GIST Cell Lines Research Articles

Abstract A013: CDK2 and CDK4/6 inhibition in GIST: Mechanisms of response and resistance

Abstract Advanced GIST is characterized by genomic perturbations of key cell cycle regulators. Oncogenic activation of CDK4/6 results in RB1 inactivation and cell cycle progression. Given that single-agent CDK4/6 inhibitor (CDK4/6i) therapy failed to show clinical activity in advanced GIST, we evaluated strategies for maximizing response to therapeutic CDK4/6 inhibition. Targeted next-generation sequencing and multiplexed protein imaging were used to detect cell cycle regulator aberrations in GIST clinical samples (N=18), including 8 metastatic TKI-resistant GISTs. Multiple metastases were analyzed in 3 patients. The impact of CDK2i (CDK2 inhibitor-II), CDK4/6i (palbociclib or abemaciclib), and CDK2/4/6i (PF-06873600) was determined through cell proliferation and protein detection assays in vitro and in vivo. Mechanisms of acquired CDK2i and CDK4/6i resistance were characterized in GIST cell lines after long-term exposure. The results demonstrate recurrent genomic aberrations in cell cycle regulators causing co-activation of the CDK2 and CDK4/6 pathways. Identical aberrations of p16, RB1, and TP53 were present in all metastases from 3 patients. We show that therapeutic co-targeting of CDK2 and CDK4/6 is synergistic in GIST cell lines with intact RB1, through inhibition of RB1 hyperphosphorylation and cell proliferation (P<0.01). Intact RB1 predicted response to treatment, whereas RB1-deficient models were resistant. Moreover, we identify RB1 inactivation and a novel oncogenic cyclin D1 resulting from an intragenic rearrangement (CCND1::chr11.g:70025223) as mechanisms of acquired CDK inhibitor resistance in GIST. The CCND1 rearrangement deleted the cyclin D1 C-terminal Thr286 and Thr288 residues which mediate cyclin D1 proteasomal degradation, resulting in overexpression of an abnormal cyclin D1. CDK inhibitor resistance properties were corroborated by lentiviral transduction of the CCND1 fusion gene into fusion-negative GIST, leiomyosarcoma, and breast cancer cells. These studies establish the biologic rationale for CDK2 and CDK4/6 co-inhibition as therapeutic strategy in patients with advanced GIST, including patients with metastatic GIST progressing on TKIs. In addition, these findings expand the spectrum of potential CDK inhibitor resistance mechanisms with translational potential for improving cell cycle targeted therapies in other cancer types. Citation Format: Inga-Marie Schaefer, Matthew L. Hemming, Meijun Z. Lundberg, Matthew P. Serrata, Isabel Goldaracena, Ninning Liu, Peng Yin, Joao A. Paulo, Steven P. Gygi, Suzanne George, Jeffrey A. Morgan, Monica M. Bertagnolli, Ewa T. Sicinska, Adrian Mariño-Enríquez, Jason L. Hornick, Chandrajit P. Raut, George D. Demetri, Wen-Bin Ou, Sinem K. Saka, Jonathan A. Fletcher. CDK2 and CDK4/6 inhibition in GIST: Mechanisms of response and resistance [abstract]. In: Proceedings of the AACR Special Conference: Sarcomas; 2022 May 9-12; Montreal, QC, Canada. Philadelphia (PA): AACR; Clin Cancer Res 2022;28(18_Suppl):Abstract nr A013.

Efficacy of SCF drug conjugate targeting c-KIT in gastrointestinal stromal tumor

BackgroundGastrointestinal stromal tumor (GIST) is a rare type of cancer that occurs in the gastrointestinal tract. The majority of GIST cases carry oncogenic forms of KIT, the receptor for stem cell factor (SCF). Small molecule kinase inhibitor imatinib is effective in prolonging the survival of GIST patients by targeting KIT. However, drug resistance often develops during the therapeutic treatment. Here, we produced a SCF-emtansine drug conjugate (SCF-DM1) with favorable drug efficacy towards GIST cells.MethodsRecombinant human SCF (rhSCF) was expressed in E. coli cells and further purified with Ni–NTA Sepharose and Phenyl Sepharose. It was then conjugated with DM1, and the conjugated product SCF-DM1 was evaluated using in vitro cell-based assays and in vivo xenograft mouse model.ResultsSCF-DM1 was effective in inhibiting imatinib-sensitive and -resistant GIST cell lines and primary tumor cells, with IC50 values of < 30 nM. It induced apoptosis and cell cycle arrest in GIST cells. In xenograft mouse model, SCF-DM1 showed favorable efficacy and safety profiles.ConclusionsrhSCF is a convenient and effective vector for drug delivery to KIT positive GIST cells. SCF-DM1 is an effective drug candidate to treat imatinib-sensitive and -resistant GIST.

