Genomics Of Drug Sensitivity In Cancer Research Articles

Association between copy-number alteration of +20q, \u221214q and \u221218p and cross-sensitivity to tyrosine kinase inhibitors in clear-cell renal cell carcinoma

BackgroundWe aim to explore association between copy number alteration (CNA) and sensitivity to common tyrosine kinase inhibitors (TKIs) used in clear-cell renal cell carcinoma (ccRCC) treatment.MethodsCNA with related sensitivity profiles were extracted from the Genomics of Drug Sensitivity in Cancer (GDSC) dataset and was cross-referenced with common CNA in ccRCC in the Cancer Genome Atlas (TCGA) dataset. Functional annotation was profiled using GSEA and NET-GE. Target genes within cytobands of interest were screened in silico and validated in vitro using proliferation assays in A498 and 786-O ccRCC cells.ResultsFour TKIs (Sunitinib, Cabozantinib, Axitinib and Sorafenib) that were clinically used in ccRCC were selected. In silico analysis showed gain of 20q (+20q) occurred in ~ 23% of cases and was associated with resistance to all four TKIs; loss of 14q (−14q) occurred in ~ 39% of cases and was associated with resistance to Sunitinib and Sorafenib; loss of 18p (−18p) occurred in ~ 39% of cases and was associated with sensitivity to Sunitinib and Sorafenib. All 3 CNAs were associated with worsened prognosis, respectively. Candidate target genes included of RBL1 on 20q, KLHL33 on 14q and ARHGAP28 on18q. In vitro validation showed RBL1 overexpression induced resistance to Sunitinib and Cabozantinib; KLHL33 silencing induced resistance to Sunitinib; ARHGAP28 silencing induced sensitivity to Cabozantinib. Functional annotation indicated FoxO signaling, hypoxic response and Wnt pathway, and Rho-related cellular adhesion were mechanistically associated with +20q, −14q and −18p, respectively.ConclusionCommon CNAs in ccRCC are associated with cancer-intrinsic cross-sensitivity to common TKIs. Further validation and functional analyses are therefore needed.

BET, SRC, and BCL2 family inhibitors are synergistic drug combinations with PARP inhibitors in ovarian cancer.

BackgroundHomologous recombination deficiencies (HRD) are present in approximately half of epithelial ovarian cancers, for which PARP inhibitors (PARPi) are becoming a preferred treatment option. However, a considerable proportion of these carcinomas acquire resistance or harbour de novo resistance, posing a significant challenge to treatment.MethodsTo identify new combinatorial therapeutics to overcome resistance to PARPi, we employed high-throughput conditional RNAi and drug screening of patient-derived ovarian cancer cells. To prioritise clinically relevant drug combinations, we integrated empirical validation with analysis of The Cancer Genome Atlas (TCGA) and Genomics of Drug Sensitivity in Cancer (GDSC) datasets to nominate candidate targets and drugs, reaching three main findings.FindingsFirstly, we found that the PARPi rucaparib enhanced the effect of BET inhibitors (CPI-203 & CPI-0610) irrespective of clinical subtype or HRD status. Additional drug combination screens identified that dasatinib, a non-receptor tyrosine kinase inhibitor, augmented the effects of rucaparib and BET inhibitors, proposing a potential broadly applicable triple-drug combination for high-grade serous and clear cell ovarian carcinomas. Secondly, rucaparib synergised with the BCL2 family inhibitor navitoclax, with preferential activity in ovarian carcinomas that harbour alterations in BRCA1/2, BARD1, or MSH2/6. Thirdly, we identified potentially antagonistic drug combinations between the PARPi rucaparib and vinca alkaloids, anthracyclines, and antimetabolites, cautioning their use in the clinic.InterpretationThese findings propose therapeutic strategies to address PARP inhibitor resistance using agents that are already approved or are in clinical development, with the potential for rapid translation to benefit a broad population of ovarian cancer patients.

Comprehensive Analysis of Prognostic Alternative Splicing Signatures in Oral Squamous Cell Carcinoma.

BackgroundAlternative splicing (AS) plays an essential role in tumorigenesis and progression. This study aimed to develop a novel prognostic model based on the AS events to obtain more accurate survival prediction and search for potential therapeutic targets in oral squamous cell carcinoma (OSCC).MethodsSeven types of AS events in 326 OSCC patients with RNA-seq were obtained from the TCGA SpliceSeq tool and the TCGA database. Cox analysis, the least absolute shrinkage and selection operator Cox regression and random forest were employed to establish prognostic models. Genomics of Drug Sensitivity in Cancer (GDSC) was adopted to estimate the possible drug sensiticity. Prognostic splicing factor (SF)-AS network was constructed by Cytoscape.ResultsThe final model included 12 AS events, showing satisfactory performance. The area under the curve for 3- and 5-year survival in the training cohort was 0.83 and 0.82, respectively while that in internal validation was 0.83 and 0.82 accordingly. The calibration curve also indicated a satisfactory agreement between the observation and the predictive values. Low-risk patients stratified by the final model presented higher sensitivity to three chemo drugs. Besides, the prognostic SF-AS regulatory network contained five key SFs and 62 AS events.ConclusionsWe developed a powerful prognostic AS signature for OSCC and deepened the understanding of SF-AS network regulatory mechanisms. Low-risk patients tended to be more sensitive to the three chemo drugs while five key SFs including CELF2, TIA1, HNRNPC, HNRNPK, and SRSF9 were identified as potential prognostic biomarkers, which may offer new prospects for effective therapies of OSCC.

