Mechanisms Of Drug Sensitivity Research Articles

Abstract C176: Cancer-specific AI identifies multi-modal biomarkers of therapeutic response for 1,951 drugs including TNG348, a highly selective USP1 inhibitor

Abstract Understanding drug mechanism-of-action and improving clinical response rates are critical challenges to cancer therapeutic development. We designed a deep learning framework that learns molecular mechanisms of drug sensitivity and resistance from preclinical and/or clinical drug response data. The trained model can then generate biomarker strategies, suggest companion diagnostics, and propose rational drug-drug combinations that increase or restore drug sensitivity. The key innovation is the model’s architecture; the framework learns an embedding that reflects an interpretable relationship between a cellular process and drug response by emulating a tumor cell’s proteomic architecture. We applied the framework to learn genetic determinants of response for 1,951 drugs and 17,386 CRISPR gene knockouts screened in human cancer cell lines. The learned embeddings reflected known and novel mechanisms. RAF inhibitors, MEK inhibitors, and BRAF CRISPR knockout colocalized (P=1.7×10-11) due to their shared dependency on hyperactivation of RAF signaling. CDK4/6 inhibitor efficacy was strongly associated with a set of physically interacting proteins that control G1-to-S transition (P=5.6×10-12). Validating the clinical applicability of the finding, the trained model stratified patient response in a retrospective analysis of 70 ER+ metastatic breast cancer patients treated with the CDK4/6 inhibitor palbociclib (P=0.05, log-rank predicted responsive vs. non-responsive). As a case study, we examined mechanisms of sensitivity and resistance to TNG348 – a potent and highly selective USP1 inhibitor – in non-small cell lung cancer (NSCLC). By integrating genetics, gene expression, and CRISPR screens, the model identified seven genes whose expression explained >50% of the variation in response to TNG348 across 32 held-out NSCLC cell lines (Spearman R=0.74; P=8.7×10-7). The findings extend TNG348’s known relationship with homologous repair to additional protein complexes involved in cell cycle and DNA repair. The model leveraged the new associations to generate a six-biomarker inclusion-exclusion logic that captures both a large population (31.4% of TCGA lung adenocarcinoma patients meet inclusion criteria) and substantially enriches for strong response amongst preclinical models (OR=14.8, P=3.0×10-3). Finally, we asked the model to design rational combinations to overcome acquired resistance mechanisms. Proposed combinations included clinically successful combinations such as RAF+MEK inhibitors and were enriched for synergistic matches in a cell line screen of 986 drug pairings (Spearman rho P=6.6×10-4). These results highlight the promise of explainable AI to learn complex cellular mechanisms-of-action and generate non-obvious biomarker hypothesis and rational drug combination strategies for clinical development. Citation Format: Adam Yaari, Lee McDaniel, Antoine Simoneau, Samuel Meier, Oliver Priebe, Eduardo Farias, Alan Huang, Jannik Andersen, Yi Yu, Maxwell Sherman. Cancer-specific AI identifies multi-modal biomarkers of therapeutic response for 1,951 drugs including TNG348, a highly selective USP1 inhibitor [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr C176.

Preclinical Modeling of Pathway-Targeted Therapy of Human Lung Cancer in the Mouse.

Animal models, particularly genetically engineered mouse models (GEMMs), continue to have a transformative impact on our understanding of the initiation and progression of hematological malignancies and solid tumors. Furthermore, GEMMs have been employed in the design and optimization of potent anticancer therapies. Increasingly, drug responses are assessed in mouse models either prior, or in parallel, to the implementation of precision medical oncology, in which groups of patients with genetically stratified cancers are treated with drugs that target the relevant oncoprotein such that mechanisms of drug sensitivity or resistance may be identified. Subsequently, this has led to the design and preclinical testing of combination therapies designed to forestall the onset of drug resistance. Indeed, mouse models of human lung cancer represent a paradigm for how a wide variety of GEMMs, driven by a variety of oncogenic drivers, have been generated to study initiation, progression, and maintenance of this disease as well as response to drugs. These studies have now expanded beyond targeted therapy to include immunotherapy. We highlight key aspects of the relationship between mouse models and the evolution of therapeutic approaches, including oncogene-targeted therapies, immunotherapies, acquired drug resistance, and ways in which successful antitumor strategies improve on efficiently translating preclinical approaches into successful antitumor strategies in patients.

