Cancer Patient-derived Xenograft Research Articles

Abstract 651: Combination of the androgen receptor inhibitor darolutamide and the PI3K inhibitor copanlisib leads to improved anti-tumor efficacy and apoptosis in prostate cancer models

Abstract Prostate cancer is mainly driven by androgen receptor (AR) signaling so that drugs directly inhibiting this pathway such as GnRH ligands and AR inhibitors are now established as mainstay treatments. Unfortunately, treatment resistance often develops and combination therapies may represent potential strategies to prolong disease remission. The PI3K/Akt/mTOR signaling pathway plays an essential role in prostate cancer and is upregulated in about 30-50% of patients. Loss of the PI3K/Akt/mTOR pathway inhibitor PTEN is observed in up to 40% of late-stage prostate cancer patients. Compounds inhibiting PI3K or Akt have shown promising efficacy in preclinical prostate cancer models, and advanced clinical trials with these inhibitors are currently ongoing in metastatic castration-resistant prostate cancer, often in combination with AR inhibitors. Here we examined the impact of combining the AR inhibitor darolutamide with the pan class I PI3K inhibitor copanlisib on preclinical models in vitro and in vivo. We found that treating different androgen-dependent prostate cancer cell lines with darolutamide and copanlisib had a very strong synergistic inhibitory effect. Induction of apoptosis was observed in the PARP cleavage and caspase 3/7 activation assays after treatment with copanlisib and more so after treatment with copanlisib plus darolutamide. Expression analysis showed that darolutamide strongly down-regulated AR target genes such as KLK3, FKBP5 and TMPRSS2, but copanlisib had no additional impact. However, when examining genes involved in the apoptotic pathway we found that the down-regulation of pro-apoptotic genes that follows androgen treatment was strongly reversed by the combined darolutamide and copanlisib treatment. Finally, in vivo efficacy studies showed that the LuCaP 96 prostate cancer patient-derived (PDX) xenograft, which has a non-functional PTEN gene, strongly responded to darolutamide, but additional treatment with copanlisib did not increase anti-tumor efficacy. In the LuCaP 35 PDX model, which expresses PTEN at low level, darolutamide or copanlisib treatment alone showed only limited efficacy, while the combination exhibited tumor inhibitory effects. In conclusion, we found that combining darolutamide with copanlisib strongly inhibited the proliferation of prostate cancer models in vitro. This was linked to strong induction of cell apoptosis and prevention of pro-apoptotic gene down-regulation. Superior efficacy was also observed in the LuCaP 35 PDX xenograft model in vivo. Additional data are needed to explore the underlying molecular mechanisms. Together these results further support the exploration of the impact of combining AR and PI3K signaling inhibitors in late-stage prostate cancer patients. Citation Format: Tatsuo Sugawara, Holly NGuyen, Eva Corey, Ekaterina Nevedomskaya, Oliver Politz, Dominik Mumberg, Bernard Haendler. Combination of the androgen receptor inhibitor darolutamide and the PI3K inhibitor copanlisib leads to improved anti-tumor efficacy and apoptosis in prostate cancer models [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 651.

Abstract 1777: Androgen receptor (AR) agonists inhibit AR- and estrogen-receptor-positive breast cancer

Abstract Introduction: Breast cancer is the most common cancer diagnosed in women. About 13% of women will develop invasive breast carcinoma in their lifetime and it is estimated that about 280,000 new cases will be diagnosed in 2021. Seventy-percent of diagnosed breast cancers are estrogen receptor (ER)-positive and about 90% of the ER-positive breast cancers are androgen receptor (AR)-positive. We have previously shown that AR agonists inhibit ER-positive breast cancer growth by sequestering the pioneer-transcription factor FOXA1 from ER-cistrome. Tissue-selective AR modulators (SARMs) such as enobosarm are in clinical trials to treat ER-positive breast cancers. Since resistance is a major impediment to sutained treatment, it is important to identify new therapeutic strategies and their mechanisms of resistance. Here, we evaluated the possible mechanisms of resistance to AR agonists in ER-positive breast cancer. Description: ER-positive breast cancer cell line xenografts and patient-derived xenografts (PDX) were used as experimental models. Once tumors grew to ~100-300 mm3, mice were randomized and treated orally with vehicle or enobosarm (30 mg/kg/day). Animals were sacrificed after 4 weeks when the tumors responded to treatment and after 10 weeks once resistance developed to treatment. Tumor volumes were measured twice weekly and tumors were collected at sacrifice for further analyses. RNA-seq and ChIP-seq were performed on tumor tissues. Summary: AR agonists, enobosarm and dihydrotestosterone (DHT), inhibited the proliferation of ER-positive breast cancer cells ZR-75-1 and T47D. Transcriptome analysis revealed that AR agonists activate AR in ER-positive breast cancer cells and inhibit ER-target gene signature. AR agonists inhibited the growth of breast cancer cell line xenograft T47D and breast cancer PDXs HCI-7 that express wildtype ER, and HCI-13 and WHIM-23, two models that express Y537S mutant ER. Treatment of WHIM-23 and T47D xenograft for over ten weeks resulted in resistance and regrowth. AR agonist-sensitive and -resistant tumors were analyzed using RNA-seq, ChIP-seq, and ATAC-seq to understand the mechanism of resistance development. The results indicate that the prolonged activation of AR will result in resistance. These mechanisms of resistance can be utilized as possible therapeutic targets. Conclusion: These results suggest that the AR is a promising therapeutic target in the treatment of ER-positive breast cancer. However, resistance development is possible with AR agonists and the mechanism provides suggestions for future combination therapies. Disclosure: This work was supported by an NCI grant (CA229164 and CA229164S1 to RN) and by a DOD grant (W81XWH2110055 to RN). Citation Format: Sarah Asemota, Suriyan Ponnusamy, Thirumagal Thiyagarajan, Ramesh Narayanan. Androgen receptor (AR) agonists inhibit AR- and estrogen-receptor-positive breast cancer [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 1777.

Abstract 3099: Biological evaluation of a novel AKR1C3 inhibitor in patient-derived prostate cancer cell line and xenograft models