Abstract 1434: Loss of PTEN confers resistance to neddylation inhibition through TOP2A in glioblastoma

Abstract As a disease that has not seen lasting advances in its standard of care for decades, glioblastoma (GBM) calls for novel therapeutic approaches, and given GBM's heterogeneous and aggressive nature, precision medicine stands out as an appealing treatment modality for this tragic disease. Neddylation is a post-translational modification process that regulates protein function and turnover. MLN4924, or Pevonedistat, inhibits the E1 ligase Nedd8 activating enzyme (NAE) and, in doing so, the E1, E2, and E3 neddylation cascade. In its most researched role, NEDD-8 is conjugated to the cullin subunit of cullin-RING ligases (CRLs), a family of E3 ligases that facilitate the ubiquitination of different substrates. Thus, inhibition of neddylation decreases ubiquitination and degradation of particular substrates: e.g., p21, p27, wee-1, and other cell cycle regulatory and pro-apoptotic proteins. We and others have shown that neddylation stands out a promising therapeutic approach to cancer and GBM, given the vulnerability to MLN4924 in subset of different in vitro and in vivo models. In this work, we establish the selective vulnerability to MLN4924 in GBM and explore the biological mechanisms that underlie this differential response. Using the Cancer Cell Line Encyclopedia (CCLE) and the Genomics of Drug Sensitivity in Cancer (GDSC) databases, we found that mutations and deletions in phosphatase and tensin homolog (PTEN) are associated with resistance to MLN4924 in both low-grade gliomas and GBM. Our established models validated these findings, as our sensitive models – GB1 (IC50= 0.28 μM) and LN18 (IC50 = 0.19 μM), – expressed PTEN, while our resistant models – M059K (IC50 = 5.5 μM) and SNU1105 (IC50 = 20.9 μM) – did not. Knocking down PTEN with siRNA in our sensitive models increased their IC50 values three- to four-fold. This relationship was further validated in multiple isogenic systems. We sought to uncover whether this PTEN-status dependent resistance was contingent on its well-studied lipid phosphatase activity and PI3K/AKT signaling. While the resistant models had higher baseline levels of phosphorylated AKT (p-AKT), we found that MLN4924 treatment did not consistently alter p-AKT levels in our sensitive and/or resistant models. We also used a PTEN null isogenic glioma stem cell system (GSC23), in which wild-type PTEN and lipid phosphatase dead (G129A) PTEN was re-expressed, and found that both PTEN models sensitized GSC23 to MLN4924 treatment equally. While independent of the PI3K/AKT axis, we did find that there were significant increases in topoisomerase II alpha (TOP2A) levels in the post-treatment lysate of our resistant (PTEN-null) models. Targeting this increase in TOP2A expression, we found strong synergy with two different TOP2 poisons – etoposide and doxorubicin – and MLN4924 in a combination DDR. For the first time, we show that PTEN status not only serves as a novel biomarker for MLN4924 response but also reveals a vulnerability to TOP2A inhibitors when used in combination with MLN4924. Citation Format: Brett Taylor, Shayesteh Ferdosi, Nanyun Tang, Rita Bybee, George Reid, Lauren Hartman, Darren Finlay, Kristiina Vuori, Matthew Lee, Sen Peng, Frank Furnari, Harshil Dhruv, Michael Berens. Loss of PTEN confers resistance to neddylation inhibition through TOP2A in glioblastoma [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 1434.

Prognosisof Lung Adenocarcinoma Patients With NTRK3 Mutations to Immune Checkpoint Inhibitors.