Abstract 5648: Response and resistance to CDK2 and CDK4/6 inhibition in GIST

Abstract Gastrointestinal stromal tumor (GIST) is the most common GI sarcoma and is generally initiated by KIT or PDGFRA mutations which are compelling therapeutic targets for tyrosine kinase inhibitors (TKI). However, the emergence of secondary mutations results in clinical resistance to available TKIs. GIST progression is driven by genomic events which incrementally target the p16-CDK4/6-RB1 and p14-TP53-RB1 pathways to create CDK4/6 and CDK2 oncogenic co-dependency. Based on limited efficacy of single-agent CDK4/6-inhibitor (CDK4/6i) therapy in GIST, we evaluated strategies of co-targeting CDK2 and CDK4/6. Multiplexed protein imaging (via Immuno-SABER) was validated for the detection of cell cycle regulator aberrations in GIST clinical samples (N=18), 7 of which were TKI-resistant, and including 3 patients in whom multiple metastases were analyzed. The impact of various CDK perturbants using CDK2i (CDK2 inhibitor-II), CDK4/6i (palbociclib or abemaciclib), and CDK2/4/6i (PF-06873600) was determined through cell proliferation and protein detection assays in GIST cell lines and murine xenografts. Mechanisms of acquired CDK2i and CDK4/6i resistance were characterized in GIST cell lines after long-term exposure. Abnormal expression/biallelic inactivation of CDKN2A/p16, RB1, and TP53 were identified in 7 (39%), 2 (11%), and 2 (11%) of 18 GISTs, respectively. Identical aberrations of p16, RB1, and TP53 were present in all metastases from 3 patients. Since 5 of 7 RB1-intact advanced GISTs had co-dysregulation of the CDK2 and CDK4/6 pathways, we evaluated co-inhibition of CDK2 and CDK4/6 in vitro and in vivo which inhibited cell proliferation (P&amp;lt;0.01) and RB1 hyperphosphorylation. Intact RB1 predicted response to treatment, whereas RB1-deficient models were resistant. Two resistant sub-lines emerged after 11 and 14 months of palbociclib exposure: one with biallelic genomic RB1 inactivation and the other with the first known example of a cyclin D1 coding sequence fusion with oncogenic properties (CCND1::chr11.g:70025223). The CCND1 fusion deleted the cyclin D1 C-terminal Thr286 and Thr288 residues which mediate cyclin D1 proteasomal degradation, resulting in overexpression of an abnormal cyclin D1. Palbociclib-resistance properties were corroborated by lentiviral transduction of the CCND1 fusion gene into fusion-negative GIST, leiomyosarcoma, and breast cancer cells. CDK2 and CDK4/6 pathway perturbations with retained RB1 are frequent in advanced GIST and can be conserved across metastases, creating a compelling biologic rationale for therapeutic cell cycle restoration. We show that co-inhibition of CDK2 and CDK4/6 is synergistic in GIST and highlight RB1 inactivation and a novel oncogenic cyclin D1 as mechanisms of acquired CDKi resistance. Hence, combination therapies targeting CDK2 and CDK4/6 with correlative biomarkers predictive of response should be evaluated in patients with metastatic or TKI-resistant GIST. Citation Format: Inga-Marie Schaefer, Meijun Z. Lundberg, Matthew L. Hemming, Sinem K. Saka, Matthew P. Serrata, Isabel Goldaracena, Ninning Liu, Peng Yin, Joao A. Paulo, Steven Gygi, George D. Demetri, Ewa Sicinska, Adrian Mariño-Enríquez, Jason L. Hornick, Chandrajit P. Raut, Wen-Bin Ou, Jonathan A. Fletcher. Response and resistance to CDK2 and CDK4/6 inhibition in GIST [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 5648.

Targeting the translational machinery in gastrointestinal stromal tumors (GIST): a new therapeutic vulnerability

Although KIT-mutant GISTs can be effectively treated with tyrosine kinase inhibitors (TKIs), many patients develop resistance to imatinib mesylate (IM) as well as the FDA-approved later-line agents sunitinib, regorafenib and ripretinib. Resistance mechanisms mainly involve secondary mutations in the KIT receptor tyrosine kinase gene indicating continued dependency on the KIT signaling pathway. The fact that the type of secondary mutation confers either sensitivity or resistance towards TKIs and the notion that secondary mutations exhibit intra- and intertumoral heterogeneity complicates the optimal choice of treatment in the imatinib-resistant setting. Therefore, new strategies that target KIT independently of its underlying mutations are urgently needed. Homoharringtonine (HHT) is a first-in-class inhibitor of protein biosynthesis and is FDA-approved for the treatment of chronic myeloid leukemia (CML) that is resistant to at least two TKIs. HHT has also shown activity in KIT-mutant mastocytosis models, which are intrinsically resistant to imatinib and most other TKIs. We hypothesized that HHT could be effective in GIST through downregulation of KIT expression and subsequent decrease of KIT activation and downstream signaling. Testing several GIST cell line models, HHT led to a significant reduction in nascent protein synthesis and was highly effective in the nanomolar range in IM-sensitive and IM-resistant GIST cell lines. HHT treatment resulted in a rapid and complete abolishment of KIT expression and activation, while KIT mRNA levels were minimally affected. The response to HHT involved induction of apoptosis as well as cell cycle arrest. The antitumor activity of HHT was confirmed in a GIST xenograft model. Taken together, inhibition of protein biosynthesis is a promising strategy to overcome TKI resistance in GIST.

KITlow Cells Mediate Imatinib Resistance in Gastrointestinal Stromal Tumor.

Gastrointestinal stromal tumor (GIST) is commonly driven by oncogenic KIT mutations that are effectively targeted by imatinib (IM), a tyrosine kinase inhibitor (TKI). However, IM does not cure GIST, and adjuvant therapy only delays recurrence in high-risk tumors. We hypothesized that GIST contains cells with primary IM resistance that may represent a reservoir for disease persistence. Here, we report a subpopulation of CD34+KITlow human GIST cells that have intrinsic IM resistance. These cells possess cancer stem cell-like expression profiles and behavior, including self-renewal and differentiation into CD34+KIThigh progeny that are sensitive to IM treatment. We also found that TKI treatment of GIST cell lines led to induction of stem cell-associated transcription factors (OCT4 and NANOG) and concomitant enrichment of the CD34+KITlow cell population. Using a data-driven approach, we constructed a transcriptomic-oncogenic map (Onco-GPS) based on the gene expression of 134 GIST samples to define pathway activation during GIST tumorigenesis. Tumors with low KIT expression had overexpression of cancer stem cell gene signatures consistent with our in vitro findings. Additionally, these tumors had activation of the Gas6/AXL pathway and NF-κB signaling gene signatures. We evaluated these targets in vitro and found that primary IM-resistant GIST cells were effectively targeted with either single-agent bemcentinib (AXL inhibitor) or bardoxolone (NF-κB inhibitor), as well as with either agent in combination with IM. Collectively, these findings suggest that CD34+KITlow cells represent a distinct, but targetable, subpopulation in human GIST that may represent a novel mechanism of primary TKI resistance, as well as a target for overcoming disease persistence following TKI therapy.