Abstract INTRODUCTION AND OBJECTIVES: Prostate cancer is a highly heterogeneous cancer type with distinct genomic and phenotypic characteristics that drive tumorigenesis and the differential response to drug therapies. A limit number of prostate cancer cell lines and patient-derived xenograft (PDX) models hinders research to improve disease outcome. Some currently available PDX models were derived from the primary tumor samples are insufficient to recapitulate the clinical response at more advanced stages. In this study, we developed patient-derived models from patients with advanced disease and evaluated a novel AKR1C3 inhibitor in these models. METHODS: Samples received from our Pathology Biorepository Shared Resource were divided into four groups and subjected to pathological staining, RNA extraction, xenografting in NSG mice via renal capsule and subcutaneous implantation in SCID mice and conditional reprogramed cultures (CRCs) or organoid culturing. The AKR1C3 inhibitor PB was modified from celecoxib. Androgen receptor (AR), AR-V7 and AKR1C3 expression were determined by western blot. The effects of the AKR1C3 inhibitor on enzalutamide sensitivity were characterized by growth assay and colony formation assay. RESULTS: Eight PDX models have been developed from prostate cancer patients with high Gleason score and/or at the castration-resistant stages. Among the PDX models, one spontaneous indefinite cell line PS1172 was established. Early passage CRCs showed the epithelial morphology with AR positive expression. Through serially passaging PS1172 PDX with castration in SCID mice, the castration resistant cell line 1172CR was re-cultured from castration-resistant PS1172 PDX tumors. 1172CR cells were resistant to enzalutamide treatment and expressed high level of AKR1C3 and AR-V7. A novel AKR1C3 inhibitor (PB) which displayed superior potential to inhibit AKR1C3 activity and suppress enzalutamide resistant prostate cancer cell growth was tested in these models. At the same dose, PB significantly suppressed 1172CR cell growth and colony formation compared to indomethacin and enzalutamide. PB also significantly suppressed AR/AR-V7 protein expression compared to indomethacin in 1172CR cells. CONCLUSION: PS1172 and castration-resistant 1172CR cells are novel models with significant characteristics such as AR-V7 and AKR1C3. These novel prostate cancer models are ideal for small molecule testing and resistant mechanism investigating. Citation Format: Joy C. Yang, Shu Ning, Hans Adomat, Martin Gleave, Allen Gao, Christopher P. Evans, Chengfei Liu. Biological evaluation of a novel AKR1C3 inhibitor in patient-derived prostate cancer cell line and xenograft models [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 3099.

Abstract 6016: Patient-derived tumor organoid cultures identify a chemo-resistant sub-population in esophageal squamous cell carcinoma

Abstract Background: Patient-derived xenograft has been a gold standard for cancer modeling. However, low establishment efficiency and high maintenance cost have hindered the development in the field. In recent years, patient-derived tumor organoid culture has been developed as an efficient and economic model to mimic and recapitulate the original tumor tissue, in contrast to conventional cancer cell lines and patient-derived xenografts (Drost & Clevers, 2018). Due to the limited availability of current esophageal squamous cell carcinoma (ESCC) models, we aimed to establish a panel of ESCC patient-derived organoid (PDO) long-term cultures for ESCC study. Results: In the current project, we have established and biobanked a panel of ESCC PDO cultures in vitro from Chinese ESCC treatment-naïve patients undergoing an endoscopic examination or upfront surgery, or patients undergoing neoadjuvant chemoradiotherapy (CRT) followed by surgery in Hong Kong. We also established and maintained PDO-derived xenograft (PDOX) in immunocompromised mice. Immunohistochemical examination confirmed the squamous origin of the established PDOs and PDOXs. Genomic profiling showed consistent mutational landscapes of the established PDO, as compared to ESCC patient profiles. Transcriptomic profiling suggests organoid culture resembles patient tissue samples better than the cell line model in terms of global gene expression patterns. Notably, PDOs from CRT-treated patients show enhanced re-growth after recovery from cisplatin treatment, indicating a candidate cancer stem cell (CSC) sub-population enriched by neoadjuvant CRT. This sub-population can be further enriched in vitro by serial cisplatin treatment and recovery cycles. Conclusions: ESCC PDOs can be efficiently established from highly limited patient tissue samples and maintained long-term in vitro and can be feasibly integrated with mouse model. It preserves the original cell hierarchy and indicates existence of candidate CSC sub-population, which may be targeted therapeutically in future. Reference: Drost J, Clevers H. Organoids in cancer research. Nat Rev Cancer. 2018 Jul; 18(7). Acknowledgment: We acknowledge the Research Grants Council (Theme-based Research Scheme T12-701/17-R to MLL) and the Food and Health Bureau (Health and Medical Research Fund 06171566 to VZY) of Hong Kong Special Administrative Region for funding supports. We acknowledge DSMZ (German Collection of Microorganisms and Cell Culture) for the KYSE cell lines. We acknowledge the University of Hong Kong Faculty of Medicine Core Facility for providing imaging facilities. Citation Format: Valen Z. Yu, Bryan C. Lung, Ian Y. Wong, Claudia L. Wong, Simon Y. Law, Maria L. Lung. Patient-derived tumor organoid cultures identify a chemo-resistant sub-population in esophageal squamous cell carcinoma [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 6016.

Abstract 3115: Establishment and characterization of patient-derived breast cancer models for cancer studies

Abstract BACKGROUND: Despite the progress and development of targeted agents and Immunotherapeutics, breast cancer is still the leading cause of death among women worldwide. At present, targeted therapy for breast cancers include endocrine blockage therapy for estrogen receptor positive (ER+) tumors and trastuzumab or lapatinib for HER2 positive (HER2+) tumors and CDK inhibitors for HER negative (HER-) tumors as well as PARP inhibitors for BRCA1/2 mutations. There is still a high unmet need for triple negative breast cancer (TNBC) and relapsed breast cancers with specific mutations. Breast cancer is a highly heterogeneous disease and reliable breast cancer models are needed to develop new clinical therapies and better reflect tumor biology. Patient-derived xenograft (PDX) models have been widely used for understanding tumor characteristics and represent human heterogeneity more faithfully and thus are suitable for mouse clinical studies for predicting drug efficacy in the clinic. Here we report the establishment and characterization of a panel of breast cancer (BC) PDX for preclinical research. METHODS: Tumor samples from patients’ breast tumor or metastastic lesion obtained via surgery, biopsy, or pleural effusions/ascites were engrafted subcutaneously in immunodeficient mice to establish PDX models. The models were classified and characterized by histopathology, immunohistochemistry (IHC), and breast cancer specific biomarkers such as estrogen receptor (ER), progesterone receptor (PR) and HER2. Genomic analysis of these models was performed by next generation sequencing (NGS). In these breast cancer models, the effects of estrogen on tumor growth and targeted endocrine therapy such as Fulvestrant were evaluated. The change of tumor volume over time was used to determine the growth rate and response to treatment in mice. RESULTS: A series of BC PDX models derived from primary patient breast tumor or metastatic lesions were successfully established and characterized. Key characteristics of parental tumors, such as histopathology, clinical markers, gene expression and copy number, as well as estrogen dependence was retained. The IHC data indicated various expression levels of ER, PR and HER2 across the panel, with TNBC being the dominant subtype. ESR1 mutation, which may play an important role in BC progression and endocrine therapy resistance, was also identified in an estrogen-independent model. In vivo, these models showed distinct tumor growth rates and different pharmacodynamic responses to standard of care agents. For example, ESR1 mutant model was responsive to Fulvestrant. CONCLUSIONS: The established BC PDX panel represent a range or ER/PR/HER2 expression status from primary and advanced cancer providing a series of preclinical models for breast cancer research and potential for mouse clinical trial enrollment. Citation Format: Leilei Chen, Xueying Yang, Likun Zhang, Xiaobo Chen, Henry Q.X. Li, Jingjing Wang. Establishment and characterization of patient-derived breast cancer models for cancer studies [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 3115.