BackgroundImmune checkpoint inhibitors (ICIs) are an important treatment modality that must be considered for patients with lung adenocarcinoma (LUAD). However, ICIs are effective only in some of these patients. Therefore, identifying biomarkers that accurately predict the prognosis of patients with LUAD treated with ICIs can help maximize their therapeutic benefits. This study aimed to identify a new potential predictor to better select and optimally benefit LUAD patients.MethodsWe first collected and analyzed a discovery immunotherapy cohort comprising clinical and mutation data for LUAD patients. Then, we evaluated whether the specific mutated genes can act as predictive biomarkers in this discovery immunotherapy cohort and further validated the findings in The Cancer Genome Atlas (TCGA) project LUAD cohort. Gene set enrichment analysis (GSEA) was used to explore possible alterations in DNA damage response (DDR) pathways within the gene mutation. Moreover, we analyzed whole-exome sequencing (WES) and drug sensitivity response data for LUAD cell lines in the Genomics of Drug Sensitivity in Cancer (GDSC) database.ResultsAmong the mutated genes screened from both the ICI treatment and TCGA-LUAD cohorts, NTRK3 mutation (mutant-type NTRK3, NTRK3-MT) was strongly associated with immunotherapy. First, significant differences in overall survival (OS) were observed between patients with NTRK3-MT and those with NTRK3-WT in the ICI treatment cohort but not in the non-ICI-treated TCGA-LUAD cohort. We then analyzed the association of NTRK3-MT with clinical characteristics and found the tumor mutation burden (TMB) to be significantly higher in both NTRK3-MT cohorts. However, significant differences in neoantigen levels and smoking history were found only for NTRK3-MT in the LUAD cohort from TCGA. Furthermore, some immune-related genes and immune cell-related genes were significantly upregulated in patients with NTRK3-MT compared to those with NTRK3-WT. In addition, NTRK3 mutation affected the deregulation of some signaling pathways and the DDR pathway.ConclusionsOur findings suggest that NTRK3-MT can predict the prognosis of patients with LUAD treated by ICIs and that it may have clinical significance for immunotherapy.

A prognostic signature based on immune-related genes for cervical squamous cell carcinoma and endocervical adenocarcinoma

Cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC) is the fourth commonest female malignancy worldwide. CESC progresses in immune-microenvironment mainly composed of infiltrating immune and stromal cells. Here, we performed an integrated analysis incorporating the expression profiles from the Cancer Genome Atlas (TCGA) database and scores of immune and stromal cells calculated by Estimation of Stromal and Immune cells in Malignant Tumours using Expression data (ESTIMATE) algorithm. A two-gene signature (CD1C and CD6 genes) was established to predict the prognosis of CESC. Based on this signature, patients were divided into the high- and low-risk groups, and this signature showed good prognostic performance according to the results of Kaplan-Meier analysis and receiver operating characteristic (ROC) analysis in train set and two validation sets. A nomogram was built for evaluating the clinical applicability of this signature. In addition, based on Tumor Immune Estimation Resource (TIMER) database, 2 hub genes showed negative correlations with tumor purity and positive correlations with infiltrating levels of immune filtrating cells. What’s more, we propose new treatment strategies for the two prognostic subtypes. Low- risk patients were found presenting with a higher level of immune checkpoint molecules and showing higher immunogenicity in immunophenoscore (IPS) analysis, which indicated a better response for immunotherapy. Meanwhile, estimated by Genomics of Drug Sensitivity in Cancer (GDSC) database, the high-risk patients showed sensitive responses to five chemotherapy drugs. Finally, 10 candidate small-molecule drugs for CESC were defined. In summary, the CD1C-CD6 signature can accurately predict the prognosis of CESC.

Computational Prediction of Drug Responses in Cancer Cell Lines From Cancer Omics and Detection of Drug Effectiveness Related Methylation Sites

Accurately predicting the response of a cancer patient to a therapeutic agent remains an important challenge in precision medicine. With the rise of data science, researchers have applied computational models to study the drug inhibition effects on cancers based on cancer genomics and transcriptomics. Moreover, a common epigenetic modification, DNA methylation, has been related to the occurrence and development of cancer, as well as drug effectiveness. Therefore, it is helpful for improvement of drug response prediction through exploring the relationship between DNA methylation and drug effectiveness. Here, we proposed a computational model to predict drug responses in cancers through integration of cancer genomics, transcriptomics, epigenomics, and compound chemical properties. Meanwhile, we applied a regularized regression model (Least Absolute Shrinkage and Selection Operator, lasso) to detect the methylation sites that were closely related to drug effectiveness. The prediction models were trained on a well-known pharmacogenomics data resource, Genomics of Drug Sensitivity in Cancer (GDSC). The cross-validation indicates that the performance of the prediction model using DNA methylation is comparable to that of using other cancer omics, including oncogene mutation and gene expression data. It indicates the important role of DNA methylation in prediction of drug responses. Encyclopedia of DNA Elements (ENCODE) and Transcriptional Regulatory Relationships Unraveled by Sentence-based Text mining (TRRUST2) database analyses suggest that the methylation sites associated with drug effectiveness are mainly located in the transcription factor (TF) binding region. Therefore, we hypothesized that the sensitivity of cancer cells to drugs could be regulated by changing the methylation modification of TF binding region. In conclusion, we confirmed the important role of DNA methylation in prediction of drug responses, and provided some methylation sites that closely related to the drug effectiveness, which may be a great regulatory target for improvement of drug treatment effects on cancer patients.

