Search

Purpose:There are several agents in early clinical trials targeting components of the adenosine pathway including A2AR and CD73. The identification of cancers with a significant adenosine drive is critical to understand the potential for these molecules. However, it is challenging to measure tumor adenosine levels at scale, thus novel, clinically tractable biomarkers are needed.Experimental Design:We generated a gene expression signature for the adenosine signaling using regulatory networks derived from the literature and validated this in patients. We applied the signature to large cohorts of disease from The Cancer Genome Atlas (TCGA) and cohorts of immune checkpoint inhibitor–treated patients.Results:The signature captures baseline adenosine levels in vivo (r2 = 0.92, P = 0.018), is reduced after small-molecule inhibition of A2AR in mice (r2 = −0.62, P = 0.001) and humans (reduction in 5 of 7 patients, 70%), and is abrogated after A2AR knockout. Analysis of TCGA confirms a negative association between adenosine and overall survival (OS, HR = 0.6, P < 2.2e–16) as well as progression-free survival (PFS, HR = 0.77, P = 0.0000006). Further, adenosine signaling is associated with reduced OS (HR = 0.47, P < 2.2e–16) and PFS (HR = 0.65, P = 0.0000002) in CD8+ T-cell–infiltrated tumors. Mutation of TGFβ superfamily members is associated with enhanced adenosine signaling and worse OS (HR = 0.43, P < 2.2e–16). Finally, adenosine signaling is associated with reduced efficacy of anti-PD1 therapy in published cohorts (HR = 0.29, P = 0.00012).Conclusions:These data support the adenosine pathway as a mediator of a successful antitumor immune response, demonstrate the prognostic potential of the signature for immunotherapy, and inform patient selection strategies for adenosine pathway modulators currently in development.

Resistance to EGFR inhibitors (EGFRi) presents a major obstacle in treating non-small cell lung cancer (NSCLC). One of the most exciting new ways to find potential resistance markers involves running functional genetic screens, such as CRISPR, followed by manual triage of significantly enriched genes. This triage process to identify ‘high value’ hits resulting from the CRISPR screen involves significant manual curation that requires specialized knowledge and can take even experts several months to comprehensively complete.To find key drivers of resistance faster we built a hybrid recommendation system on top of a heterogeneous biomedical knowledge graph integrating preclinical, clinical, and literature evidence. Genes were ranked based on trade-offs between diverse types of evidence linking them to potential mechanisms of EGFRi resistance. This unbiased approach identified 36 resistance markers from >3,000 genes, reducing hit identification time from months to minutes. In addition to reproducing known resistance markers, our method identified novel resistance mechanisms that we prospectively validated.

Immuno-oncology (IO) therapies have transformed the therapeutic landscape of non-small cell lung cancer (NSCLC). However, patient responses to IO are variable and influenced by a heterogeneous combination of health, immune, and tumor factors. There is a pressing need to discover the distinct NSCLC subgroups that influence response. We have developed a deep patient graph convolutional network, we call “DeePaN”, to discover NSCLC complexity across data modalities impacting IO benefit. DeePaN employs high-dimensional data derived from both real-world evidence (RWE)-based electronic health records (EHRs) and genomics across 1937 IO-treated NSCLC patients. DeePaN demonstrated effectiveness to stratify patients into subgroups with significantly different (P-value of 2.2 × 10−11) overall median survival of 20.35 months and 9.42 months post-IO therapy. Significant differences in IO outcome were not seen from multiple non-graph-based unsupervised methods. Furthermore, we demonstrate that patient stratification from DeePaN has the potential to augment the emerging IO biomarker of tumor mutation burden (TMB). Characterization of the subgroups discovered by DeePaN indicates potential to inform IO therapeutic insight, including the enrichment of mutated KRAS and high blood monocyte count in the IO beneficial and IO non-beneficial subgroups, respectively. Our work has proven the concept that graph-based AI is feasible and can effectively integrate high-dimensional genomic and EHR data to meaningfully stratify cancer patients on distinct clinical outcomes, with potential to inform precision oncology.

