and patient-derived tumor cell lines (n = 14). The expression levels of mERV varied largely both in the PDX samples and in the mouse tissues. No mERV signal was, however, obtained from cDNA or genomic DNA of CRC cell lines. Expression of EcoMLV was higher in PDX than in murine tissues; for EndoPP it was the opposite. These two were thus further investigated in 40 additional PDX. In addition, four patient-derived cell lines free of any mERV expression were subcutaneously injected into immunodeficient mice. Outgrowing cell-derived xenografts barely expressed EndoPP. In contrast, the expression of EcoMLV was even higher than in surrounding mouse tissues. This expression gradually vanished within few passages of re-cultivated cells. In summary, these results strongly imply that: (i) PDX and murine tissues in general are likely to be contaminated by mERV, (ii) mERV are expressed transiently and at low level in fresh PDX-derived cell cultures, and (iii) mERV integration into the genome of human cells is unlikely or at least a very rare event. Thus, mERVs are stowaways present in murine cells, in PDX tissues and early thereof-derived cell cultures. We conclude that further analysis is needed concerning their impact on results obtained from studies performed with PDX but also with murine tumor models.