Integrative Drug and Genomic Profiling Identify Therapeutic Vulnerabilities and Inform Precision Medicine for Pediatric Acute Myeloid Leukemia

Background/Aims: Despite advances in chemotherapy-based treatment protocols, the outcomes of children with acute myeloid leukemia (AML) remain suboptimal. Implementation of targeted therapy based solely on genomics is challenging due to the complex mutational patterns and scarcity of pharmacologic agents for most lesions. In addition, pediatric and adult AML are genetically and biologically distinct, which poses a major hurdle for extrapolation of new agents approved for adult AML to the pediatric population. This study aims to adopt a functional approach that directly measure the response of patient-derived leukemic cells to targeted agents, and to establish the drug sensitivity pattern and identify candidates of immediate clinical relevance for precision usage in high-risk pediatric AML.Methods: A high-throughput drug screening, comprising 39 targeted agents (2 in Phase I, 10 in Phase II, 5 in Phase III, 22 FDA-approved) and 6 conventional chemotherapeutics, was performed on 30 pediatric AML samples collected at diagnosis or relapse using a serum-free, cytokine-supported culture system. A counter-screen of active drugs on cord blood hematopoietic stem cells was accomplished to reveal leukemia-selective activities. The robustness of the drug testing platform for predicting in vivo activities was validated in xenograft models. Genomic profiling was complementarily performed to identify the genetic markers and underlying mechanisms of drug sensitivity. Patients with refractory AML were treated with targeted agents based on drug profiling results, and assessed for clinical responses.Results: Unsupervised clustering revealed 5 distinct clusters of drug response: highly active compounds (IC50 <15 nM, 5 drugs); generally active compounds (IC50 <250 nM, 11 drugs); compounds with bimodal activities (wide IC50 ranges, 3 drugs); generally inactive compounds (16 drugs); and inactive compounds (IC50>2000 nM, 10 drugs). Targeted agents, including Bcl-2, HDAC, proteasome, HSP and survivin inhibitors, had substantially higher potency and selectivity over standard chemotherapeutic agents. New agents approved for adult AML were essentially inactive in pediatric AML. Drug sensitivity ex vivo accurately predicted in vivo single-agent and combinatorial activities with cytarabine in cell line- and patient-derived xenografts. Targeted resequencing of a 141-gene panel revealed novel mutations of prognostic relevance, such as KMT2C, in pediatric AML and their vulnerability to targeted agents. Whole-genome RNA-sequencing identified distinct gene expression signatures shaping the response to individual drugs. Administration of venetoclax to a child with refractory AML resulted in rapid blast clearance and achieved long-term remission. Complementary genomic profiling on serial specimens dictated the dynamic drug responses during disease evolution.Conclusions: Our study establishes a reliable drug testing platform and a pediatric-specific drug response profile of AML, which enables an evidence-based selection of targeted agents for patients without treatment options and endows therapies increasingly precise and personalized. The study also generates a valuable gene-drug-clinical dataset that could be leveraged to address the fundamental and translational biology of pediatric AML. It will ultimately impact the future design of clinical trials and protocols for managing this life-threatening malignancy. DisclosuresNo relevant conflicts of interest to declare.

Mouse Lymphoma Model-Based Senescence Exploitation to Predict Treatment Outcome of DLBCL Patients

Introduction: Molecularly informed treatment decisions are the basis of personalized cancer precision medicine. While gene expression profiling has been used to map the molecular landscape of tumors and to study the impact of molecularly defined subtypes on treatment outcome, biological effector principles, specifically cellular senescence, remain largely understudied. Syngeneic mouse models of cancer that recapitulate core molecular features of human malignancies could serve as useful models to explore mechanisms of drug sensitivity, and, likewise, to dissect molecular mechanisms of treatment failure. In particular, we focus here on the role of therapy-induced senescence, a histone H3 lysine 9 trimethylation (H3K9me3)-governed terminal cell-cycle arrest program.Methods: The Eµ- myc transgenic lymphoma mouse was used as a cross-species model of chemoresistance for patients diagnosed with diffuse large B-cell lymphoma (DLBCL). Gene expression profiles (GEP) of primary Eµ- myc lymphomas were obtained at diagnosis and after exposure to cyclophosphamide chemotherapy in vivo, with long-term outcome monitored in a clinical trial-like fashion. Lymphoma senescence capacity, a pivotal drug effector principle, was interrogated in mice by unbiased approaches as well as genetic perturbations.Results: DLBCL-established gene expression-based classifiers related to cell-of-origin (COO - i.e. GCB/ABC) subtypes and distinct DLBCL biologies (e.g. comprehensive consensus clusters [CCC]) were instrumental to identify comparable lymphoma subsets in the murine system. In turn, mice carrying senescence-defective lymphomas due to engineered loss of the H3K9me3 methyltransferase Suv39h1 or overexpression of the H3K9-active demethylase LSD1 succumbed faster to chemotherapy, as DLBCL patients with high LSD1 or low H3K9me3 expression in their lymphomas presented with significantly inferior outcome. Likewise, cross-species application of senescence-associated characteristics identified in unengineered primary mouse lymphomas unveiled molecular features important for treatment outcome in human DLBCL. Ongoing mouse model-based work aims at specific targeting of aberrant H3K9 demethylase activity to assess therapeutic long-term effects and to determine the conditions for future early-phase clinical testing in DLBCL patients.Conclusions: Primary Eµ- myc transgenic lymphomas serve as a faithful model for human DLBCL, since they share key pathologic and genetic features, including CCC/COO-based clustering. In turn, mouse lymphoma-based molecular exploration, genetic dissection and functional interrogation of treatment effector mechanisms, which is difficult to conduct in patient material, can be re-applied to molecularly and clinically co-annotated datasets from DLBCL patients to inform novel, lesion- and state-driven personalized treatment approaches and the design of future clinical trials. DisclosuresNo relevant conflicts of interest to declare.