Summary High-throughput drug screens in cancer cell lines test compounds at low concentrations, thereby enabling the identification of drug-sensitivity biomarkers, while resistance biomarkers remain underexplored. Dissecting meaningful drug responses at high concentrations is challenging due to cytotoxicity, i.e., off-target effects, thus limiting resistance biomarker discovery to frequently mutated cancer genes. To address this, we interrogate subpopulations carrying sensitivity biomarkers and consecutively investigate unexpectedly resistant (UNRES) cell lines for unique genetic alterations that may drive resistance. By analyzing the GDSC and CTRP datasets, we find 53 and 35 UNRES cases, respectively. For 24 and 28 of them, we highlight putative resistance biomarkers. We find clinically relevant cases such as EGFRT790M mutation in NCI-H1975 or PTEN loss in NCI-H1650 cells, in lung adenocarcinoma treated with EGFR inhibitors. Interrogating the underpinnings of drug resistance with publicly available CRISPR phenotypic assays assists in prioritizing resistance drivers, offering hypotheses for drug combinations., Graphical Abstract, Highlights • Presenting a strategy for identifying rare drug-resistance markers in cancer cells • Effectively stratifying drug resistance from cytotoxicity in pharmacology screens • Generating hypotheses for resistance mechanisms underlying certain biomarkers • CRISPR screens support resistance hypotheses and shed light on cancer vulnerabilities, The Bigger Picture Cancer drug resistance is the major challenge of modern oncology. Identifying resistance and its biomarkers will empower the next generation of precision medicines. High-throughput pharmacology screens in cancer cell lines have successfully identified drug-sensitivity biomarkers, but drug-resistance biomarkers are underexplored. Intrinsic drug-resistance events are often rare and experimentally indistinguishable from cytotoxicity or artifacts without prior knowledge. To address this, we investigate cell-line populations sensitized to a drug treatment (i.e., carrying established sensitivity biomarkers) and characterize those cell lines that do not respond as expected. We highlight unique genetic features harbored by these cell lines and confirm their linkage to drug resistance using CRISPR gene essentiality data. Our analysis and results pave the way for enhanced precision medicine, guide further CRISPR screens, and identify potential drug combinations to tackle resistance., Identifying cancer drug resistance and its biomarkers will empower the next generation of anti-cancer medicines, tailoring treatments to individual patients. Detecting drug resistance in high-throughput pharmacology screens is experimentally challenging. We present a computational framework identifying rare intrinsically resistant cancer cell lines. Our observations provide hypotheses for associated drug-resistance biomarkers, which we validate with independent CRISPR essentiality screens. Our results pave the way for enhancing cancer precision medicine and effective drug combinations to overcome resistance.

Targeted mono-therapies in cancer are hampered by the ability of tumor cells to escape inhibition through rewiring or alternative pathways. Drug combination approaches can provide a means to overcome these resistance mechanisms. Effective use of combinations requires strategies to select combinations from the enormous space of combinations, and to stratify patients according to their likelihood to respond. We here introduce two complementary workflows: One prioritising experiments in high-throughput screens for drug synergy enrichment, and a consecutive workflow to predict hypothesis-driven synergy stratification. Both approaches only need data of efficacy of single drugs. They rely on the notion of target functional similarity between two target proteins. This notion reflects how similarly effective drugs are on different cancer cells as a function of cancer signaling pathways’ activities on those cells. Our synergy prediction workflow revealed that two drugs targeting either the same or functionally opposite pathways are more likely to be synergistic. This enables experimental prioritisation in high-throughput screens and supports the notion that synergy can be achieved by either redundant pathway inhibition or targeting independent compensatory mechanisms. We tested the synergy stratification workflow on seven target protein pairs (AKT/EGFR, AKT/MTOR, BCL2/MTOR, EGFR/MTOR, AKT/BCL2, AKT/ALK and AKT/PARP1, representing 29 combinations and predicted their synergies in 33 breast cancer cell lines (Pearson’s correlation r=0.27). Additionally, we experimentally validated predicted synergy of the BRAF/Insulin Receptor combination (Dabrafenib/BMS−754807) in 48 colorectal cancer cell lines (r=0.5). In conclusion, our synergy prediction workflow can support compound prioritization in large scale drug screenings, and our synergy stratification workflow can select where the efficacy of drugs already known for inducing synergy is higher.

Patient-derived tumor xenograft (PDX) mouse models have emerged as an important oncology research platform to study tumor evolution, mechanisms of drug response and resistance, and tailoring chemotherapeutic approaches for individual patients. The lack of robust standards for reporting on PDX models has hampered the ability of researchers to find relevant PDX models and associated data. Here we present the PDX models minimal information standard (PDX-MI) for reporting on the generation, quality assurance, and use of PDX models. PDX-MI defines the minimal information for describing the clinical attributes of a patient's tumor, the processes of implantation and passaging of tumors in a host mouse strain, quality assurance methods, and the use of PDX models in cancer research. Adherence to PDX-MI standards will facilitate accurate search results for oncology models and their associated data across distributed repository databases and promote reproducibility in research studies using these models. Cancer Res; 77(21); e62–66. ©2017 AACR.