Abstract 1756: Antitumor effect of XMT1660, a B7H4 targeting antibody drug conjugate, in an unselected panel of patient derived xenograft models of breast cancer

Abstract Introduction: XMT-1660 is a first-in-class Dolasynthen Antibody-Drug Conjugate (ADC) targeting B7-H4 and carrying a DolaLock payload with controlled bystander effect. B7-H4 is an immunoregulatory protein that is expressed in breast, ovarian and endometrial tumors. Expression of B7-H4 has been described in both tumor cells and in tumor associated macrophages, and B7-H4 shows infrequent co-expression with PD-L1, suggesting distinct immunoregulatory functions. XMT-1660 is built on the Dolasynthen platform which incorporates several attributes, including site-specific bioconjugation and precise control over drug-antibody ratio (DAR). The XMT-1660 ADC contains 6 DolaLock Auristatin F-hydroxypropyl amide (AF-HPA) anti-tubulin payloads per antibody (DAR-6). Previously, we demonstrated the cytotoxic effect of XMT-1660 in cell line xenograft and patient derived xenograft models of breast cancer known to have B7-H4 target expression. The purpose of this study was to evaluate the anti-tumor activity of XMT-1660 across an unselected panel of breast cancer patient derived xenograft (PDX) models, evaluate the level of B7-H4 expression across this panel of breast PDX models, and assess the correlation between B7-H4 expression and anti-tumor activity following a single dose of XMT-1660. Methods: A panel of 30 breast cancer patient-derived xenograft models, annotated by prior treatment history, and divided between TNBC and ER-positive subtypes, was implanted into athymic Nude-Foxn1n mice. When tumors reached an average volume of 150-300 mm3, animals (n=3) were treated with a single administration of either XMT-1660 0.15mg/kg (calculated by payload dose)/4.71 mg/kg (calculated by antibody dose) IV/QD X1 or saline vehicle. Tumor volumes were evaluated until the planned endpoint of mean tumor volume of control group of 1500 mm3 or day 28. At the endpoint, xenografts, or tumor beds (in the case of no palpable mass) were collected as formalin fixed paraffin embedded material. The xenograft material was evaluated for B7-H4 expression by protein and RNA methods and expression was compared to the anti-tumor activity. Results: In this panel of breast patient derived xenograft models, a range of anti-tumor activities has been observed following administration of a single dose of XMT-1660. The relationship between XMT-1660 anti-tumor activity and B7-H4 expression was demonstrated. Conclusion: XMT-1660 elicits a range of anti-tumor activity across a series of primary breast cancer xenograft models. XMT-1660 is currently in IND-enabling studies and is expected to enter a Phase I dose escalation clinical study in 2022. The efficacy/expression relationship of B7-H4 will be further evaluated in an upcoming clinical study with a goal to identify patients most likely to respond to XMT-1660. Citation Format: Scott D. Collins, Pamela S. Shaw, Shawn P. Fessler, Jason Wang, Rebecca Mosher. Antitumor effect of XMT1660, a B7H4 targeting antibody drug conjugate, in an unselected panel of patient derived xenograft models of breast cancer [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 1756.

Abstract 6161: HER3 expression is associated with cold tumor immune microenvironment and the regulation of its signal in humanized breast cancer patient derived xenograft models converts to hot tumor

Abstract Background: HER3 signal pathways in cancer cells are thought to be involved in up-regulation of PD-L1 expression, resulting in “cold” tumor immune microenvironment (TIME). However, little is known on the association of HER3 expression and immune related molecular network in breast cancer and the functional role of HER3 signals in antitumor immune responses. Methods: The dataset of the TCGA study was analyzed to determine the correlation of HER3 expression and antitumor immune molecules. Two breast cancer patient-derived xenograft (PDX) models were treated with patritumab (anti-HER3 antibody), polyclonal activated autologous T cells (PATCs) or a combination of patritumab and PATCs (P-PATCs) in order to test immunological modulation by the regulation of HER3 signals. RT-PCR, immunohistochemical staining and CyTOF were performed in tumors or liver tissue for CD3, CD4, CD8a, CD137, CD134/OX40, CD80, IFN-γ and PD-L1, LAG3, TIM3, FOXP3. In addition, HER3-knockdown human breast cancer cell lines and PBMC were co-cultured and comprehensive cytokine expression in supernatant was analyzed by bioplex. Results: In TCGA dataset analysis, HER3 expression was inversely corelated with expression of CD4, CD8A, IFN-γ, granzyme A/B. In humanized PDX models, although treatment of PATCs or patritumab alone showed limited anti-tumor effects, P-PATCs treatment showed a significantly greater anti-tumor response. Interestingly, the proportion of CD137 expressing T cell infiltration and mRNA expression of CD137 was significantly higher in P-PATCs group compared to patritumab or the PATCs group, while the proportion of CD4/8 T cell infiltration was also higher in the P-PATCs group. CyTOF revealed that the proportion of CD137 and CD134/OX40 expressing cells was higher in P-PATCs group compared to the PATCs group. In co-culture of PBMC with HER3 knockdown cancer cells, the secretion level of immune stimulatory factors (IL2, IFN-g, TNF-a) increased and immune inhibitors (IL6, IL10) decreases. Conclusions: These data suggest that HER3 expression in breast cancer associated with “cold” TIME and the regulation of its signal converts “cold” TIME to “hot”. Thus, inhibition of HER3 signal might be of interesting concept for both regulation of oncologic signal and immunological modulation in HER3-expressing breast cancer. Further HER3 related immune modulation exploration in breast cancer treatment strategy needs to be clarified in the future. Citation Format: Sunao Tanaka, Eiji Suzuki, Takeshi Kotake, Shinpei Kawaoka, Kosuke Kawaguchi, Tomomi Nishimura, Fengling Pu, Masakazu Toi. HER3 expression is associated with cold tumor immune microenvironment and the regulation of its signal in humanized breast cancer patient derived xenograft models converts to hot tumor [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 6161.

Abstract 1768: Anti-tumor activity and biomarker analysis for datopotamab deruxtecan in breast cancer PDX models