Heterogenous chemosensitivity of a panel of organoid lines derived from small cell neuroendocrine carcinoma of the uterine cervix.

Small cell neuroendocrine carcinoma (SCNEC) of the uterine cervix is a rare disease with a poor prognosis. The lack of established disease models has hampered therapy development. We generated a panel of cancer tissue-originated spheroid (CTOS) lines derived from SCNEC of the uterine cervix using a method based upon cell-cell contact throughout the preparation and culturing processes. Using 11 CTOS lines, we assessed the sensitivity of various drugs used in clinical practice. Drug sensitivity assays revealed significant heterogeneous inter-CTOS chemosensitivity. Microarray analyses were then performed to identify sensitivity-related gene signatures. Specific gene sets were identified which likely contribute to the sensitivity to the tested drugs. We identified a line (Cerv54) that was exceptionally sensitive to irinotecan. Cerv54 had increased levels of CES1, which catalyzes the conversion of irinotecan to the active form, SN38, although in Cerv54 cells, SN38 was undetectable, CES1 expression and activity were markedly low compared to the liver, and a CES1 inhibitor had no effect on irinotecan sensitivity. These results suggested a novel irinotecan mode of action in Cerv54. Our CTOS lines may be useful for understanding the variation and mechanism of drug sensitivity, contributing to the understanding and development of chemotherapeutic drugs.

Ex Vivo Drug Sensitivity and Functional Genomics Platform Identifies Novel Combinations Targeting Intrinsic and Extrinsic Apoptotic Signaling Pathways in Multiple Myeloma

Introduction: Despite some long-term remissions, eventual drug resistance in most patients remains a critical obstacle in the treatment of multiple myeloma (MM). The development of new drugs/drug combinations with novel mechanisms of action are needed for continued improvement in patient outcomes. Initiation of tumor cell death via activation of the intrinsic (mitochondrial) and/or extrinsic (death receptor) apoptotic signaling pathways has been shown to be an effective therapeutic strategy in MM. Venetoclax (Ven) is a selective, small-molecule inhibitor of BCL-2 that exhibits clinical activity in MM cells, particularly in patients harboring the t(11;14) translocation. Navitoclax (Nav) is a small-molecule that targets multiple antiapoptotic BCL-2 family proteins, including BCL-XL, BCL-2, and BCL-W to initiate the intrinsic apoptotic pathway. Eftozanermin alfa (Eftoza) is a novel, second generation TRAIL receptor agonist that induces cell death via death receptor pathways and is under investigation in multiple solid and heme malignancies. In addition, the pan-BET inhibitor mivebresib (Miv) and the BDII selective BET inhibitor ABBV-744 have shown synergistic activity with Ven in cell line models of multiple heme malignancies. Results reported here describe ex vivo drug sensitivities and functional genomic analyses of Ven, Nav, Eftoza, Miv, and ABBV-744 alone or in combination with standard-of-care agents, including bortezomib, carfilzomib, panobinostat, daratumumab, or pomalidomide. Methods: A high-throughput ex vivo drug screening assay using a coculture system of bone marrow (BM)-derived MM and stromal cells was used to assess the sensitivity of MM patient tumor cells (Figure 1A). Paired whole exome sequencing (WES) and RNA sequencing (RNA-seq) analyses were performed. Results: Primary MM patient specimens (n=52) were evaluated in the ex vivo platform, including treatment-naïve, early relapse (1-3 prior lines), and late relapse (4-8 prior lines) patients treated with proteasome inhibitors, immunomodulatory drugs, and monoclonal antibodies. As expected, t(11;14)-positive MM patient specimens were more sensitive than wildtype to Ven ex vivo (D AUC, -18.6, P=0.002), however MM cells harboring amp(1q) were more resistant than wildtype (D AUC, +5.07, P=0.032), suggesting MCL1 (1q21 gene locus) is a key resistance factor to Ven single-agent activity in MM. Gene set enrichment analysis identified B-cell receptor signaling (normalized enrichment score (NES), 1.96, adjusted P=0.010) and MYC pathway (NES, 1.95, adjusted P=0.010) overexpression as predictors of increased sensitivity to Ven ex vivo. A t(11;14) gene expression signature was also generated using a penalized regression model approach in an additional MMWG/ORIEN MM patient cohort (n=155). The t(11;14) predictive gene expression signature was confirmed by correlation with Ven AUC in the ex vivo model. Additional pathway analyses were performed to identify potential predictive markers of sensitivity/resistance for each single agent and drug combination. Although ex vivo activity of Nav was higher in t(11;14) specimens compared to non-t(11;14) (D AUC, -17.8, P=0.011), ex vivo activity in non-t(11;14) specimens was also observed, indicating additional anti-MM activity by cotargeting of BCL-XL and BCL-2. Both Miv and ABBV-744 showed single-agent activity ex vivo, however Miv demonstrated higher activity (median LD50=88.4nM), suggesting that pan-BET inhibition is more effective than BDII-specific BET inhibition in MM. Finally, a novel drug-combination effect analysis was used that identified novel synergistic ex vivo combinations including Ven and panobinostat (P=0.0013) and Eftoza with bortezomib (P=1.8E-7) or carfilzomib (P=7E-4). Additionally, single-agent induction of macrophage-mediated phagocytosis was observed in both Ven and daratumumab, which was synergistic when the 2 drugs were combined (Figure 1B). Conclusion: An ex vivo functional genomic screen of MM patient specimens demonstrated the usefulness of this approach to identify candidate drugs and potential predictive biomarkers for continued evaluation in clinical trials. This approach confirmed known mechanisms of drug sensitivity and identified new ones, including a novel characterized immune-mediated synergy between Ven and daratumumab, and potential combination strategy for Eftoza and proteasome inhibitors. Figure 1 Disclosures Siqueira Silva: Karyopharm: Research Funding; NIH/NCI: Research Funding; AbbVie: Research Funding. Kulkarni:M2GEN: Current Employment. Mitchell:AbbVie: Other: payment for bioinformatics analysis, Research Funding; M2GEN: Current Employment, Research Funding. Dai:Cygnal Therapeutics: Current Employment; M2GEN: Ended employment in the past 24 months. Hampton:M2GEN: Current Employment. Lu:AbbVie: Current Employment, Current equity holder in publicly-traded company. Modi:AbbVie: Current Employment, Other: may own stock or stock options. Motwani:AbbVie: Current Employment, Current equity holder in publicly-traded company. Harb:AbbVie: Current Employment, Other: may hold stock or stock options. Ross:AbbVie: Current Employment, Current equity holder in publicly-traded company. Shain:Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; GlaxoSmithKline: Speakers Bureau; Sanofi/Genzyme: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Karyopharm: Research Funding, Speakers Bureau; AbbVie: Research Funding; Takeda: Honoraria, Speakers Bureau; Janssen: Honoraria, Speakers Bureau; Amgen: Speakers Bureau; Adaptive: Consultancy, Honoraria; BMS: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. OffLabel Disclosure: While this is a preclinical study, venetoclax for treatment of multiple myeloma is not an approved indication