The critical role of BAP1 mutation in the prognosis and treatment selection of kidney renal clear cell carcinoma.

BackgroundThe BAP1 mutation is commonly found kidney renal clear cell carcinoma (KIRC) and a potential biomarker of individualized therapy. We evaluated the clinical significance of BAP1 mutation in the prognosis and treatment therapies for KIRC. Potential key pathways and related genes associated with these mechanisms were also identified in this investigation.MethodsWe identified the relevant data of patients BAP1 mutated on the cBioPortal and the compounds with significant selectivity to BAP1 mutations on the Genomics of Drug Sensitivity in Cancer (GDSC). And then, we identified the differences in mRNA expression levels of biological function annotation and pathways between mutated and wild type BAP1 patients by GSEA analysis. Furthermore, we screened the differentially expressed genes (DEGs) between BAP1 mutated and wild typed in KIRC patients and performed the GO and KEGG analysis. Finally, we conducted a protein-protein interaction (PPI) network to investigate the interaction between proteins encoded by candidate DEGs.ResultsReview of the TCGA data revealed 41 patients (10%) with KIRC displayed the BAP1 mutation. Further analysis led to the identification of 730 DEGs, while 617 genes were shown to be down-regulated, with 113 genes displaying upregulation. GO and KEGG pathway analysis indicated DEGs as enriched in metabolism, drug metabolism-cytochrome P450, and Drug-metabolizing enzymes. Subsequently, the top 10 hub genes, ranked by the degree in the PPI network were identified. Furthermore, our findings verify that the BAP1 mutation was associated with the deterioration of prognosis in patients with KIRC. Additionally, analysis of the GDSC database revealed that KIRC patients with BPP1 mutation are more prone to responding to Linsitinib.ConclusionsOur investigation identified the main pathways and relevant genes related to the BAP1 mutation in KIRC, which can contribute to the development of targeted treatment strategies for enhanced prognostic predictions of KIRC.

An Improved Anticancer Drug-Response Prediction Based on an Ensemble Method Integrating Matrix Completion and Ridge Regression

In this study, we proposed an ensemble learning method, simultaneously integrating a low-rank matrix completion model and a ridge regression model to predict anticancer drug response on cancer cell lines. The model was applied to two benchmark datasets, including the Cancer Cell Line Encyclopedia (CCLE) and the Genomics of Drug Sensitivity in Cancer (GDSC). As previous studies suggest, the dual-layer integrated cell line-drug network model was one of the best models by far and outperformed most state-of-the-art models. Thus, we performed a head-to-head comparison between the dual-layer integrated cell line-drug network model and our model by a 10-fold crossvalidation study. For the CCLE dataset, our model has a higher Pearson correlation coefficient between predicted and observed drug responses than that of the dual-layer integrated cell line-drug network model in 18 out of 23 drugs. For the GDSC dataset, our model is better in 26 out of 28 drugs in the phosphatidylinositol 3-kinase (PI3K) pathway and 26 out of 30 drugs in the extracellular signal-regulated kinase (ERK) signaling pathway, respectively. Based on the prediction results, we carried out two types of case studies, which further verified the effectiveness of the proposed model on the drug-response prediction. In addition, our model is more biologically interpretable than the compared method, since it explicitly outputs the genes involved in the prediction, which are enriched in functions, like transcription, Src homology 2/3 (SH2/3) domain, cell cycle, ATP binding, and zinc finger.

The application of artificial intelligence to drug sensitivity prediction

The development of computational methods for the prediction of effective therapeutic strategies based on the genomic information of patients is the main challenge of precision medicine. Since the 21st century, next-generation sequencing (NGS) has opened up new possibilities for personalized medicine. Extensive characterization at the molecular level for hundreds of cancer cell lines has been brought to the public eye by many organizations and agencies around the world. For example, the National Cancer Institute 60 Human Cancer Cell Line Screen (NCI-60), Cancer Cell Line Encyclopedia (CCLE) and Genomics of Drug Sensitivity in Cancer (GDSC) have provided large-scale omics data such as genomic, transcriptomic and epigenomic data characterizing cancer cell lines, and The Cancer Genome Atlas (TCGA) has molecularly characterized over 20000 primary cancers of patients. Combined with the drug response data of cancer cell lines, multiomics data could be used to analyse the mechanisms of action of anticancer drugs, which could be incorporated into precision medicine strategies. Over several decades, artificial intelligence (AI) technologies based on big data have revolutionized bioinformatics. AI has built a bridge between genomics and drug sensitivity by promoting the development of predictive models for the drug response of cancer cell lines. The 2012 NCI-DREAM drug prediction challenge has been particularly influential, as the innovative applications of machine learning that emerged from it have laid the groundwork for future studies. However, classic machine learning models are still challenging in terms of predictability because they limit the systematic integration of high-dimensional multiomics data. Therefore, network-based approaches, including link prediction and network representation, have become mainstream methods for drug response prediction. On the one hand, network-based approaches have not faced the “small n, large p” problem since the multiomics features are either represented in a gene/protein network or embedded in similarity networks between cell lines. On the other hand, the introduction of gene regulatory networks (GRNs) and protein-protein interactions (PPIs) into the predictive model can provide a functional background for the integration of genomic data and thereby improve the predictive performance of drug response. In addition to network-based approaches, multimodal deep learning models can systematically integrate multiomic data by considering them as different modalities. Generally, there are three feature fusion methods in deep neural networks: Input-level feature fusion (early fusion), intermediate feature fusion and decision-level fusion (late fusion). Intermediate feature fusion is predominant in drug response prediction studies, by which features are learned separately for each type of omics data and then integrated into one unified representation to be used as the input for a classifier or a regressor. Moreover, the features of drug structures can be used as a model to improve the performance. In brief, we summarize the characteristics of publicly accessible genomic databases and discuss the trends of artificial intelligence applications in drug sensitivity prediction for cancer cell lines, including machine learning, networks and multimodal deep neural networks.