Abstract 3767: FBXO32 ubiquitin ligase mediates apoptosis evasion and adaptation to KIT inhibition in gastrointestinal stromal tumor - GIST

Abstract Introduction: GIST is the most common malignant mesenchymal neoplasm and features myogenic differentiation. Most GISTs depend upon oncogenic signaling through constitutively activated forms of KIT and exhibit remarkable responses to the first-line KIT inhibitor imatinib. Secondary resistance involves emergence of subclonal populations harboring heterogeneous KIT-secondary mutations that are not collectively inhibited by any approved therapy. We dissected KIT-downstream signaling nodes to identify mediators of the KIT oncogenic program, seeking targets of overarching relevance in polyclonal imatinib-resistant GIST. Methods: using clinically-representative human GIST cell models, we undertook pharmacological and functional screening to identify key targetable nodes within KIT-downstream PI3K/mTOR and RAS/MAPK pathways and to model therapeutic strategies in vitro and in vivo. Transcriptomic analysis (RNAseq) identified genes co-regulated by PI3K/mTOR and RAS/MAPK, and functional evaluations by lentiviral construct overexpression and gene knockdown elucidated roles in GIST biology. Results: PI3K/mTOR and MEK1/2 were the most essential KIT-dependent nodes irrespective of the type of KIT primary or secondary mutation. Single node ablation did not yield sustained antiproliferative and apoptotic effects, but concurrent intermittent inhibition was synergistic in vitro and in vivo, supporting the critical and complementary roles of these two pathways. Transcriptomic analyses underscored FBXO32 (a SCF E3 ubiquitin ligase and the main effector of muscular atrophy) as the most differentially expressed gene co-regulated by the PI3K/mTOR and RAS/MAPK pathways. FBXO32 was also the most differentially expressed gene after KIT inhibition in GIST cell lines, and enrichment in FBXO32 expression was found in post-imatinib GIST samples compared to baseline matched tumor tissue. We demonstrated that FBXO32 is transcriptionally upregulated by FOXO3a upon KIT or concurrent PI3K/mTOR and MEK1/2 blockage, paralleling human muscle regulation. Notoriously, FBXO32 proved to have a pro-survival role in GIST cells, enabling evasion of apoptosis. Microarray and in vitro and in vivo functional assays showed that FBXO32 mediates cell cycle exit and quiescence upon KIT or concurrent KIT-downstream pathways inhibition, thereby minimizing imatinib-induced apoptosis and facilitating GIST cell survival. Conclusions: PI3K/mTOR and MEK1/2 are critical mediators of KIT oncogenic signaling in GIST. FBXO32 emerges as a KIT-dependent E3 ubiquitin ligase transcriptionally repressed through KIT downstream pathways. However, FBXO32 induction upon therapeutic KIT signaling inhibition triggers quiescence and subsequent apoptosis evasion, thus emerging as a novel mechanism of adaptation to therapeutic inhibition of oncogenic KIT. Citation Format: Alfonso García Valverde, Jordi Rosell, Daniel Pilco Janeta, Sergi Sayols, Claudia Valverde, Anna Esteve, Marta Gut, Jordi Barretina, George D. Demetri, Jonathan A. Fletcher, Joan Carles, Joaquín Arribas, César Serrano. FBXO32 ubiquitin ligase mediates apoptosis evasion and adaptation to KIT inhibition in gastrointestinal stromal tumor - GIST [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3767.

Abstract 3540: Exploring novel treatment agent in GIST by using connectivity map

Abstract Gastrointestinal stromal tumors (GIST) is gastrointestinal tract sarcoma which commonly contains a mutation in the KIT or PDGFRA. Imatinib mesylate, a specific tyrosine kinase inhibitor of the KIT and PDGFRA, causes regression or stabilization of disease in most of the patients with metastatic GIST. However, acquired resistance to imatinib is inevitable. Recently, E26 variant 1 (ETV1) pathway was found to be a key downstream effector of KIT. In this study, we explored the role of ETVI and its downstream effectors pathway in GIST by bioinformatics analysis, and identified potential novel agents that target ETV1 pathway of GISTs by using expression profile and a pattern-matching software “Connectivity Map” (CMAP), and confirmed the efficacy of novel agents by using GIST cell line in vitro assay. By using CMAP, four drugs with potential ETV1-inhibition effect were identified: suberanilohydroxamic acid (SAHA,Vorinostat) and trichostatin, two histone deacetylase inhibitors (HDACI), and trifluoperazine (TFP) and thioridazine (TDZ), two drugs of phenothiazine class. All four drugs were found to have ETV1-down regulating effect by immunoblotting. Phenothiazine could induce apoptosis and autophagy in GIST. Treatment of phenothiazine was found paradoxically up-regulate MEK activity. Combination of MEK inhibitor and phenothiazine showed great synergistic effect in exerting cytotoxicity against GIST cell lines. Our study established phenothiazine as a new class of agent with therapeutic effect in GIST, and combination of phenothiazine and MEK inhibitor showed great potential in GIST treatment. Citation Format: Chueh-Chuan Yen, Li-Tzong Chen, Chien-Feng Li. Exploring novel treatment agent in GIST by using connectivity map. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3540.