Abstract Background: Trophoblast cell surface antigen (TROP2), is expressed on many tumors including breast cancer, thus TROP2 antibody drug conjugates are being pursued as a therapeutic strategy. Datopotamab deruxtecan (Dato-DXd) is composed of a humanized anti-TROP2 IgG1 monoclonal antibody attached to a highly potent topoisomerase I inhibitor payload (an exatecan derivative DXd) via a stable tetrapeptide-based cleavable linker. Dato-DXd has shown promise in the treatment of TROP2 expressing lung and breast cancers. However, the role of TROP2 as a predictor of therapeutic sensitivity is yet to be elucidated. We sought to determine efficacy of Dato-DXd in breast cancer patient-derived xenografts (PDXs) as well as explore biomarkers of response and combinations with PARP inhibitors. Methods: Membrane expression of TROP2 was assessed in 31 breast cancer PDXs and matching patient tumors by immunohistochemistry (IHC). Schlafen family member 11 (SLFN11) nuclear expression was also assessed by IHC. The antitumor efficacy of two doses (1 and 10 mg/kg, q3wk, IV) of Dato-DXd and the isotype control-DXd (IgG-DXd) were tested against 9 breast cancer PDXs derived from residual tumors after neoadjuvant chemotherapy. The PDXs represented a range of TROP2 expression levels, including 3 TROP2 negative/low PDXs. The antitumor activity of Dato-DXd in combination with PARP inhibition (olaparib) was assessed in 3 PDXs with intermediate Dato-DXd activity. Tumor volumes were measured twice weekly; antitumor activity was assessed by treatment-to-control ratio (T/C) and event-free survival (EFS-2). T/C ratio was calculated as (Vt,21/Vt,0)/(Vc,21/Vc,0), where t = treatment, c = control (no treatment or IgG-DXd), and V = tumor volume (mm3). Stable disease (SD) and response (R) were based on day 21 (-30- to 20% and <-30%, respectively). An event was defined as tumor doubling. Results: TROP2 H-score in PDXs correlated with TROP2 expression in matched patients (r= 0.7264, p<0.0001), however expression was lower in PDXs (p= 0.04). Dato-DXd caused R/SD in 2 of 9 models at 1 mg/kg, and in 4 of 9 models at 10 mg/kg. All models that regressed with Dato-DXd expressed TROP2. TROP2 expression was associated with higher antitumor activity compared to IgG-DXd based on T/C ratio (r= -0.7448, p= 0.0213) and EFS-2 (r= 0.9318, p= 0.0068) at 10 mg/kg but not at 1 mg/kg. Three models with low TROP2 expression had >50% SLFN11 positivity in the nucleus by IHC; 2 of 3 had doubling of EFS-2 both by Dato-DXd and IgG-DXd. The combination of Dato-DXd and olaparib had improved activity over single agents in 2 of 3 models. Additional comparative predictive and pharmacodynamic biomarker studies will be presented. Conclusion: Dato-DXd is a promising therapy for breast cancer patients including those resistant to standard chemotherapy. Additional biomarkers may better integrate DXd sensitivity into patient selection. Citation Format: Erkan Yuca, Kurt Evans, Argun Akcakanat, Gabriela Raso, Yasmeen Q. Rizvi, Fei Yang, Lauren Byers, Senthil Damodaran, Okajima Daisuke, Funda Meric-Bernstam. Anti-tumor activity and biomarker analysis for datopotamab deruxtecan in breast cancer PDX models [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 1768.

Abstract 1794: The combination of Talazoparib with anti-her2 drugs shows efficacy in drug resistant Her2+ and low Her2 PDX models

Abstract Background: Her2+ breast cancer accounts for 15% of breast cancer cases and is primarily treated with a combination of Her2-targeted agents and chemotherapy. While use of targeted agents has substantially improved survival of these patients, clinical drug resistance is prevalent. DNA repair defects are a feature of cancer that has been exploited by targeted agents such as inhibitors of poly-ADP ribose polymerase (PARP). PARP inhibitors have been shown to be effective in patients with germline mutations in other DNA repair factors such as BRCA1/2. There is limited data for the use of PARP inhibitors in HER2+ breast cancers, and in tumors without BRCA1/2 mutations. Using a collection of HER2+ and HER2-low expressing breast cancer patient-derived xenografts (PDXs) that have differential responses to the Her2-targeted agents Trastuzumab, Pertuzumab, and T-DM1, we tested the combination of these agents with Talazoparib (Talzenna™), a clinically approved Parp inhibitor. Methods: Orthotopically engrafted PDX mouse models resistant to either Trastuzumab or T-DM1 were randomized into treatment arms including Trastuzumab or T-DM1 in combination with Talazoparib, and the single agents alone. Mice are monitored for body weight and change in tumour volume every 2-3 days, then sacrificed at endpoint for survival analysis. We have also generated PDX-derived cells (PDCs) from one model with which to validate in vivo results and perform mechanistic studies. All models are being characterized for copy number changes, somatic mutations, and protein expression by immunohistochemistry. Results: The combination of Talazoparib and either Trastuzumab or T-DM1 was effective in 3 PDX models, including a Trastuzumab-resistant HER2+ PDX, a T-DM1-resistant HER2+ PDX and a T-DM1-resistant HER2-low expressing PDX. The combinations resulted in significant delays in tumour growth (p<0.05), including a transient tumour regression with the combination T-DM1+Talazoparib in 1 PDX. Median overall survival time was improved with the combination of Trastuzumab +Talazoparib (p<0.05), while the combinations of T-DM1+Talazoparib increased survival time although not significantly. Notably all models have maintained Her2 expression through treatment. Conclusion: Our results suggest that combining HER-targeted agents with Talazoparib may benefit patients with advanced HER2-therapy resistant HER2+ and low HER2+ breast cancers. Exploratory biomarker analysis and mechanistic studies using PDCs are currently underway in our laboratory. Citation Format: Kathryn Bozek, Marguerite Buchanan, Cathy Lan, Josiane Lafleur, Cedric Darini, Urszula Krzemien, Adriana Aguilar-Mahecha, Mark Basik. The combination of Talazoparib with anti-her2 drugs shows efficacy in drug resistant Her2+ and low Her2 PDX models [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 1794.

Abstract 3107: Identifying therapeutic vulnerabilities in HR-positive, HER2-negative advanced breast cancer patient-derived xenograft models refractory to CDK4/6 inhibition

Abstract Background: Clinically, the addition of CDK4/6 inhibitors (e.g. palbociclib) to endocrine therapy (ET) significantly delays progression of advanced hormone receptor-positive (HR+), HER2-negative (HER2-) breast cancer patients. However, most patients develop resistance and progress following long-term treatment. Thus, an ongoing clinical challenge has been the identification of biomarkers that predict response to CDK4/6 inhibitors. There is also an unmet need to identify actionable targets for patients who have progressed. Currently, there are no clinically useful biomarkers to predict response and/or resistance to anti-CDK4/6 therapy. Hence, the goal of our study was to identify the therapeutic vulnerabilities of CDK4/6 inhibitor resistant patient derived xenograft (PDX) models and identify key markers that longitudinally correlate with development of resistance and inform new treatment directions. Methods: We obtained 4 different PDX models, two from patients who progressed after 2-4 months on treatment with palbociclib plus ET (i.e. intrinsic resistance) and the other two who developed resistance over time (i.e. acquired resistance) with disease progression between 12-18 months while on palbociclib plus ET. To elucidate mechanisms of resistance, we performed genome-wide expression analysis via RNA-sequencing. Results: The molecular assessment of the PDX models revealed different transcriptomic signatures for intrinsic resistance compared to acquired resistance. We identified 3,386 upregulated and 2,829 downregulated significant differentially expressed genes (DEGs) in the acquired compared to intrinsic resistant PDX models. Further, gene set enrichment analysis (GSEA) revealed enrichment of distinct driver pathways in the acquired-compared to the intrinsic resistant models. Conclusions: Collectively, these results have identified potential unique targets for intrinsic and acquired resistance to guide treatment strategies. Thus, our ongoing studies are geared towards identifying and targeting distinct therapeutic vulnerabilities of intrinsic and acquired resistance to palbociclib. Citation Format: Nicole M. Kettner, Juliana Navarro-Yepes, Tuyen N. Bui, Xiayu Rao, Jing Wang, Funda Meric-Bernstam, Kelly K. Hunt, Debu Tripathy, Khandan Keyomarsi. Identifying therapeutic vulnerabilities in HR-positive, HER2-negative advanced breast cancer patient-derived xenograft models refractory to CDK4/6 inhibition [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 3107.