Feature selection strategies for drug sensitivity prediction

Drug sensitivity prediction constitutes one of the main challenges in personalized medicine. Critically, the sensitivity of cancer cells to treatment depends on an unknown subset of a large number of biological features. Here, we compare standard, data-driven feature selection approaches to feature selection driven by prior knowledge of drug targets, target pathways, and gene expression signatures. We asses these methodologies on Genomics of Drug Sensitivity in Cancer (GDSC) dataset, evaluating 2484 unique models. For 23 drugs, better predictive performance is achieved when the features are selected according to prior knowledge of drug targets and pathways. The best correlation of observed and predicted response using the test set is achieved for Linifanib (r = 0.75). Extending the drug-dependent features with gene expression signatures yields the most predictive models for 60 drugs, with the best performing example of Dabrafenib. For many compounds, even a very small subset of drug-related features is highly predictive of drug sensitivity. Small feature sets selected using prior knowledge are more predictive for drugs targeting specific genes and pathways, while models with wider feature sets perform better for drugs affecting general cellular mechanisms. Appropriate feature selection strategies facilitate the development of interpretable models that are indicative for therapy design.

Multiscale omic assessment of EWSR1-NFATc2 fusion positive sarcomas to identify conserved fusion breakpoint and activation of the mTOR pathway.

e23536 Background: EWSR1-NFATc2 ( E-N) fusion positive sarcomas are rare cancers historically categorized as Ewing sarcomas. Emerging evidence suggests unique molecular characteristics and chemotherapy sensitivities in these cancers. Here we present the largest cohort of genomically profiled E-N fusion positive sarcomas and evaluate their breakpoints and putative pathway alterations. Methods: Comprehensive genomic profiling of 1,024 EWSR1 fusion positive sarcomas, including 14 E-N fusions, was obtained through the Foundation Medicine research database (FMI). Additional data from the Gene Expression Omnibus, the Genomics of Drug Sensitivity in Cancer (GDSC), and The Cancer Genome Atlas (TCGA) datasets were used. Descriptive statistics, principle component analysis, differential expression, and pathway analysis were used as appropriate. Results: E-N fusions identified in the FMI database demonstrated consistent breakpoints located between exon 2 and 3 on NFATc2; consistent with previous reports suggesting that the primary transactivation domain and regulatory domains of NFATc2 are conserved in the EWRS1-NFATc2 fusion. Cluster analysis demonstrated separate E-N fusion positive sarcoma grouping than other EWRS1 sarcomas, while identifying an enrichment of mTOR pathway variants in E-N fusion positive samples (p = 0.05). Gene expression data from E-N identified a significant activation of the mTOR pathway in E-N positive sarcomas compared to either EWSR1-ETS (p = 0.002) or CIC-DUX4 positive sarcomas (p < 0.001). In the GDSC database, NFATc2-High lines exhibited activation of the mTOR pathway (p < 0.0001) and were more sensitive to the mTOR inhibitor rapamycin (p = 0.0006). Pan-cancer analysis of 33 non-overlapping TCGA datasets, elevated NFATc2 mRNA expression was significantly associated with activated RICTOR and mTOR signaling (p < 0.0001). Elevated NFATc2 correlated with higher tumor grade and poorer progression-free survival and overall survival. Conclusions: Taken together, the data presented here suggests that E-N fusion positive sarcomas represent a distinct molecularly entity in comparison to traditional Ewing sarcomas. Conserved action of NFATc2 may correlate to mTOR over activation.

YM155-Adapted Cancer Cell Lines Reveal Drug-Induced Heterogeneity and Enable the Identification of Biomarker Candidates for the Acquired Resistance Setting.