Identifying Drug Sensitivity Subnetworks with NETPHIX

SummaryPhenotypic heterogeneity in cancer is often caused by different patterns of genetic alterations. Understanding such phenotype-genotype relationships is fundamental for the advance of personalized medicine. We develop a computational method, named NETPHIX (NETwork-to-PHenotype association with eXclusivity) to identify subnetworks of genes whose genetic alterations are associated with drug response or other continuous cancer phenotypes. Leveraging interaction information among genes and properties of cancer mutations such as mutual exclusivity, we formulate the problem as an integer linear program and solve it optimally to obtain a subnetwork of associated genes. Applied to a large-scale drug screening dataset, NETPHIX uncovered gene modules significantly associated with drug responses. Utilizing interaction information, NETPHIX modules are functionally coherent and can thus provide important insights into drug action. In addition, we show that modules identified by NETPHIX together with their association patterns can be leveraged to suggest drug combinations.

The Role of Non-Coding RNAs as Prognostic Factor, Predictor of Drug Response or Resistance and Pharmacological Targets, in the Cutaneous Squamous Cell Carcinoma.

Simple SummaryCutaneous Squamous Cell Carcinoma is poorly understood at a molecular level, however emerging evidence have highlighted that non-coding RNAs play a key role in the oncogenesis and tumor progression. Herein we focused on the translational potential of non-coding RNAs, such as miRNAs, circular RNAs and lncRNAs, by reporting cutting edge findings showing that they can be considered as potential predictors of response to diverse class of drugs, including chemotherapeutics and immune checkpoints inhibitors and as reliable prognosticators. Additionally, by reporting the molecular mechanisms of drug sensitivity and resistance in which non-coding RNAs are involved, we highlighted their potential role as druggable targets and suggested new therapeutic options for CSCC patients.Cutaneous squamous cell carcinoma (CSCC) is the most common keratinocyte-derived skin cancer in the Caucasian population. Exposure to UV radiations (UVRs) represents the main risk carcinogenesis, causing a considerable accumulation of DNA damage in epidermal keratinocytes with an uncontrolled hyperproliferation and tumor development. The limited and rarely durable response of CSCC to the current therapeutic options has led researchers to look for new therapeutic strategies. Recently, the multi-omics approaches have contributed to the identification and prediction of the key role of non-coding RNAs (ncRNAs), such as microRNAs (miRNAs), circularRNAs (circRNAs) and long non-coding RNAs (lncRNAs) in the regulation of several cellular processes in different tumor types, including CSCC. ncRNAs can modulate transcriptional and post-transcriptional events by interacting either with each other or with DNA and proteins, such as transcription factors and RNA-binding proteins. In this review, the implication of ncRNAs in tumorigenesis and their potential role as diagnostic biomarkers and therapeutic targets in human CSCC are reported.