Abstract 3491: Rational combination of targeted agents and BH3 mimetics to improve cancer treatment

Abstract Background: In the clinic is often observed that melanoma cancer patients first respond to targeted therapeutic agents like dabrafenib showing tumor burden shrinkage, but subsequently relapse with tumors that are more aggressive, no longer responsive to therapy and showing acquired metastatic capacity. Most chemotherapeutic agents kill via the mitochondrial pathway of apoptosis, but unfortunately tumors frequently are able to survive, often adapting their antiapoptotic strategy. We used a variant of the functional predictive test dynamic BH3 profiling (DBP) to determine the tumor's antiapoptotic defense adaptation over time, and rationally identify combination of therapies to restore programmed cell death. Hypothesis: By using different BH3 peptides in DBP we can determine the adaptive tumor's antiapoptotic defense and adjust therapy to improve cancer cell death. Results: In BRAF mutant melanoma, cKIT mutant GIST and HER2 overexpressing breast cancer cell lines, we performed DBP over time using different BH3 peptides: BIM, BAD, PUMA, MS1 and NOXA. We observed that when they are treated with specific targeted agents against their oncogenic drivers, there is an increase in MCL-1 dependence over time. This observation points to a possible tumor defense mechanism conferring resistance against anticancer drugs. Based on these observations, we exploited these altered dependencies by pretreating the cancer cells with these targeted agents and combining them with novel BH3 mimetics targeting MCL-1 to synergistically enhance cell death. Exploiting DBP's capacity to measure changes in priming without the requirement for prolonged ex vivo culture, we assessed if this antiapoptotic evolution was also present in primary melanoma samples. Using FACS-based DBP to select for specific tumor populations, we tested the possible use of BH3 mimetics in combination with targeted agents to improve cancer treatment in the clinic. Conclusions: Our preliminary results reveal the use of dynamic BH3 profiling to be used not only as a powerful real-time tool to predict chemotherapy response but also to evaluate the rational combination of targeted agents with BH3 mimetics in vitro and in the clinic. [R.H. and A.L. contributed equally to this work.] Citation Format: Joan Montero, Dorota E. Sadowicz, Jeremy Ryan, Charles H. Yoon, Rizwan Haq, Anthony Letai. Rational combination of targeted agents and BH3 mimetics to improve cancer treatment. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3491.

Abstract A40: MicroRNA-494 inhibits growth of the gastrointesintal stromal tumor by targeting multiple cell cycle regulatory genes

Abstract Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal neoplasms occurring throughout the entire region of the gastrointestinal tract. Two oncogenes of the receptor tyrosine kinase family, KIT and PDGFRA, have gain-of-function mutations in approximately 70% and 15% of GISTs. Mutations of these 2 genes cause sustained activation, resulting in constant stimulation of the downstream signaling pathways of KIT and platelet-derived growth factor receptor, alpha polypeptide. Although KIT mutations are present in approximately 70% of GISTs,KIT overexpression is observed in more than 90% of GISTs. Dysregulation of microRNAs(miRNA)-494 was demonstrated to be a major cause for the overexppression of mutatnt KIT proteins in GISTs. It was also demonstrated that miR-494 treatment inhibits proliferation of GISTs and promotes apoptosis however, no details mechanisms are not known. The purpose of this study was to provide evidence regarding the roles of miR-494 in GIST suppression. To elucidate the specific pathways affected by miRNA-494 treatment on GIST cell lines, transcriptome analysis of GIST430 and GIST882 cells transfected with nontargeting microRNA or with miR-494 was performed and subsequently validated by qRT-PCR. Based on the transcriptome analysis, we found that multiple cell cycle relevant pathways were down-regulated by miR-494 treatment and importantly, core genes such as KIT, CASP2, CDCA2, AURKB, CDCA8, E2F7, BIRC5 and CKS1B involved in the regulation of cell cycle pathways were identified as novel targets by miR-494. Using reporter assays, we further validated that miR-494 directly binds to the 3'UTR regions of BRAC5, CKS1B and E2F7 and induces biological alterations. To demonstrate the perturbation of cell cycle pathways by miR-494, FACS analysis using five GIST cell lines was performed and we found that all of the five GIST cell lines transfected with miR-494 subjected to cell cycle arrest and apoptosis. Colony forming and proliferation assays also demonstrated that miR-494 treatment induced the growth inhibition of GIST cells which was derived from the alterations of cell cycle pathways. Taken together, our findings suggest that the efficient suppression of GISTs by miR-494 is driven not only by targeting KIT, but also multiple genes involved in cell cycle pathways Citation Format: SeongJu Yun, Won Kyu Kim, Hoguen Kim. MicroRNA-494 inhibits growth of the gastrointesintal stromal tumor by targeting multiple cell cycle regulatory genes. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr A40.

Abstract C6: Antiapoptotic defense evolution in cancer: novel therapeutic strategies to restore cell death using dynamic BH3 profiling

Abstract Background: There is a lack of effective predictive biomarkers to efficiently assess the optimal treatment for each patient and overcome resistance. When effective death signaling is initiated by a targeted therapy, an increase in mitochondrial apoptotic sensitivity (or ‘priming for death’) can be observed within hours with Dynamic BH3 Profiling (DBP). Most chemotherapeutic agents kill via the mitochondrial pathway of apoptosis, but unfortunately tumors frequently are able to survive adapting their antiapoptotic strategy. By using different BH3 peptides we can determine the tumor's defense adaptation over time and find novel ways to rationally develop combination therapies and restore apoptotic cell death. Hypothesis: Analyzing the tumor's antipoptotic defense over time with DBP we can predict how the tumor will adapt to the first-line therapy and restore cell death using BH3 mimetics. Results: Our first set of experiments consisted in characterizing the tumor's adaptation to therapy over time. We performed DBP over time using different BH3 peptides: BIM, BAD, MS1, PUMA and NOXA. In GIST and melanoma cell lines, we observed that when they were effectively treated with imatinib (KIT inhibitor) and dabrafenib (BRAFV600E inhibitor) respectively, or to selumetinib (MEK inhibitor) in both cases, there was an increased dependence in MCL-1 between 24 and 48 hours. This observation explains a possible mechanism by which tumors adapt their antiapoptotic dependencies. Based on these observations, we exploit these altered dependencies by pretreating the cancer cells with the above mentioned targeted therapies and then add a specific MCL-1 inhibitor to restore cell death. Other cell lines, including breast cancer and melanoma, were subsequently studied, and the combination of specific targeted therapies with BH3 mimetics (ABT-199, ABT-263 and MCL-1 inhibitors) was also assessed. Exploiting DBP's capacity to measure changes in priming without the requirement for prolonged ex vivo culture, we assessed if this antipoptotic evolution was also present in primary melanoma samples. We tested combinations of targeted therapies with BH3 mimetics to incorporate antiapoptotic targets to cancer treatment. Conclusions: Our cell line and clinical experiments demonstrate the potential for Dynamic BH3 profiling to be used as a powerful real-time tool to predict chemotherapy response and evaluate induced dependencies in relapsed patients. By studying combinations of targeted agents with BH3 mimetics we aim to improve cancer treatment. Citation Format: Joan Montero, Dorota E. Sadowicz, Rizwan Haq*, Anthony Letai*. Antiapoptotic defense evolution in cancer: novel therapeutic strategies to restore cell death using dynamic BH3 profiling. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr C6.