Survivin is a drug target and its suppressant YM155 a drug candidate mainly investigated for high-risk neuroblastoma. Findings from one YM155-adapted subline of the neuroblastoma cell line UKF-NB-3 had suggested that increased ABCB1 (mediates YM155 efflux) levels, decreased SLC35F2 (mediates YM155 uptake) levels, decreased survivin levels, and TP53 mutations indicate YM155 resistance. Here, the investigation of 10 additional YM155-adapted UKF-NB-3 sublines only confirmed the roles of ABCB1 and SLC35F2. However, cellular ABCB1 and SLC35F2 levels did not indicate YM155 sensitivity in YM155-naïve cells, as indicated by drug response data derived from the Cancer Therapeutics Response Portal (CTRP) and the Genomics of Drug Sensitivity in Cancer (GDSC) databases. Moreover, the resistant sublines were characterized by a remarkable heterogeneity. Only seven sublines developed on-target resistance as indicated by resistance to RNAi-mediated survivin depletion. The sublines also varied in their response to other anti-cancer drugs. In conclusion, cancer cell populations of limited intrinsic heterogeneity can develop various resistance phenotypes in response to treatment. Therefore, individualized therapies will require monitoring of cancer cell evolution in response to treatment. Moreover, biomarkers can indicate resistance formation in the acquired resistance setting, even when they are not predictive in the intrinsic resistance setting.

Gene regulatory network analysis with drug sensitivity reveals synergistic effects of combinatory chemotherapy in gastric cancer

The combination of docetaxel, cisplatin, and fluorouracil (DCF) is highly synergistic in advanced gastric cancer. We aimed to explain these synergistic effects at the molecular level. Thus, we constructed a weighted correlation network using the differentially expressed genes between Stage I and IV gastric cancer based on The Cancer Genome Atlas (TCGA), and three modules were derived. Next, we investigated the correlation between the eigengene of the expression of the gene network modules and the chemotherapeutic drug response to DCF from the Genomics of Drug Sensitivity in Cancer (GDSC) database. The three modules were associated with functions related to cell migration, angiogenesis, and the immune response. The eigengenes of the three modules had a high correlation with DCF (−0.41, −0.40, and −0.15). The eigengenes of the three modules tended to increase as the stage increased. Advanced gastric cancer was affected by the interaction the among modules with three functions, namely cell migration, angiogenesis, and the immune response, all of which are related to metastasis. The weighted correlation network analysis model proved the complementary effects of DCF at the molecular level and thus, could be used as a unique methodology to determine the optimal combination of chemotherapy drugs for patients with gastric cancer.

Ensembled machine learning framework for drug sensitivity prediction

Drug sensitivity prediction is one of the critical tasks involved in drug designing and discovery. Recently several online databases and consortiums have contributed to providing open access to pharmacogenomic data. These databases have helped in developing computational approaches for drug sensitivity prediction. Cancer is a complex disease involving the heterogeneous behaviour of same tumour-type patients towards the same kind of drug therapy. Several methods have been proposed in the literature to predict drug sensitivity. However, these methods are not efficient enough to predict drug sensitivity. The present study has proposed an ensemble learning framework for drug-response prediction using a modified rotation forest. The proposed framework is further compared with three state-of-the-art algorithms and two baseline methods using Genomics of Drug Sensitivity in Cancer (GDSC) and Cancer Cell Line Encyclopedia (CCLE) drug screens. The authors have also predicted missing drug response values in the data set using the proposed approach. The proposed approach outperforms other counterparts even though gene mutation data is not incorporated while designing the approach. An average mean square error of 3.14 and 0.404 is achieved using GDSC and CCLE drug screens, respectively. The obtained results show that the proposed framework has considerable potential to improve anti-cancer drug response prediction.

Alterations in TP53 Are a Potential Biomarker of Bladder Cancer Patients Who Benefit From Immune Checkpoint Inhibition.

In recent years, immune checkpoint inhibitors (ICIs) targeting CTLA-4 or PD1/PDL1 have achieved remarkable success in the treatment of bladder cancer (BLCA), but only a few patients have shown durable clinical benefits. The prognostic role of a mutant form of the tumor suppressor gene TP53 (TP53-MT) in predicting the efficacy of ICIs is highly controversial; therefore, in this study, we obtained data for 210 patients from an immunotherapy cohort, 412 patients from The Cancer Genome Atlas (TCGA)-BLCA cohort and 18 BLCA cell lines from Genomics of Drug Sensitivity in Cancer (GDSC), and we performed integrated bioinformatic analysis to explore the relationships between TP53-MT and clinical benefits derived from ICI treatment and the underlying mechanisms. We conclude that TP53-MT is a potential indicator of a relatively good response to ICIs and associated with prolonged overall survival (OS) (log-rank test, hazard ratio (HR) = 0.65 [95% confidence interval (CI), 0.44-0.99], p = 0.041). Through integrated analysis with several platforms, we found that TP53-MT patients were more likely to benefit from ICIs than wild-type P53 (TP53-WT) patients, which may be the result of 2 major mechanisms. First, the patients with TP53-MT showed stronger tumor antigenicity and tumor antigen presentation, as indicated by a higher tumor mutational load, a higher neoantigen load and increased expression of MHC; second, the antitumor immunity preexisting in tumors was stronger in samples with TP53-MT than in those with TP53-WT, including enrichment of interferon-gamma, positive regulation of TNF secretion pathways and increased expression of some immunostimulatory molecules, such as CXCL9 and CXCL10. This study provided some clues for identifying patients who would potentially benefit from ICIs at the somatic genomic level, developing new indications for targeted second-generation sequencing and promoting the development of precision medicine.