Abstract A006: Preclinical models from rare aggressive tumors: Establishment, drug sensitivity and genomic analyses of mesothelioma PDX as models for oncology drug development

Abstract Malignant mesothelioma is a rare but aggressive disease with few therapeutic options and a poor prognosis. The main cause of the disease is the exposition to asbestos, even after 50 years of latency. Intended therapeutic goals, such as increase of patient survival and improvement of quality of life have only been achieved to a limited extent. Curative treatment is currently still not possible. Therefore, there is a strong clinical need for translational models. For this, we have characterized patient-derived xenograft models (PDX) of mesothelioma to evaluate their suitability for preclinical research. For PDX establishment, tumor tissue from surgically resected pleura or peritoneum mesotheliomas, was implanted subcutaneously into immunodeficient mice and serially passaged for at least 3 in vivo-passages. Established PDX models were characterized towards sensitivity to standard of care drugs. PDX tissue was collected for RNA sequencing and immunohistochemistry. The generated molecular profiles (comparative gene expression and sequence variations) of the models were analyzed comprehensively. A panel of 6 mesothelioma PDX has been established. In these PDX models, characteristic histological structure of the original patient tumor was preserved. The PDX models responded individually to standard of care drugs like pemetrexed, cisplatin, and gemcitabine as well as combinations thereof. The combination of pemetrexed or gemcitabine with cisplatin showed a stronger tumor growth inhibition than the respective monotherapies. The analyzed mutational and gene expression landscape resembles the known pattern of mesotheliomas. In summary, our established mesothelioma PDX will serve as a preclinical tool to investigate cellular and molecular mechanisms of drug sensitivity and resistance. They will provide a valuable preclinical model system to identify and validate biomarkers and to develop new therapeutic approaches. Citation Format: Michael Becker, Bernadette Brzezicha, Jana Rolff, Theresia Conrad, Beate Rau, Wolfgang Walther, Jens Hoffmann. Preclinical models from rare aggressive tumors: Establishment, drug sensitivity and genomic analyses of mesothelioma PDX as models for oncology drug development [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr A006. doi:10.1158/1535-7163.TARG-19-A006

Abstract A014: Gastric cancer PDX models for predictive preclinical studies: Establishment, drug sensitivity, and genomic characterization

Abstract Gastric cancer is a common malignant disease representing the fifth common cancer worldwide. Its prognosis is determined by the stage and localization (distal vs. proximal), making gastric cancer the third most lethal cancer type. In early and localized stages, the therapeutic aim is curative, in metastatic stages however, palliative treatment is the only option. Despite some progress in the last 10 years, mortality for late stage gastric cancer is still high due to late diagnosis of the disease. Therefore, we have established patient-derived xenograft models (PDX) of gastric cancer to determine their potential as a preclinical tool to improve diagnosis and therapies of this cancer. For PDX generation, tumor tissue specimens were collected from the gastric cancers immediately after surgery and transplanted subcutaneously into immunodeficient mice. Each PDX was serially passaged for at least three times. Established PDX models were treated in monotherapy with standard of care drugs, inter alia, from the European FLOT standard (5-FU, leucovorine, oxaliplatin and docetaxel). PDX tumor tissue was collected for immunohistochemistry as well as for next generation RNA sequencing and a comprehensive transcriptome analysis was performed. We established 7 new PDX models of gastric cancer and stably expanded them for further analyses. The histology of the primary tumor was preserved in the PDX model proven by histopathological evaluation. Treatment with standard of care drugs such as 5-FU, docetaxel, or oxaliplatin showed an individual response towards these drugs. The mutational and gene expression landscape of the PDX models resembles the pattern known from public gastric cancer data sets. In conclusion, the newly established gastric cancer PDX panel represents a tool for preclinical analyses of molecular mechanisms of drug sensitivity and resistance. They further provide a clinically relevant platform to identify new biomarkers and to develop novel therapies. Citation Format: Michael Becker, Bernadette Brzezicha, Theresia Conrad, Jana Rolff, Michael Linnebacher, Christoph Treese, Beate Rau, Wolfgang Walther, Jens Hoffmann. Gastric cancer PDX models for predictive preclinical studies: Establishment, drug sensitivity, and genomic characterization [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr A014. doi:10.1158/1535-7163.TARG-19-A014

Abstract 5113: A flexible pipeline for precision medicine biomarker detection and prediction of treatment response

Abstract Recent years have seen an explosion in the availability of paired molecular profiling and drug screen data, providing an unprecedented opportunity for the development of targeted therapies based on an individual’s genetic background. Despite a number of recent successes in diseases ranging from cystic fibrosis to cancer, significant hurdles remain in our ability to accurately predict treatments based on molecular profiling data. In particular, few such tools exist that allow the integration of heterogeneous data types (e.g. genomic, transcriptomic, and somatic mutations), along with high-throughput drug screen data to make predictions about treatment efficacy. Here, we describe a generalized open-source pipeline developed for the analysis of precision medicine data, Pharmacogenomics Prediction Pipeline, or “P3”. The modular design of P3 enables the inclusion of arbitrary input data types and the selection from multiple alternative machine learning algorithms, while automated statistical and visualization reporting steps incorporated throughout the pipeline assist in parameter tuning and early detection of problematic data components. Molecular profiling data is further enriched by the incorporation of external biological information in the form of pathway and gene set annotations such and Gene Ontology (GO) and The Molecular Signatures Database (MSigDB). To demonstrate the use of P3 for preclinical biomarker prediction, we applied P3 to an unpublished multiple myeloma dataset consisting of exome, RNA-Seq, CNV, and drug screen data for 1,912 compounds across 47 tumor cell lines. Specifically, P3 was used to predict molecular features associated with response to treatment for all drugs where a differential response to treatment was observed across patients. Furthermore, molecular profiling and drug screen data for 267 drugs and over a thousand cell lines spanning multiple cancer types from the Genomics of Drug Sensitivity in Cancer (GDSC) project were analyzed using P3, providing insights into shared mechanisms of drug sensitivity and resistance across different cancer and treatment types. Citation Format: V Keith Hughitt, Sayeh Gorjifard, Aleksandra M. Michalowski, John K. Simmons, Ryan Dale, Eric C. Polley, Jonathan J. Keats, Beverly A. Mock. A flexible pipeline for precision medicine biomarker detection and prediction of treatment response [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 5113.