Pharmacological Inhibition of KIT Activates MET Signaling in Gastrointestinal Stromal Tumors.

Gastrointestinal stromal tumors (GIST) are the most common adult sarcomas and the oncogenic driver is usually a KIT or PDGFRA mutation. Although GISTs are often initially sensitive to imatinib or other tyrosine kinase inhibitors, resistance generally develops, necessitating backup strategies for therapy. In this study, we determined that a subset of human GIST specimens that acquired imatinib resistance acquired expression of activated forms of the MET oncogene. MET activation also developed after imatinib therapy in a mouse model of GIST (KitV558del/+ mice), where it was associated with increased tumor hypoxia. MET activation also occurred in imatinib-sensitive human GIST cell lines after imatinib treatment in vitro. MET inhibition by crizotinib or RNA interference was cytotoxic to an imatinib-resistant human GIST cell population. Moreover, combining crizotinib and imatinib was more effective than imatinib alone in imatinib-sensitive GIST models. Finally, cabozantinib, a dual MET and KIT small-molecule inhibitor, was markedly more effective than imatinib in multiple preclinical models of imatinib-sensitive and imatinib-resistant GIST. Collectively, our findings showed that activation of compensatory MET signaling by KIT inhibition may contribute to tumor resistance. Furthermore, our work offered a preclinical proof of concept for MET inhibition by cabozantinib as an effective strategy for GIST treatment.

Abstract 5136: HDACi inhibits liposarcoma via targeting of the MDM2-p53 signaling axis and PTEN, irrespective of p53 mutational status

Abstract Liposarcoma (LPS) is the most common sarcoma of humans. There are no systemic therapeutic regimens known to improve survival when complete surgical resection is not feasible, underscoring the need for an improved molecular understanding of LPS to stimulate the development of effective targeted therapies. The MDM2-p53 pathway plays a prominent role in WDLPS pathogenesis, with the vast majority of human tumors harboring either MDM2 amplifications or p53 mutations. Nutlin-3, a small-molecular antagonist of MDM2, inhibits cell proliferation via blocking the interaction of MDM2-p53 in LPS with wild-type p53. We wonder whether MDM2 overexpression due to amplification or/and mutant p53 still plays a crucial oncogenic role in LPS containing p53 mutation. Herein, we developed isogenic liposarcoma lines in which the parental forms were MDM2 amplification and p53 wild-type, whereas sublines had mutation p53 expression, showed Nutlin-3 resistance. mRNA sequencing and immunoblotting showed that MDM2 and CDK4 were overexpression in the parental lines and sublines, whereas MDM2 and CDK4 expression was low in mesothelioma and GIST cell lines. HDAC inhibitor (HDACi, SAHA and LBH589) treatment resulted in the dephosphorylation and degradation of MDM2 and p53, but little affected CDK4 and JUN expression, irrespective of p53 mutational status in LPS with MDM2 amplification, and in a mesothelioma cell line JMN1B with p53 mutation, but not two mesothelioma lines containing normal MDM2 level and wild-type p53. Our findings indicate that regulation of wild-type 53 degradation by HDACi is MDM2 amplification-dependent. HDAC inhibition by SAHA and LBH589 had a substantially effect on LPS and mesothelioma proliferation and survival associated with upregulation of the PTEN and p21, inhibition of cell proliferation marker cyclin A and PCNA expression, induction of G1 or G2 phase arrest; induction of apoptosis showing an increase of caspase 3/7 activity and expression, PARP cleavage, and the sub-G1 apoptotic population. Moreover, we characterize biological functions of MDM2-p53 axis in nutlin-3-sensitive and nutlin-3-resistant LPS, showing MDM2 knockdown in the four LPS lines and p53 knockdown in the three mutant-p53 LPS lines resulted in anti-proliferative effects. Additive effects were obtained through a coordinated attack on MDM2 and p53, as demonstrated by immunoblots, cell viability and cell cycle analyses, showing that MDM2 and p53 knockdown, in LPS cell lines containing the p53 mutation, induced greater cell apoptosis and anti-proliferative effects, compared to either intervention alone, which is comparable to the effects seen after HDAC inhibition by SAHA and LBH589. These compelling pro-apoptotic and anti-proliferative responses indicate that HDAC inhibition warrants clinical evaluation as a novel therapeutic strategy in LPS, including nutlin-3 resistant sublines with the p53 mutation. Note: This abstract was not presented at the meeting. Citation Format: Wen-bin Ou, Hailong Li, Li Liu, Shengmei Zhou, Ye Kuang, Ziqin Yan, Zhe Jiang, Si Shi, Fanguo Meng, Qing Sheng, Haimeng Zhou, Jonathan A. Fletcher. HDACi inhibits liposarcoma via targeting of the MDM2-p53 signaling axis and PTEN, irrespective of p53 mutational status. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 5136. doi:10.1158/1538-7445.AM2014-5136

Abstract 3396: Dual lineage inhibition of ETV1 and KIT disrupts the ETV1-KIT feed forward circuit and potentiates imatinib antitumor effect in GIST oncogenesis