Abstract 1548: Examination of miRNA dysregulation in pancreatic cancer PDX models with parallel proteomic and transcriptomic investigations

Abstract Introduction: Pancreatic cancer is a high-fatality cancer. An estimated 57,600 new cases were diagnosed in the United States in 2020 with a corresponding 47,050 deaths. Despite improvements in the molecular understanding of pancreatic cancer, the prognosis for this disease has not improved in nearly 4 decades, largely because of the lack of early stage disease detection and limited efficiency of systemic therapies. MicroRNAs (miRNAs) are a class of non-coding RNAs approximately 22 nucleotides (nt) in length that regulate gene expression. in recent years, the dysregulation of miRNAs has been implicated in the development of various human cancers. Methods: Parallel analysis of global mRNA, miRNA and protein expression was conducted on pancreatic cancer patient-derived xenografts. The data from this analysis of these F1 tumours, was used to refine/identify miRNA targets of interest to generate 2 differential expression (DE) lists for miRNA; 1) Tumour vs Normal and 2) Tumour vs F1. Predictive software was used to identify targets and pathways of the DE priority miRNAs. These targets were further refined by integrating the targets with the global mRNA and proteomic profiling (unpublished data) performed on the same clinical samples. Alteration of cancer phenotype was investigated using overexpression and/or sponge vectors for each miRNA. Results: Analysis generated a list of priority miRNAs that had inverse interactions with predicted protein targets and were shown to be differentially expressed in our proteomic data. From 39 miRNAs investigated from the priority list, 7 miRNAs targeted 2 or more proteins across the 2 comparisons. These miRNAs - miRs-2467, -4742, -509, -615-3p, -4534, -222 and 2-06 when altered in expression MIA PaCa-2 and PANC-1 pancreatic cancer cells resulted in modulated cell proliferation, colony forming ability and anoikis resistance. Studies are ongoing on the changes to the chemotherapy and radiosensitivity of each cell line in response to changing these miRNA levels. Conclusions: The study identified miRNAs which displayed increased expression between normal and tumour and again increased between tumour and F1. From this list using predicative software and integration with additional global analysis of mRNA and Protein expression a priority list of 7 miRNAs were identified. The core concept in this study was to use differential expression of miRNAs as a tool to further unravel the biology of human pancreatic cancer, in the hope of uncovering new targets or vulnerabilities which might be exploited therapeutically or finding new bio markers which might contribute to better or earlier diagnosis. Citation Format: Taylor-Jade Allen. Examination of miRNA dysregulation in pancreatic cancer PDX models with parallel proteomic and transcriptomic investigations [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 1548.

Abstract 1791: Epigenetically upregulating TROP2 enhances therapeutic efficacy of TROP2 ADC sacitizumab govitecan

Abstract Purpose: TROP2 is overexpressed in many tumor types and is being actively pursued as a target. Sacituzumab govitecan (SG), a humanized anti-TROP2 antibody conjugated with SN-38, was approved for treatment of metastatic triple negative breast cancer, with the greatest efficacy in patients with medium or high TROP2 levels. We sought to enhance efficacy of TROP2-targeted therapies by pharmacological regulation of TROP2 expression. Methods: TROP2 levels were assessed by immunohistochemistry (IHC) in two sets of breast tumors: a set of surgical samples and a tissue microarray. TROP2 mRNA expression was assessed in surgical samples and breast cancer patient-derived xenografts (PDXs) with RNAseq and assessed in the TCGA. In cell lines, expression of TROP2, E-cadherin (E-cad), and Schlafen family member 11 (SLFN11) were assessed by immunoblotting and qPCR following drug treatment or cell line manipulation. Epithelial-mesenchymal transition was evaluated by cell migration, cell invasion, and anchorage-independent growth assays. Antitumor efficacy of drug combination was assessed by cell survival, cell colony formation, and apoptosis assays. Results: By IHC, TROP2 was expressed in only 40% of metaplastic breast cancers (MpBC), but nearly all non-MpBC tumors. TCGA database evaluation further showed higher TROP2 levels in non-MpBC tumors than metaplastic tumors. In breast cancer surgical specimens, breast cancer PDXs, and the TCGA, there was a strong correlation between TROP2 and E-cad expression. In vitro, we demonstrated that downregulating transcriptional factor zinc finger E-box binding homeobox 1 (ZEB1) led to mesenchymal-epithelial transition with upregulation of both E-cad and TROP2 expression in breast cancer cells, leading to increased sensitivity to SG treatment. Screening of epigenetic modulators identified DNA methyltransferase inhibitor decitabine as an enhancer of TROP2 and E-cad expression in PDX cell lines of metaplastic cancer origin and mesenchymal subtype breast cancer cell lines. Decitabine increased TROP2 expression by decreasing TROP2 promoter methylation. Decitabine was significantly synergistic with SG, and enhanced apoptosis. Similarly, overexpression of TROP2 (by plasmid) in cell lines enhanced activity of SG. Furthermore, decitabine increased expression of SLFN11, a putative biomarker of SN38 sensitivity, and was synergistic with SG in TROP2 expressing, SLFN11 low breast cancer cell lines. Conclusion: TROP2 is expressed in most breast cancers, but is expressed less frequently in MpBC, an aggressive subtype unresponsive to traditional therapies. Epigenetic modulator decitabine upregulates TROP2 and SLFN11 expression and enhances antitumor efficacy of SG. Combinatorial treatment of TROP2 ADCs with epigenetic modulators of TROP2 represent a novel therapeutic strategy for tumors with low TROP2 or SLFN11 expression. Citation Format: Ming Zhao, Timothy P. DiPeri, Gabriela Raso, Yasmeen Q. Rizvi, Xiaofeng Zheng, Kurt Evans, Argun Akcakanat, Fei Yang, Debu Tripathy, Ecaterina Ileana Dumbrava, Senthil Damodaran, Funda Meric-Bernstam. Epigenetically upregulating TROP2 enhances therapeutic efficacy of TROP2 ADC sacitizumab govitecan [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 1791.