Effect of miR-506 on pirarubicin sensitivity and on mesenchymal to epithelial transition and DNA damage homologous recombination repair process in leiomyosarcoma.

e22503 Background: MicroRNAs has been shown to be involved in many processes of tumorigenesis. Furthermore, miR-506 has been confirmed to be related to the occurrence of epithelial to mesenchymal transition (EMT) and tumor chemoresistance. However, the role of miR-506 in leiomyosarcoma has not yet been reported. Methods: A retrospective review of the 61 Chinese patients diagnosed with soft tissue leiomyosarcoma from January 2000 to December 2012 was investigated. Patient survival trends were examined by Kaplan-Meier analysis and log-rank tests. In vitro experiments were used to determine the biological function and physiological mechanism of drug sensitivity of miR-506. Results: In Chinese leiomyosarcoma patient cohort, we detected a decreased expression of miR-506 in the tumor tissues than that in marginal normal tissues. Tissue microarray analysis detected and confirmed a novel mesenchymal to epithelial transition (MET) in soft tissue leiomyosarcoma, also this MET phenomena was correlated with survival. In vitro, miR-506 inhibited the proliferation, migration, and invasion of leiomyosarcoma cells by promoting the occurrence of MET. Furthermore, RAD51, RAD52, ATM and ATR, which are defined as DNA damage homologous recombination repair related proteins, were highly expressed and associated with worse overall survival of leiomyosarcoma patients. Through targeting MET and DNA damage homologous recombination repair related proteins, miR-506 promoted the sensitivity of leiomyosarcoma cells to Pirarubicin. Conclusions: The current study reveals a novel mechanism that miR-506 increases the drug sensitivity of tumor cells, which are probably in association with the MET and DNA damage homologous recombination repair process.

Integrative Modeling Identifies Key Determinants of Inhibitor Sensitivity in Breast Cancer Cell Lines.

Cancer cell lines differ greatly in their sensitivity to anticancer drugs as a result of different oncogenic drivers and drug resistance mechanisms operating in each cell line. Although many of these mechanisms have been discovered, it remains a challenge to understand how they interact to render an individual cell line sensitive or resistant to a particular drug. To better understand this variability, we profiled a panel of 30 breast cancer cell lines in the absence of drugs for their mutations, copy number aberrations, mRNA, protein expression and protein phosphorylation, and for response to seven different kinase inhibitors. We then constructed a knowledge-based, Bayesian computational model that integrates these data types and estimates the relative contribution of various drug sensitivity mechanisms. The resulting model of regulatory signaling explained the majority of the variability observed in drug response. The model also identified cell lines with an unexplained response, and for these we searched for novel explanatory factors. Among others, we found that 4E-BP1 protein expression, and not just the extent of phosphorylation, was a determinant of mTOR inhibitor sensitivity. We validated this finding experimentally and found that overexpression of 4E-BP1 in cell lines that normally possess low levels of this protein is sufficient to increase mTOR inhibitor sensitivity. Taken together, our work demonstrates that combining experimental characterization with integrative modeling can be used to systematically test and extend our understanding of the variability in anticancer drug response.Significance: By estimating how different oncogenic mutations and drug resistance mechanisms affect the response of cancer cells to kinase inhibitors, we can better understand and ultimately predict response to these anticancer drugs.Graphical Abstract: http://cancerres.aacrjournals.org/content/canres/78/15/4396/F1.large.jpg Cancer Res; 78(15); 4396-410. ©2018 AACR.