Abstract Gastrointestinal stromal tumour (GIST) is one of the most common types of human sarcoma and is primarily defined by activating mutations in the KIT or PDGFRA receptor tyrosine kinases. Despite the initial clinical success of imatinib that specifically target mutant KIT and PDGFRA, imatinib resistance has become the biggest challenge in the management of advanced GIST patient. Novel therapeutics that can improve the efficacy of first line imatinib therapy and/or prevent and overcome imatinib resistance is imperative. The ETS family member, ETV1, has been previously identified as a lineage-specific survival factor that cooperates with mutant KIT by ETV1 protein stabilization by active MAP kinase signaling downstream of KIT signaling in GIST oncogenesis. Here we demonstrate that ETV1 is required for GIST initiation and maintenance in vivo using compound genetically engineered mouse models (GEMM). We further identified that ETV1 forms a feed forward circuit in GIST oncogenesis where the ETV1 protein is stabilized by active MAP kinase signaling downstream of KIT and stabilized ETV1enhances KIT expression through direct binding to the KIT enhancer regions in both GIST mouse models and human GIST cell lines. The dual lineage targeting of KIT by imatinib and ETV1 by MEK162 (an MEK inhibitor) disrupts the ETV1-KIT feed forward circuit and induces more apoptosis than single agent imatinib or MEK162 in human GIST cells. The combination therapy resulted in complete tumor regression whereas single agent imatinib or MEK162 resulted in stabilization of disease in human GIST xenograft studies. Moreover, the combination therapy also induced more tumor fibrosis than single agent imatinib or MEK162 in GIST GEMM. These observations demonstrate the in vivo role of ETV1 in GIST oncogenesis and the feasibility of targeting ETV1 protein stability by inhibiting MAP kinase signaling. They also suggest that the dual lineage targeting of ETV1 and KIT by the combination therapy may provide a more effective therapeutic strategy than imatinib alone in GIST management. Citation Format: Leili Ran, Inna Sirota, Zhen Cao, Devan Murphy, Shipra Shukla, Ferdinando Rossi, John Wongvipat, William D. Tap, Peter Besmer, Cristina R. Antonescu, Yu Chen, Ping Chi. Dual lineage inhibition of ETV1 and KIT disrupts the ETV1-KIT feed forward circuit and potentiates imatinib antitumor effect in GIST oncogenesis. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 3396. doi:10.1158/1538-7445.AM2014-3396

A Panel of Gastrointestinal Stromal Tumours (Gist) Xenograft Models for in Vivo Preclinical Drug Testing

ABSTRACT Aim: Most GIST are dependent on the KIT/PDGFRA signalling pathway and therefore can be treated with specific tyrosine kinase inhibitors (TKI). With time GIST develop resistance to the TKI. We need reliable models which can be used for preclinical drug testing. Methods: Patient-derived xenografts were established in nu/nu NMRI mice by subcutaneous implantation of fresh, surgically resected tumour specimens from consenting patients with GIST treated at the University Hospitals Leuven. Once tumour growth was observed, pieces of tumour were re-transplanted to next generations of mice. At each passage tumour fragments were collected for histological and molecular characterization. A model was considered to be established after observing stable histological and molecular features for at least two passages. Results: At present we have access to five successfully established GIST models, with different KIT mutations and different sensitivity to standard TKIs (Table). Established GIST xenograft models and their sensitivity to standard treatment. In vivo efficacy was assessed by % of relative tumour volume after 3 weeks under either imatinib (50mg/kg BID) or sunitinib (40mg/kg QD) Model* KIT mutation In vivo efficacy Imatinib Sunitinib UZLX-GIST1 p.V560D 20 20 UZLX-GIST2 p.A502_Y503dup 170 80 UZLX-GIST3 p.W557_V559delinsF 25 25 UZLX-GIST4 p.K558_G565delinsR 45 20 UZLX-GIST9 p.P577del; p.W557LfsX5; p.D820G 250 170 *models which showed tumor growth for at least two passages Untreated tumours from subsequent passages show morphological and immunohistochemical features resembling the original patient biopsy, verifying the stability GIST xenografts growth. KIT mutational status has also been confirmed in samples obtained from all models. These models are used for subsequent testing of new therapeutic approaches and current results have had an impact on at least four prospective clinical trials. If needed, we parallel our patient-derived models with in vitro or in vivo studies using one of the few established GIST cell lines. Conclusions: Our established GIST xenografts maintain the histological and molecular characteristics of the original patient sample and are a very useful and reliable preclinical platform for testing the new targeted approaches in GIST. Disclosure: All authors have declared no conflicts of interest.

In vitro and in vivo activity of regorafenib (REGO) in drug-resistant gastrointestinal stromal tumors (GIST).

10510 Background: KIT and PDGFRA mutations (mut) are the crucial transforming events in most GISTs, and tyrosine kinase inhibitors (TKIs) with activity against KIT and PDGFRA, such as imatinib (IM) (front-line therapy) and sunitinib (SU) (second-line therapy), are effective treatments in GIST patients (pts). Resistance to IM and SU is commonly associated with evolution of secondary kinase mut. REGO is a multi-targeted TKI that inhibits KIT, PDGFR, and other oncologic targets and has recently shown benefit in pts with metastatic GIST after progression on standard treatments. We evaluated the in vitro and in vivo activity of REGO compared with IM, SU, and sorafenib (SOR) (a multi-TKI structurally related to REGO). Methods: REGO, IM, SU, and SOR inhibition of viability and KIT phosphorylation was assessed in human GIST cell lines and in Ba/F3 cells transformed by KIT oncoproteins with IM-resistant ATP binding pocket or activation-loop mut. KIT/PDGFRA genotyping was performed in GISTs responding or progressing on REGO in the academic phase II clinical trial. Results: In GISTs with KIT exon 11 mutant oncoproteins, REGO potently inhibited viability, KIT phosphorylation, and downstream effector phosphorylation (AKT, MAPK, S6). IM-resistant activation loop mut were more potently inhibited by REGO than SU, whereas the gatekeeper IM-resistant mut T670I was inhibited by both REGO and SU, and the common ATP-binding pocket mutant V654A was more potently inhibited by SU than REGO. Two GIST metastases progressing in one pt after initial response to REGO contained KIT V654A mut. SOR and REGO demonstrated comparable in vitro overall activity. Representative GIST cell line viability IC50s are shown in the Table (values in bold indicate expected clinical relevance). Conclusions: In vitro studies confirm REGO is a potent inhibitor of KIT exon 11 mut in GIST and appears to have stronger activity than SU against the most common KIT activation-loop mut observed in GIST. Ongoing clinical correlative analyses from REGO-treated study patients will be presented. [Table: see text]

Using pegylated arginine deiminase (ADI-PEG20) for the treatment of sarcomas that lack argininosuccinate synthesase 1 expression.