Abstract 330: Targeting B7-H3 in prostate cancer: Preclinical proof of concept with MGC018, an investigational anti-B7-H3 antibody-drug conjugate

Abstract Introduction: Prostate cancer is the second most common cancer among men worldwide. In 2021, it is estimated that 248,530 men in the United States will be diagnosed with prostate cancer, and 34,130 will die from the disease. Although current treatments have success initially, development of resistance commonly leads to recurrence of an incurable castrate-resistant form of the disease. Thus, significant need for novel therapies to improve the outcome of castrate-resistant prostate cancer remains. B7-H3 (CD276), a member of the B7 family of immunomodulatory molecules, is overexpressed in primary and metastatic prostate cancer, and correlates with disease severity and poor clinical outcome. MGC018, a duocarmycin-based B7-H3 antibody-drug conjugate, is currently being evaluated in clinical studies. Here, MGC018 was explored preclinically to assess the potential for targeting B7-H3 in prostate cancer. Methods: Immunohistochemistry studies were performed to define the expression of B7-H3 in prostate cancer tissue microarrays (TMA). In vivo efficacy studies were conducted with human prostate cancer cell line-derived xenograft (CDX) models to explore the antitumor activity of MGC018 as a single agent and in combination with Poly (ADP-ribose) polymerase (PARP) and androgen receptor (AR) inhibitors. Based on the results in the CDX studies, in vivo efficacy studies were extended to a panel of metastatic prostate cancer patient-derived xenograft (PDX) models, which exhibit heterogenous expression of B7-H3 and more closely mimic the biological characteristics of patient tumors. Results: Staining of prostate tumor TMAs revealed high expression of B7-H3 in primary and metastatic prostate cancer. Of the prostate samples evaluated, 95% (38/40) of the tumor samples were positive for B7-H3 (H-score ≥ 20): 65% had H-scores greater than 200, while 20% and 10% had H-scores between 101-200 and 1-100, respectively. MGC018 demonstrated in vitro cytotoxicity toward B7-H3-positive human prostate cancer cell lines. The in vitro cytotoxicity translated to potent antitumor activity in vivo toward prostate cancer CDX models, and the antitumor activity of MGC018 was enhanced when combined with inhibitors of PARP or AR. In PDX models of metastatic prostate cancer, MGC018 was active as a single agent toward heterogeneous B7-H3-expressing tumors, and combining MGC018 with inhibitors of PARP or AR led to a greater response in some models. Conclusion: B7-H3 is frequently overexpressed in prostate cancer. MGC018 demonstrated potent antitumor activity in vivo toward CDX and PDX models of prostate cancer, and enhanced antitumor activity when combined with inhibitors of PARP or AR. These results support prostate cancer as an indication that may be responsive to ADC-based treatments directed toward B7-H3. MGC018 is being investigated in metastatic prostate cancer in a Phase 1/2 clinical study. Citation Format: Juniper A. Scribner, Francine Z. Chen, Anushka De Costa, Ying Li, Michael Chiechi, Thomas Son, Jeff Hooley, Jonathan Li, Scott Koenig, Chet Bohac, Ezio Bonvini, Paul A. Moore, Deryk Loo. Targeting B7-H3 in prostate cancer: Preclinical proof of concept with MGC018, an investigational anti-B7-H3 antibody-drug conjugate [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 330.

Abstract 695: Patient derived organoids model lung adenocarcinoma metastases

Abstract Introduction: While lung cancer patient-derived xenografts (PDX) models predict response to therapy, PDX tumors rarely metastasize thus offering little personalized assessment of the biology of metastases. We hypothesize that the development of an in vivo patient-derived organoid (PDO) model will improve our understanding of the underlying genomic and transcriptomic alterations that help drive the biology of metastases for specific lung adenocarcinoma (LUAD) histologic subtypes. Methods: Fresh LUAD specimens were processed and cultured using optimized organoid culture conditions. Established PDOs were characterized by histologic analyses, RNA-sequencing and whole genomic sequencing (WGS). Response to therapy of PDOs was evaluated by treatment with combined chemotherapies, targeted therapies, or co-culturing with patient-matched peripheral blood mononuclear cells (PMBC) in vitro and/or in vivo. The metastatic capabilities of PDOs were assessed in vivo using intracardiac injection and scoring for tumor outgrowth in distant organs using in vivo bioluminescence imaging system. Metastatic tumors from specific organs were isolated, PDOs reestablished, and those tumors underwent additional genomic and transcriptomic characterization. Phylogenetic analyses of tumor metastatic progression and putative metastasis markers were investigated by WGS and RNA-sequencing. Results: We successfully generated 14 PDOs from early-stage (I-II) primary LUAD specimens and 12 from human LUAD metastatic sites (brain, bone and pleural effusion). All PDOs recapitulate genetic alterations and transcriptomic signatures of their original tumors. Moreover, we found that individual PDOs exhibit distinct responses to different chemotherapy regimens and targeted therapies. Pre-treatment of PDOs with low doses of platinum-based doublet chemotherapy regimens activated patient-matched tumor specific T cells and enhanced PD-1 checkpoint inhibitor Nivolumab induced CD8+ T cell-mediated tumor killing. Importantly, utilizing our in vivo PDO metastasis model we successfully developed a platform to evaluate: 1) metastatic potential, 2) response to targeted therapies, 3) tumor evolution with respect of genomic instability, tumor phylogeny, and colonization, and 4) molecular signatures promoting tumor metastases. Conclusion: Our PDO system reliably models human LUAD in vitro and in vivo, by recapitulating histologic architectures, genomic alterations and transcriptomic signatures of the parent LUAD. Using LUAD PDOs we have successfully created multiple platforms to assess response to chemo- and targeted- therapies, develop an in vivo PDO metastasis model to assess tumor evolution, and use of patient matched PMBCs and PDOs to assess immune-priming strategies for individual patient tumors. Citation Format: Yuan Liu, David Jones. Patient derived organoids model lung adenocarcinoma metastases [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 695.

Abstract 1143: ATG-022, an antibody-drug conjugate targeting Claudin 18.2, demonstrated potent in vivo efficacy in gastric cancer patient-derived xenografts

Abstract Background: Antibody-drug conjugates (ADCs) have become promising antitumor agents in recent years. ADCs are comprised of a monoclonal antibody, targeting antigens expressed at higher levels on tumor cells than on normal cells, cross-linked to small molecule payloads with cytotoxic activity. Human Claudin 18.2 (CLDN18.2) is overexpressed in a large proportion of gastric and pancreatic cancers, with restricted normal tissue expression. The monoclonal antibody targeting CLDN18.2 (IMAB362) has demonstrated promising clinical benefit in combination with chemotherapies for gastric cancer patients. However, the same antibody shows suboptimal efficacy in patients with low CLDN18.2 expression. Here we report the potent in vivo efficacy of ATG-022, a CLDN18.2 ADC, in multiple gastric cancer patient-derived xenograft (PDX) models, including those with low CLDN18.2 expression. Methods: The binding affinity of ATG-022 with CLDN18.2 was detected by SPR and FACs analysis. The in vitro 50% inhibition concentration (IC50) of ATG-022 was determined in CLDN18.2 positive cell lines using CellTiter-Glo luminescent cell viability assay. In vivo efficacy of ATG-022 was evaluated in a series of gastric cancer PDX models with different expression level of CLDN18.2 The expression level of CLDN18.2 was determined using IHC staining. ATG-022 was dosed twice at 1 mg/kg, 3 mg/kg or 10 mg/kg every two weeks. Results: ATG-022 binds to CLDN18.2 protein with sub-nM affinity. It induced potent in vitro cytotoxicity in CHOK1 cells overexpressing CLDN18.2, with an IC50 of 5-7nM. Bystander killing by ATG-022 was observed. ATG-022 demonstrated potent in vivo antitumor efficacy in a gastric cancer PDX model with high CLDN18.2 expression. Intravenous dosing of 1 mg/kg, 3 mg/kg, and 10 mg/kg ATG-022 induced 35.41%, 71.56% tumor growth inhibition (%TGI) and tumor regression respectively. In a PDX model with low CLDN18.2 expression, 3 mg/kg ATG-022 or 10 mg/kg clinical benchmark CLDN18.2-ADC did not inhibit tumor growth, while 10 mg/kg ATG-022 induced tumor regression. Conclusions: ATG-022 demonstrated potent in vitro and in vivo antitumor effects, with in vivo efficacy observed in CLDN18.2-low expression PDX models, suggesting a promising therapeutic strategy for gastric cancer patients with a broad range of CLDN18.2 expression levels. Citation Format: Peng Chen, Yun Liu, Min Deng, Linjie Tian, Kevin Lynch, Bo Shan, Jay Mei, Bing Hou. ATG-022, an antibody-drug conjugate targeting Claudin 18.2, demonstrated potent in vivo efficacy in gastric cancer patient-derived xenografts [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 1143.