Abstract 2887: Evaluating the effectiveness of MAPK, AKT, and mTOR inhibitors in reducing proliferation in cellular models of papillary and follicular thyroid cancer

Abstract Thyroid cancer is the most common endocrine malignancy, and incidence has been steadily increasing, suggesting occurrence will exceed that of colon cancer by 2030. Papillary and follicular thyroid cancer subtypes are most common, which are activated by BRAF and HRAS mutations, respectively. Interestingly, papillary and follicular thyroid cancers are associated with different pathologies and metastases. Currently, both papillary and follicular thyroid cancers are treated by surgical removal of the thyroid, followed by radioactive iodine treatment to eliminate any remaining tumor cells. Unfortunately, for individuals with progressive thyroid cancer, these treatment options are not effective, highlighting a need for increased investigation into mechanisms of drug sensitivity. In this study, we sought to evaluate mechanisms of differential drug sensitivity through RNA-seq analysis of cell lines derived from mouse models of papillary and follicular thyroid cancer. Additionally, we evaluated the effects of MAPK, AKT, and mTOR inhibitors on murine papillary and follicular thyroid cancer cell proliferation. To accurately assess drug efficacy, we calculated GR50 concentrations for each inhibitor in six papillary and follicular thyroid cancer cell lines. We discovered that papillary and follicular thyroid cancer cell lines responded differently to MAPK and AKT inhibitors, suggesting that treatment approaches should be tailored to subtype despite having mutations in the same signaling pathway. Citation Format: Brianna LeBoeuf, Braxton Anderson, Margaret Young, Aime Franco, Laura MacDonald. Evaluating the effectiveness of MAPK, AKT, and mTOR inhibitors in reducing proliferation in cellular models of papillary and follicular thyroid cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2887.

PO-467 Integrative modelling to understand and predict cancer drug response

IntroductionResponse to cancer therapeutics, whether in cell lines or patients, is variable and often difficult to predict. While many oncogenic drivers and drug resistance mechanisms have been described, understanding of how the interplay of these factors impacts response is still limited, precluding implementation in clinical practice.Material and methodsTo better understand the mechanisms behind the variability in drug response in breast cancer, we profiled both cell lines and tumour samples, in the absence of drug treatment, for the presence of mutations, copy number aberrations, mRNA and protein expression as well as protein phosphorylation. Response of the cell lines to several drugs in the RTK/PI3K/MAPK pathways was also determined. The molecular characteristics together with response data were used to construct a knowledge-based, Bayesian computational model that integrates all data types and estimates the relative contribution of the various drug sensitivity mechanisms. Using a set of patient samples, with known response to a combination treatment of trastuzumab, paclitaxel and carboplatin in a neoadjuvant setting (TRAIN trial), we have constructed a preliminary model to explain the observed variability in pathological complete response (pCR). Upon further refinement, the predictive ability of this model will be tested using an independent validation set of patient samples.Results and discussionsOur model of regulatory signalling is able to explain most of the variability observed in drug response in cell lines. It also identified cell lines with an unexplained response, and provided an opportunity for us to search for novel explanatory factors. Among others, we found that the 4E-BP1 protein expression level – and not just the extent of its phosphorylation – is a determinant of mTOR inhibitor sensitivity, which we validated experimentally.Extending the modelling approach to focus on a predictive clinical application, specifically response to chemotherapy combined with trastuzumab treatment in a neoadjuvant setting for HER2 +breast cancer, we have constructed a preliminary model able to explain part of the variability in pCR, the refinement and validation of which will be presented.ConclusionCombining molecular characterisation with integrative modelling can be used to systematically test and extend our understanding of the variability in anticancer drug response. Such approaches pave the way for establishing effective personalised treatments.

Integrated multi-omics data analysis identifying novel drug sensitivity-associated molecular targets of hepatocellular carcinoma cells.

Hepatocellular carcinoma (HCC) is the most common type of liver cancer and the third-leading cause of malignancy-associated mortality worldwide. HCC cells are highly resistant to chemotherapeutic agents. Therefore, there are currently only two US Food and Drug Administration-approved drugs available for the treatment of HCC. The objective of the present study was to analyze the results of previously published high-throughput drug screening, and in vitro genomic and transcriptomic data from HCC cell lines, and to integrate the obtained results to define the underlying molecular mechanisms of drug sensitivity and resistance in HCC cells. The results of treatment with 225 different small molecules on 14 different HCC cell lines were retrieved from the Genomics of Drug Sensitivity in Cancer database and analyzed. Cluster analysis using the treatment results determined that HCC cell lines consist of two groups, according to their drug response profiles. Continued analyses of these two groups with Gene Set Enrichment Analysis method revealed 6 treatment-sensitive molecular targets (epidermal growth factor receptor, mechanistic target of rapamycin, deoxyribonucleic acid-dependent protein kinase, the Aurora kinases, Bruton's tyrosine kinase and phosphoinositide 3-kinase; all P<0.05) and partially effective drugs. Genetic and genome-wide gene expression data analyses of the determined targets and their known biological partners revealed 2 somatically mutated and 13 differentially expressed genes, which differed between drug-resistant and drug-sensitive HCC cells. Integration of the obtained data into a short molecular pathway revealed a drug treatment-sensitive signaling axis in HCC cells. In conclusion, the results of the present study provide novel drug sensitivity-associated molecular targets for the development of novel personalized and targeted molecular therapies against HCC.

Haploid genetic screens identify genetic vulnerabilities to microtubule-targeting agents.