10526 Background: To better understand the metabolism of sarcomas, we explored the role of argininosuccinate synthase 1 (ASS1), the rate-limiting enzyme in the conversion of citrulline to arginine in the urea cycle. When ASS1 is not expressed the amino acid arginine becomes an essential amino acid that must be delivered from the diet. The potential for therapeutic manipulation of this pathway with pegylated arginine deiminase (ADI-PEG20) was explored using metabolomic mass spectroscopy. Methods: Immunohistochemistry (IHC) of ASS1 expression on 701 patient specimens representing 45 subtypes of sarcoma was performed. Proliferation and cell death were analyzed in a panel of cell lines. Clonogenic assays were performed using ADI-PEG20 and chloroquine. Autophagy induction and inhibition was assessed. Xenograft models were tested. Metabolomic mass spectroscopy was performed in the presence and absence of ADI-PEG20. Results: lHC analysis demonstrated that ASS1 is not expressed in over 85% of sarcomas (619 of 701 patient samples). Treatment of a panel of LMS, synovial, ASPS, GIST, MPNST, chondrosarcoma, osteosarcoma, and Ewings sarcoma cell lines with ADI-PEG20 resulted in cell cycle arrest but not apoptosis when ASS1 expression was low (13/15 lines). Xenografts of the ASS1 low MNNG/HOS cells and SKLMS1 cells into nude mice followed by treatment with ADI-PEG20 +/- chloroquine demonstrated significantly slower tumor growth as compared to PBS or chloroquine treated controls. Treatment of sarcomas that lack ASS1 with ADI-PEG20 induced autophagy. Global analysis of metabolomic mass spectroscopy demonstrated alterations in PKM2, glutamine dependence, the Warburg effect, the TCA cycle and glutathione biology. Conclusions: This suggests that sarcomas may be sensitive to the arginine depletion therapy using ADI-PEG20. Autophagy induction has global metabolic consequences for sarcoma. Successful control of tumor burden by using arginine deprivation therapy and autophagy inhibition in our preclinical studies will help to orient clinical trials for this innovative treatment in patients with sarcoma that lack ASS1 expression.

Abstract 1824: Functional characterization of DOG1 in Gastrointestinal Stromal Tumors (GIST)

Abstract Introduction: DOG1, a calcium-dependent chloride channel, has been shown to be strongly and specifically expressed in GIST and is commonly used as diagnostic marker. Notably, KIT-negative GIST cell lines do not express DOG1. We have therefore sought to investigate the role of DOG1 for KIT expression and GIST growth and its potential as therapeutic target. Methods: GIST cell lines (GIST882: K642E, IM-sensitive, GIST-T1: del558-614, IM-sensitive, GIST48: V560D/D820A, IM-resistant) were transduced with lentiviral particles containing DOG1 sh-RNA. Expression of KIT and DOG1 was measured by immunoblotting. Effects of DOG1 knock-down on proliferation (with/without imatinib) were studied by BrdU incorporation assays and effects on chloride currents using patch clamp technique. Xenografts from transduced cells (and scrambled control) were generated and evaluated for tumor growth and expression of DOG1, KIT and Ki-67. Expression profiles from cell lines and xenograft tumors were compared using microarrays. Results: Lentiviral knock-down of DOG1 resulted in a near-complete inhibition of DOG1 while expression of KIT or the KIT-depending signalling pathways were not affected. Patch clamp measurements showed a strong inhibition of chloride currents in the majority of cells (DOG1-knock-down) tested compared to scrambled control. No difference was found in the growth rate of cells with DOG1-knockdown alone or in the presence of imatinib compared to cells transduced with scrambled sh-RNA. Xenografts generated from transduced cell lines retained DOG-1 knockdown as well as strong KIT-expression. Notably, xenografts with knock-down of DOG1 showed a lower proliferative activity (Ki-67: immunoreactivity score 12 (strongly positive) versus 6 (moderately positive)) with a reduction of tumor size of 43% after 19 days of tumor growth (n=8; p=0.003) when compared with the scrambled control. Expression analyses revealed MNK1 (MAP kinase-interacting serine/threonine-protein kinase 1) to be down-regulated in DOG1 negative GIST xenografts compared to scrambled control. Conclusion: DOG1 knock-down results in a reduction of intracellular chloride currents but does not affect KIT-expression or sensitizes cells to imatinib treatment. Proliferation is not affected by DOG1 knock-down in vitro but DOG1 appears to contribute to tumor growth in vivo. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 1824. doi:1538-7445.AM2012-1824

Abstract 2239: The effect combining the KIT inhibitor Imatinib with the PI3K inhibitor BKM120 or the dual PI3K/mTOR inhibitor BEZ235 on the proliferation of gastrointestinal stromal tumor cell lines

Abstract Gastrointestinal stromal tumors (GIST) are the most common type of sarcoma found in the gastrointestinal tract, with 5,000-6,000 cases per year diagnosed in the United States. Activating mutations in KIT, or occasionally in PDGFR are present in up to 90% of GISTs. The PI3K/AKT pathway has been demonstrated to be a crucial survival pathway in imatinib-resistant GISTs. In this study, we evaluated single agent activities of the KIT inhibitor imatinib, the PI3K inhibitor BKM120 and the dual PI3K/mTOR inhibitor BEZ235, as well as the synergistic activity of the combination of imatinib with BKM120 and imatinib with BEZ235 in a panel of GIST cell lines. For example, when the antiproliferative effects of imatinib and BKM120 were evaluated in combination at concentrations from 0.37 µM to 3.3 µM, growth suppression was observed in excess of the percent inhibition achieved by imatinib or BKM120 single agent treatment in the GIST882 and GIST430 cell lines. Our results suggest that combining imatinib with BKM120 or BEZ235 may provide a therapeutic strategy for imatinib-resistant GISTs. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2239. doi:1538-7445.AM2012-2239

Abstract A166: PTEN inactivation in gastrointestinal stromal tumors (GIST): Possible relevance for treatment of imatinib-resistant disease.