In silico Prediction Followed By in Vitro validation Identifies the Survivin Inhibitor YM155 As a Potent Secondary Drug Against Pi-Resistant Myeloma

Drug resistance is a major cause of concern in cancer chemotherapy that not only reduces efficacy but also causes toxicity through overdosing. Multiple myeloma (MM) is the second-most common haematological malignancy in the USA that remains a challenging disease to cure due to the presence of significant complexity and heterogeneity at molecular level with high treatment cost. Proteasome inhibitor, the standard of care drug, has proven to remarkably improve myeloma patient outcomes but is associated with dose limiting toxicities and drug resistance. Therefore, new combination treatment regimens are urgently needed for the treatment of PI-resistant MM that can lower the required dose of standard-of-care drug without compromising on treatment efficacy. We have created an optimization-regularization based computational prediction method called secDrug that uses large-scale pharmacogenomics databases to identify novel secondary drugs against PI-resistance when used as single agent or in combination. We applied secDrug to PIs-resistant cell lines (>100) in the Genomics of Drug Sensitivity in Cancer (GDSC) database and predicted the top drugs that can be best combined with PIs to overcome PI-resistance in MM. These include: Survivin inhibitor (YM155), Nicotinamide phospho ribosyl transferase or Nampt inhibitor (FK866), PIKfyve inhibitor (YM201636), Raf inhibitor (PLX-4720), Bcl2 inhibitor (Navitoclax),HSP90 inhibitor (17-AAG), AKT inhibitor (KIN00102), HDAC inhibitors (Panobinostat, SAHA), S6K1-specific inhibitor (PF-4708671), and the neddylation inhibitor (MLN4924). To validate our prediction results, we treated human multiple myeloma cell lines (HMCLs) highly resistant to the proteasome inhibitors Bortezomib, Carfilzomob, Oprozomib and Ixazomib and >10 clonally derived acquired resistant cells with the top predicted best secondary drugs. Next, we treated three pairs of parental PI-sensitive (P) and clonally-derived PI-resistant (R) cell lines from PI-sensitive (P) HMCLs representing acquired PI-response with the top predicted drug combinations. Our preliminary in vitro data confirmed our in silico predictions of secondary drugs based on secDrug analysis. Among these, YM155 and YM155 + Ixazomib showed strikingly very high in vitro cytotoxicity against PI- resistant MM cell lines (Figure 1). The results, based on Chou-Talalay's combination index (CI) theorem, show that this combination works highly synergistically where the IC50 of ixazomib in MM cells was significantly reduced in presence of YM155. Similar results were obtained for the other top predicted combination regimens. Our findings provide a strong case for combining YM155 with ixazomib to enhance sensitivity or overcome resistance to ixazomib and thereby improve patient outcome. Further, this study provides additional support towards using secDrug for predicting secondary drugs in cancers resistant to the standard-of-care therapy. Currently, we are exploring the molecular mechanisms behind the synergism between proteasome inhibitors and the top drugs identified by our algorithm as candidates for secondary combination therapies using next-gen transcriptomics and single-cell analysis approaches. Earlier studies have shown that YM155 potentially induces selective cell death in human pluripotent stem cells. Further, the presence of cancer stem cells sub population/ MM-CSCs in drug-resistant tumors have been shown to significantly contribute to the development of PI-resistance in MM. Hence, we propose that YM155 reverses PI-resistance by reducing the stem cell load in multiple myeloma cell line model system. Therefore, our future plan is to assess the effects of YM155 on MM-cancer stem cells like subclones/ sub population/ MM-CSCs including CD19- CD138- quiescent stem cells, ALDH+ and side populations/SP) in drug-resistant tumors. Overall, our in vitro validation results corroborated with our in silico prediction of secondary drugs. Disclosures No relevant conflicts of interest to declare.