Abstract 1689: Heterogeneity of tumor-initiating cell related signaling revealed by lentiviral signaling reporters in breast cancer patient-derived xenografts

Abstract Classical markers of tumor-initiating cells (TIC), aldehyde dehydrogenase 1 (ALDH1) activity and the cell-surface markers, CD44pos/CD24low/neg, show insufficient cell type specificity to identify TIC definitively, nor do they allow robust enrichment by FACS. As a consequence, analyses of TIC function and gene expression to date are not definitive. Identification and characterization of the TIC subpopulation is critically important in breast cancer since these cells are thought to mediate treatment resistance, as well as local and distant recurrence. This is particularly critical for “triple negative” breast cancer (TNBC), since currently there are no targeted therapies available for this clinical subtype. Lentiviral signaling reporter vectors allow purification of enriched TIC subpopulations thereby facilitating single cell analyses to identify distinct sets of cells with TIC features. We used a collection of fluorescent lentiviral reporters that enable visualization and flow cytometric enrichment of cells undergoing STAT3-, GLI-, WNT-, or OCT4/SOX2-mediated transcriptional activity. Our collection of patient derived xenograft (PDX) models of TNBC were infected with the lentiviral reporters, re-transplanted, and screened for reporter activity upon regrowth. TIC function of reporter-expressing cells relative to negative cells was assayed by mammosphere formation and limiting dilution transplantation. Reporter expressing cells with enriched TIC function were then analyzed by single cell RNAseq for expression of classical markers to identify the subset of cells most likely to represent the TIC subpopulation. Our results indicate that STAT3 and GLI transcriptional activity can serve as uniquely PDX-dependent indicators for TIC function for a subpopulation of cancer cells. Also, multiple potential TIC signaling pathways, including STAT3, GLI, WNT, and OCT4/SOX2, can be active in a single PDX model, but do not necessarily associate with TIC function, indicating TIC diversity in TNBC. These data in conjunction with treatment response and single-cell RNA-seq sheds light on the molecular diversity of TIC populations in TNBC, identifies candidate targets for therapeutic intervention, and will indicate whether TIC diversity is associated with differential response to different chemotherapeutic agents. Citation Format: John Landua, Ping Gong, Michael T. Lewis. Heterogeneity of tumor-initiating cell related signaling revealed by lentiviral signaling reporters in breast cancer patient-derived xenografts [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 1689.

Abstract 1792: FGFR3 mediated PARP1 tyrosine 158 phosphorylation promotes PARP inhibitor resistance

Abstract Poly(ADP-ribose) polymerase (PARP) inhibitors (PARPi), which induce DNA damage by inhibiting PARP1 enzymatic activity and trapping PARP on the damaged DNA, are used to eliminate BRCA1/2-mutated (BRCAm) cancer. However, clinical observations suggest that BRCAm tumors develop PARPi resistance. Current strategies to overcome PARPi resistance include impeding multiple DNA repair pathways to induce excessive DNA damage. Here, we propose a novel strategy targeting oncogenic receptor tyrosine kinases to enhance PARP trapping. By developing triple-negative breast cancer (TNBC) cells with acquired talazoparib resistance, we observed a high prevalence of activated fibroblast growth factor receptor 3 (FGFR3) kinase in these cells through kinase antibody array analysis. Mass spectrometry analysis and in vitro kinase assay suggested that FGFR3 phosphorylated PARP1 at tyrosine residues 158 and 176. Biochemistry studies suggested that only PARP1 tyrosine 158 phosphorylation contributes to PARPi resistance in the cells we developed. We then developed a monoclonal antibody against tyrosine 158 phosphorylated PARP1, and found that high-level PARP1 tyrosine 158 phosphorylation positively correlated with PARPi resistance in breast cancer patient-derived xenograft models. We further demonstrated that the combination of FGFR inhibitor and PARPi delayed DNA repair with prolonged PARP trapping. Moreover, synergy between PARPi and FGFR inhibition was observed in multiple TNBC cell lines with PARPi resistance in vitro. The combination of PARPi and FGFR inhibitor also showed synergism in vivo, and treatment with the combination of PARPi and FGFR inhibitor was tolerated in mouse models. These findings reveal that PARP1 tyrosine 158 phosphorylation facilitates resolving of the PARPi-induced PARP-trapping, and that the tyrosine 158 phosphorylated PARP1 may be an effective biomarker to indicate FGFR3 mediated PARPi resistance. Citation Format: MeiKuang Chen, Yuan Gao, Weiya Xia, Yu-Han Wang, Jennifer K. Litton, Yu-Yi Chu, Funda Meric-Bernstam, Helen Piwnica-Worms, Banu Arun, Jordi Rodon Ahnert, Yongkun Wei, Wei-Chao Chang, Hung-Ling Wang, Coya Tapia, Constance T. Albarracin, Shao-Chun Wang, Ying-Nai Wang, Gabriel N. Hortobagyi, Chunru Lin, Liuqing Yang, Dihua Yu, Mien-Chie Hung. FGFR3 mediated PARP1 tyrosine 158 phosphorylation promotes PARP inhibitor resistance [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 1792.