The absence of biomarkers to accurately predict anticancer therapy response remains a major obstacle in clinical oncology. We applied a genome‐wide loss‐of‐function screening approach in human haploid cells to characterize genetic vulnerabilities to classical microtubule‐targeting agents. Using docetaxel and vinorelbine, two well‐established chemotherapeutic agents, we sought to identify genetic alterations sensitizing human HAP1 cells to these drugs. Despite the fact that both drugs act on microtubules, a set of distinct genes were identified whose disruption affects drug sensitivity. For docetaxel, this included a number of genes with a function in mitosis, while for vinorelbine we identified inactivation of FBXW7,RB1, and NF2, three frequently mutated tumor suppressor genes, as sensitizing factors. We validated these genes using independent knockout clones and confirmed FBXW7 as an important regulator of the mitotic spindle assembly. Upon FBXW7 depletion, vinorelbine treatment led to decreased survival of cells due to defective mitotic progression and subsequent mitotic catastrophe. We show that haploid insertional mutagenesis screens are a useful tool to study genetic vulnerabilities to classical chemotherapeutic drugs by identifying thus far unknown sensitivity factors. These results provide a rationale for investigating patient response to vinca alkaloid‐based anticancer treatment in relation to the mutational status of these three tumor suppressor genes, and could in the future lead to the establishment of novel predictive biomarkers or suggest new drug combinations based on molecular mechanisms of drug sensitivity.

Link synthetic lethality to drug sensitivity of cancer cells.

Synthetic lethal (SL) interactions occur when alterations in two genes lead to cell death but alteration in only one of them is not lethal. SL interactions provide a new strategy for molecular-targeted cancer therapy. Currently, there are few drugs targeting SL interactions that entered into clinical trials. Therefore, it is necessary to investigate the link between SL interactions and drug sensitivity of cancer cells systematically for drug development purpose. We identified SL interactions by integrating the high-throughput data from The Cancer Genome Atlas, small hairpin RNA data and genetic interactions of yeast. By integrating SL interactions from other studies, we tested whether the SL pairs that consist of drug target genes and the genes with genomic alterations are related with drug sensitivity of cancer cells. We found that only 6.26%∼34.61% of SL interactions showed the expected significant drug sensitivity using the pooled cancer cell line data from different tissues, but the proportion increased significantly to approximately 90% using the cancer cell line data for each specific tissue. From an independent pharmacogenomics data of 41 breast cancer cell lines, we found three SL interactions (ABL1-IFI16, ABL1-SLC50A1 and ABL1-SYT11) showed significantly better prognosis for the patients with both genes being altered than the patients with only one gene being altered, which partially supports the SL effect between the gene pairs. Our study not only provides a new way for unraveling the complex mechanisms of drug sensitivity but also suggests numerous potentially important drug targets for cancer therapy.

Global view of a drug-sensitivity gene network.

An important challenge in drug development is to gain insight into the mechanism of drug sensitivity. Looking for insights into the global relationships between drugs and their sensitivity genes would be expected to reveal mechanism of drug sensitivity. Here we constructed a drug-sensitivity gene network (DSGN) based on the relationships between drugs and their sensitivity genes, using drug screened genomic data from the NCI-60 cell line panel, including 181 drugs and 1057 sensitivity genes, and 1646 associations between them. Through network analysis, we found that two drugs that share the same sensitivity genes tend to share the same Anatomical Therapeutic Chemical classification and side effects. We then found that the sensitivity genes of same drugs tend to cluster together in the human interactome and participate in the same biological function modules (pathways). Finally, we noticed that the sensitivity genes and target genes of the same drug have a significant dense distance in the human interactome network and they were functionally related. For example, target genes such as epidermal growth factor receptor gene can activate downstream sensitivity genes of the same drug in the PI3K/Akt pathway. Thus, the DSGN would provide great insights into the mechanism of drug sensitivity.

Establishment and proteomic characterization of patient-derived clear cell sarcoma xenografts and cell lines.

Clear cell sarcoma (CCS) is an aggressive mesenchymal malignancy characterized by the unique chimeric EWS-ATF1 fusion gene. Patient-derived cancer models are essential tools for the understanding of tumorigenesis and the development of anti-cancer drugs; however, only a limited number of CCS cell lines exist. The objective of this study was to establish patient-derived CCS models. We established patient-derived CCS models from a 43-yr-old female patient. We prepared the patient-derived xenografts (PDXs) from tumor tissues obtained through biopsy or surgery and isolated stable cell lines from PDXs and the original tumor tissue. The presence of gene fusions was examined by RT-PCR, and Sanger sequencing. The established cell lines were characterized by short tandem repeat, viability, colony and spheroid formation, and invasion analyses. Differences in gene enrichment between the primary tumor and cell lines were examined by mass spectrometry and KEGG pathway analysis. The cell lines were maintained for more than 80 passages, and had tumorigenic characteristics such as colony and spheroid formation and invasion. Mass spectrometric proteome analysis demonstrated that the cell lines were enriched for similar but distinct molecular pathways, compared to those in the xenografts and original tumor tissue. Next, tyrosine kinase inhibitors were screened for their suppressive effects on viability. We found that ponatinib, vandetanib, and doxorubicin suppressed the growth of cell lines, and had equivalent IC50 values. Further in-depth investigation and understanding of drug-sensitivity mechanisms will be important for the clinical applications of our cell lines.