Abstract Background: Inactivation of PTEN occurs in numerous tumor types and has been implicated as a resistance mechanism to receptor tyrosine kinase (RTK)-targeted therapies in human cancer. The role of PTEN inactivation in GIST in the acquisition of late resistance to imatinib (IM), a RTK inhibitor, remains elusive. Aim: We investigated the incidence and the mechanisms of PTEN inactivation in a cohort of GIST specimens, and the impact of PI3K inhibition on KIT signaling in IM-sensitive vs. IM-resistant GIST cell lines. Methods: Molecular characterization of PTEN was performed on 65 (23 primary, 10 recurrent/metastatic, and 32 IM-resistant) GIST by CGH array/FISH, mutational analysis, PTEN promoter methylation and QPCR analysis. The PTEN expression and KIT downstream signaling were assessed by Western blotting (WB). The effect of PTEN knock-down by siRNA on KIT signaling was examined in KIT-positive (GIST48) and KIT-negative (KUL-5) GIST cell lines. The impact of NVP-BEZ235 (BEZ), a dual PI3K/mTOR inhibitor, was evaluated in vitro, alone and in combination with IM on IM-sensitive (GIST882, GIST882LY) and IM-resistant (GIST48, GIST48B) cell lines. GIST882LY is a sub-line with bi-allelic PTEN loss. GIST48 has an IM-resistant KIT-D820A mutation, whereas GIST48B has lost its former dependence on KIT oncoprotein. Biological consequences of PI3K inhibition were determined by WB for KIT signaling pathways and by cell proliferation and apoptosis assays. Results: On genomic level PTEN was affected mainly by mono-allelic loss (9%, 16% and 36% of primary, recurrent/metastatic and IM-resistant tumors, respectively). PTEN mutations or PTEN promoter hypermethylation were not found. By QPCR, a positive correlation was observed between low PTEN expression and its genomic loss (p&amp;lt;0.01). Reduced PTEN expression was confirmed in GIST with mono-allelic PTEN by WB. In vitro, siRNA silencing of PTEN in GIST cell lines resulted in up-regulation of PI3K/AKT/S6 and ERK1/2 pathways, which was partially reverted by PI3K inhibition. By WB, both GIST882 and GIST882LY were sensitive to IM and BEZ. However, GIST 882LY required higher dose of IM (250nM) and BEZ (50nM) to show the same inhibitory effects on AKT and S6 phosphorylation in comparison with GIST882 (100nM and 25nM, respectively). Inhibition of PI3K pathway resulted in up-regulation of ERK in these models. In contrast, combination of BEZ (25nM) and IM (50nM) induced a synergistic effect on the inhibition of both pathways. IM did not show effects on GIST48 or GIST48B. Conversely, both cell lines were sensitive to BEZ (at 200nM and 100nM, respectively). Antiproliferative effects of BEZ were strong in GIST882, GIST882LY and GIST48 cell lines in a dose and time dependent manner (IC50's of 25, 50 and 75nM after 72 h, respectively); these effects were potentiated in combination with IM. In contrast, BEZ alone (up to 100nM) or in combination with IM (up to 2000nM) did not inhibit proliferation of the KIT-negative GIST48B cell line. No induction of apoptosis was observed in any of our models or conditions. Conclusions: Downregulation of PTEN occurs frequently in IM-resistant GIST in parallel with PI3K/AKT pathway hyperactivation. Inhibitors of the PI3K/AKT pathway represent a potential treatment option for GIST and additional research efforts are required to fully explore this possible treatment strategy. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr A166.

PAWS-GIST national initiative: A joint patient/carer/professionals initiative in the United Kingdom for pediatric/adolescent/wild-type/syndromic gastrointestinal stromal tumors.

e20523 Background: Gastrointestinal stromal tumours (GISTS) are the most common mesenchymal tumours of the gastrointestinal tract. Over 90% of the GISTS have either KIT or PDGFRA mutations. The pediatric/adolescent/wild-type/syndromic (PAWS) GISTS are a heterogeneous group of patients with a different biology and natural history compared with the adult sporadic GISTS. Standard therapeutic approaches using tyrosine kinase inhibitors like imatinib or sunitinib have been less successful in PAWS-GIST patients. Better understanding of the biology and newer therapeutic strategies are needed to make any advances in the treatment of these extremely rare cancers. Methods: A joint collaborative group comprising of GIST specialists (Oncologists, pathologists, surgeons, radiologists, paediatricians, specialist nurses) and patients/carers was established in October 2010 to form the PAWS-GIST Focus group in UK. The first meeting took place in London in November 2010. The aims of the group are the following: (a) clinically gather UK-wide information on the number of patients, natural history, treatment, etc; (b) act as a facilitator between various specialist groups treating these patients; (c) provide information and support to patients and carers and opportunities for fundraising activities; and (d) encourage collaborative research work leading to better understanding of the biology, exploring newer small molecules/kinase inhibitors and explore establishing GIST cell lines. Results: As of January 2011, we have been able to identify 13 paediatric/adolescent GIST patients in United Kingdom. Work is in progress as part of the UK national GIST registry to prospectively identify all PAWS-GIST patients in the UK. Conclusions: PAWS-GIST initiative is the first of its kind in UK and is fully supported by the UK GIST professional and patient groups. Future plans include establishing national clinics based on the NIH USA model. Fund raising activities are ongoing.