Abstract 5824: Fully-humanized, bioengineered miR-1291 modulates key nutrient transport and metabolism to control tumor metabolism

Abstract Tumor progression is accompanied by metabolic reprogramming to allow carcinoma cells to access essential nutrients under the tumor microenvironment. To meet energetic and material needs for high-rate proliferation and tumorigenesis, cancer cells generally rely more heavily on vital nutrients (e.g., glucose, amino acids, and vitamins) than normal cells. Therefore, controlling nutrient transport and metabolism in cancer cells is a promising therapeutic strategy. Our previous studies have suggested that microRNA-1291-5p (miR-1291-5p or miR-1291) is downregulated in pancreatic cancer (PC) cells and acts as a tumor suppressor. The current study is to delineate the role of miR-1291 in the control of PC cell metabolism and develop miR-1291 therapy. We first employed our newly-invented RNA bioengineering technology to achieve high-level and large-scale, in vivo fermentation production of a fully-humanized biologic miR-1291 agent at a high degree of homogeneity (> 4 mg BioRNA per liter of culture and > 97% purity). RNA sequencing studies revealed that BioRNA/miR-1291 was selectively processed to target miR-1291-5p to modulate the transcriptome in PC cells. Proteomics study showed that a number of proteins were reduced remarkably in AsPC-1 cells by biologic miR-1291-5p, among which PNPO, an enzyme that mediates the synthesis of essential nutrient vitamin B6 (VB6), was revealed as a leading target of miR-1291-5p. Besides PNPO and the previously-reported glucose transporter SLC2A1/GLUT1, amino acid transporter SLC7A5/LAT1 was verified as a direct target for miR-1291-5p. Downregulation of PNPO, LAT1, GLUT1 protein expression by biological miR-1291 led to sharp alteration of homeostasis of glucose, amino acids, and VB6 in human PC cells, and subsequent suppression of glycolysis capacity and mitochondrial function as well as the increase of oxidative stress. Combination treatment with bioengineered miR-1291 and 5-FU exerted strong synergism in the inhibition of PC cell and organoid proliferation in vitro. In addition, miR-1291 mono- and combination therapy were effective to control tumor growth in pancreatic cancer patient-derived xenograft (PDX) mouse models. These results demonstrate a critical role for miR-1291 in the regulation of PC cell metabolism which provides insight in developing new therapeutic strategies for PC. Citation Format: Mei-Juan Tu, Frank J. Gonzalez, Ai-Ming Yu. Fully-humanized, bioengineered miR-1291 modulates key nutrient transport and metabolism to control tumor metabolism [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 5824.

Abstract 327: Antitumor efficacy of trastuzumab deruxtecan in combination with adavosertib in HER2-expressing Cyclin E amplified gastroesophageal cancers

Abstract Purpose: Trastuzumab deruxtecan (T-DXd) is a HER2-directed antibody-drug conjugate approved for patients with HER2-positive gastric/gastroesophageal junction (GEJ) cancers. Amplification of CCNE1, which encodes cyclin E, has been proposed as a mechanism of resistance to HER2-targeted therapies. The Wee1 kinase inhibitor adavosertib (AZD1775) has demonstrated enhanced antitumor activity in the setting of cyclin E overexpression. We investigated the efficacy of T-DXd in combination with adavosertib in HER2-expressing cancers with and without co-amplification of CCNE1. Methods: Frequency of CCNE1 amplification was determined among patients with ERBB2 amplification in a clinical genomic database at the MD Anderson Cancer Center (MDACC) and within The Cancer Genome Atlas (TCGA). Sulforhodamine B assay, western blotting, and Annexin V staining were used to determine the effects of T-DXd and DXd in combination with adavosertib in vitro. Four HER2-expressing gastric/GEJ patient-derived xenograft (PDX) models with or without co-amplification of CCNE1 were tested with T-DXd in combination with adavosertib in vivo. Tumor volume and body weights were measured twice weekly and treatment responses were assessed by the relative treatment-to-control ratio, waterfall plots, and event-free survival (event defined as the day which tumors doubled in size from baseline). Results: Of 515 patients identified at MDACC with ERBB2 amplification, 60 (11.6%) had co-amplification of CCNE1. Of the 405 patients identified within the TCGA with ERBB2 amplification, 43 (10.6%) had co-amplification of CCNE1. Tumor types with the highest frequency of ERBB2/CCNE1 co-amplification were gastric/GEJ (38.1% MDACC, 31.0% TCGA), endometrial (25.0% MDACC, 29.6% TCGA), and ovarian (15.4% MDACC, 15.4% TCGA). Adavosertib was synergistic in combination with DXd in four cell lines with ERBB2 amplification, CCNE1 amplification, or ERBB2/CCNE1 co-amplification. Both DXd and T-DXd upregulated cyclin E and γH2AX protein expression at 24 and 96 hours, and the combination of DXd or T-DXd with adavosertib enhanced apoptosis. T-DXd induced durable tumor regression in two HER2 amplified and HER2 overexpressing gastroesophageal PDX cancer models with HER2 amplification/3+ overexpression and concomitant cyclin E amplification/expression. Adavosertib enhanced the antitumor activity of T-DXd in two HER2 low, cyclin E amplified gastroesophageal cancer PDX models (one with an ERBB2 G778A mutation). Conclusions: ERBB2 and CCNE1 are frequently co-amplified. Our findings provide supporting rationale for combining T-DXd with adavosertib in HER2-expressing cancers with CCNE1 co-amplification. Further study is needed to determine if therapeutic induction of cyclin E with DXd may sensitize to Wee1 kinase inhibition in patients without cyclin E amplification as well. Citation Format: Timothy P. DiPeri, Kurt W. Evans, Gabriela Raso, Yasmeen Q. Rizvi, Xiaofeng Zheng, Bryce Kirby, Kathleen Kong, Khandan Keyomarsi, Jaffer A. Ajani, Timothy A. Yap, Funda Meric-Bernstam. Antitumor efficacy of trastuzumab deruxtecan in combination with adavosertib in HER2-expressing Cyclin E amplified gastroesophageal cancers [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 327.

MODL-29. Molecular Landscape of a comprehensive panel of pediatric brain cancer Patient-derived orthotopic xenograft (PDOX) models inform unique targets for drug responsiveness

Abstract Brain tumor is a leading cause of cancer related death in children. In addition to replicating histopathology, animal models faithfully replicating genetic/epigenetic abnormalities, molecular subtypes and broad inter-tumoral heterogeneities are needed. Through direct implantation of patient surgical or autopsied tumor tissues into matching locations in the brains of SCID mice, we developed a panel of 150 PDOX mouse models. Here, we report the analysis of 74 of the 150 PDOX models, 45 matching patient tissues and 60 non-tumorigenic samples to a well-annotated reference cohort of 2,801 methylation profiles of primary brain tumors. Our data showed that the lack of tumorigenicity was neither correlated with molecular subtypes nor predicted by low cell viabilities of the patient samples. Methylation profiling identified PDOX models representing nearly a full spectrum of molecular subtypes of pediatric brain tumors including GBM, medulloblastoma, ependymoma and ATRT. Direct comparison with the original patient tumors confirmed the replication of molecular subtypes. ONCOplot [FB1] analysis of PDOX models derived from matching pairs of primary and recurrent tumors (n=8) revealed close clustering with the patient tumors. Investigation of metastatic properties was performed in 13 MB models by harvesting and sub-transplanting matching PDOX primary tumors in the cerebella and metastatic tumors in the spinal cords. To confirm the potential and power of PDOX models in preclinical drug testing, we applied fractionated radiation (2 Gy/day x 5 days) and optimized multi-agent combinatory chemotherapies in MB models of the four major subgroups. High-throughput combination drug screening with ~ 8,000 drugs in PDOX-derived GBM cell lines and primary cultures of MB PDOX cells identified a library of ~ 3,500 drugs that were active in pediatric brain tumors. In summary, this study provides detailed information on molecular subclassification of a uniquely large cohort of PDOX models to serve as essential tools for brain tumor research.