Network-based Biased Tree Ensembles (NetBiTE) for Drug Sensitivity Prediction and Drug Sensitivity Biomarker Identification in Cancer

We present the Network-based Biased Tree Ensembles (NetBiTE) method for drug sensitivity prediction and drug sensitivity biomarker identification in cancer using a combination of prior knowledge and gene expression data. Our devised method consists of a biased tree ensemble that is built according to a probabilistic bias weight distribution. The bias weight distribution is obtained from the assignment of high weights to the drug targets and propagating the assigned weights over a protein-protein interaction network such as STRING. The propagation of weights, defines neighborhoods of influence around the drug targets and as such simulates the spread of perturbations within the cell, following drug administration. Using a synthetic dataset, we showcase how application of biased tree ensembles (BiTE) results in significant accuracy gains at a much lower computational cost compared to the unbiased random forests (RF) algorithm. We then apply NetBiTE to the Genomics of Drug Sensitivity in Cancer (GDSC) dataset and demonstrate that NetBiTE outperforms RF in predicting IC50 drug sensitivity, only for drugs that target membrane receptor pathways (MRPs): RTK, EGFR and IGFR signaling pathways. We propose based on the NetBiTE results, that for drugs that inhibit MRPs, the expression of target genes prior to drug administration is a biomarker for IC50 drug sensitivity following drug administration. We further verify and reinforce this proposition through control studies on, PI3K/MTOR signaling pathway inhibitors, a drug category that does not target MRPs, and through assignment of dummy targets to MRP inhibiting drugs and investigating the variation in NetBiTE accuracy.

Nonlinear mixed-effects models for modeling in vitro drug response data to determine problematic cancer cell lines

Cancer cell lines (CCLs) have been widely used to study of cancer. Recent studies have called into question the reliability of data collected on CCLs. Hence, we set out to determine CCLs that tend to be overly sensitive or resistant to a majority of drugs utilizing a nonlinear mixed-effects (NLME) modeling framework. Using drug response data collected in the Cancer Cell Line Encyclopedia (CCLE) and the Genomics of Drug Sensitivity in Cancer (GDSC), we determined the optimal functional form for each drug. Then, a NLME model was fit to the drug response data, with the estimated random effects used to determine sensitive or resistant CCLs. Out of the roughly 500 CCLs studies from the CCLE, we found 17 cell lines to be overly sensitive or resistant to the studied drugs. In the GDSC, we found 15 out of the 990 CCLs to be excessively sensitive or resistant. These results can inform researchers in the selection of CCLs to include in drug studies. Additionally, this study illustrates the need for assessing the dose-response functional form and the use of NLME models to achieve more stable estimates of drug response parameters.

P1.14-41 The Combination of the PARP Inhibitor Olaparib and the ATR Inhibitor VE-821 Selectively Targets ATM-Deficient Lung Cancer Cells

The driving principle behind precision medicine is to specifically target genetic variations that arise in tumorigenesis while leaving normal cells unaffected. Mutations in Ataxia Telangiectasia Mutated (ATM) may offer such a therapeutic target. ATM is mutated in approximately 12% of lung cancers and up to 40% of lung adenocarcinoma have been reported to lack ATM protein expression. ATM is an apex signaling kinase that responds to DNA-double strand breaks, playing a direct role in DNA repair as well as the initiation of signaling cascades that can lead to cell cycle arrest and apoptosis. We asked whether ATM-deficient human lung cancer cells are sensitive to the poly-ADP ribose polymerase (PARP) inhibitor olaparib, and investigated the mechanism of action of olaparib in these cells. We analyzed drug sensitivity for 61 lung adenocarcinoma cell lines from the Genomics of Drug Sensitivity in Cancer (GDSC) project and deleted ATM from lung adenocarcinoma A549 cells using CRISPR/Cas9. We determined the effects on cell viability using trypan blue exclusion and clonogenic survival assays. To investigate the mechanism of sensitivity of ATM-deficient cells to PARP and ATR inhibitors we used flow cytometry and cell viability assays as above. We observed a positive correlation between olaparib IC50 values and ATM mRNA expression. ATM mutant cell lines were more sensitive to olaparib compared to ATM wild-type cell lines or cell lines with amplified ATM. Additionally, ATM-deficient lung cancer cells were sensitive to olaparib, as are lung cancer cells (A549) with CRISPR/Cas9 deletion of ATM. Mechanistically, olaparib caused the temporary and reversible accumulation of G2 phase cells in ATM-deficient cells which manifested as a decrease in proliferation in both the trypan blue exclusion assay and clonogenic survival assay. Olaparib did not induce cell death in ATM-deficient cells, however cell death was induced when olaparib was used in combination with the ATR inhibitor VE-821. We show that olaparib acts as a cytostatic agent in ATM-deficient lung cancer cells, inducing a reversible and temporary growth arrest in G2 phase. Only when combined with the ATR inhibitor VE-821 was cell death observed and only in ATM-deficient cells. Our data suggest that patients with ATM-deficient lung cancer could benefit from combinatorial treatment with PARP and ATR inhibitors.