Current understanding of the mutation spectrum of relapsed/refractory (RR) tumors is limited. We performed whole exome sequencing (WES) on 47 diffuse large B cell lymphoma (DLBCL) tumors that persisted after R-CHOP treatment, 8 matched to primary biopsies. We compared genomic alterations from the RR cohort against two treatment-naïve DLBCL cohorts (n=112). While the overall number and types of mutations did not differ significantly, we identified frequency changes in DLBCL driver genes. The overall frequency of MYD88 mutant samples increased (12% to 19%), but we noted a decrease in p.L265P (8% to 4%) and increase in p.S219C mutations (2% to 6%). CARD11 p.D230N, PIM1 p.K115N and CD79B p.Y196C mutations were not observed in the RR cohort, although these mutations were prominent in the primary DLBCL samples. We observed an increase in BCL2 mutations (21% to 38% of samples), BCL2 amplifications (3% to 6% of samples) and CREBBP mutations (31% to 42% of samples) in the RR cohort, supported by acquisition of mutations in these genes in relapsed compared to diagnostic biopsies from the same patient. These increases may reflect the genetic characteristics of R-CHOP RR tumors expected to be enriched for during clinical trial enrollment. These findings hold significance for a number of emerging targeted therapies aligned to genetic targets and biomarkers in DLBCL, reinforcing the importance of time-of-treatment biomarker screening during DLBCL therapy selection.

Precision medicine is empowered by biomarker discovery. High-throughput drug screens in cancer cell lines enable the identification of drug sensitivity biomarkers, whilst resistance biomarkers remain underexplored: dissecting meaningful drug responses at high concentrations is impossible due to cytotoxicity, making non-responders and resistant cell lines indistinguishable, thus limiting resistance biomarker discovery to frequently mutated cancer genes. In order to address this, we developed a statistical method to indirectly identify resistance biomarkers. Here, we interrogate subpopulations carrying sensitivity biomarkers, and consecutively investigate resistant outliers for unique genetic alterations that putatively drive this resistance. Out of 78 cases with resistant outliers, our method predicts infrequent resistance biomarkers for 43 of them, e.g. the clinically relevant EGFRT790M mutation in lung adenocarcinoma treated with EGFR inhibitors. Interrogating the underpinnings of drug resistance with publicly available CRISPR phenotypic assays assists in prioritising resistance drivers, offering hypotheses for potentially effective drug combinations, and thereby paving the way for the next generation of precision medicine.

Immuno-oncology (IO) therapies have transformed the therapeutic landscape of non-small cell lung cancer (NSCLC). However, patient responses to IO are variable and influenced by a heterogeneous combination of health, immune and tumor factors. There is a pressing need to discover the distinct NSCLC subgroups that influence response. We have developed a deep patient graph convolutional network, we call “DeePaN”, to discover NSCLC complexity across data modalities impacting IO benefit. DeePaN employs high-dimensional data derived from both real world evidence (RWE) based electronic health records (EHRs) and genomics across 1,937 IO treated NSCLC patients. DeePaN demonstrated effectiveness to stratify patients into subgroups with significantly different (p-value of 2.2 × 10−11) overall survival of 20.35 months and 9.42 months post-IO therapy. Significant differences in IO outcome were not seen from multiple non-graph based unsupervised methods. Furthermore, we demonstrate that patient stratification from DeePaN has the potential to augment the emerging IO biomarker of tumor mutation burden (TMB). Characterization of the subgroups discovered by DeePaN indicates potential to inform IO therapeutic insight, including the enrichment of mutated KRAS and high blood monocyte count in the IO beneficial and IO non-beneficial subgroups, respectively. To the best of our knowledge, our work for the first time has proven the concept that graph based AI is feasible and can effectively integrate high-dimensional genomic and EHR data to meaningfully stratify cancer patients on distinct clinical outcomes, with potential to inform precision oncology.

BRAF and MEK1/2 inhibitors are effective in melanoma but resistance inevitably develops. Despite increasing the abundance of pro-apoptotic BIM and BMF, ERK1/2 pathway inhibition is predominantly cytostatic, reflecting residual pro-survival BCL2 family activity. Here, we show that uniquely low BCL-XL expression in melanoma biases the pro-survival pool towards MCL1. Consequently, BRAF or MEK1/2 inhibitors are synthetic lethal with the MCL1 inhibitor AZD5991, driving profound tumour cell death that requires BAK/BAX, BIM and BMF, and inhibiting tumour growth in vivo. Combination of ERK1/2 pathway inhibitors with BCL2/BCL-w/BCL-XL inhibitors is stronger in CRC, correlating with a low MCL1:BCL-XL ratio; indeed the MCL1:BCL-XL ratio is predictive of ERK1/2 pathway inhibitor synergy with MCL1 or BCL2/BCL-w/BCL-XL inhibitors. Finally, AZD5991 delays acquired BRAFi/MEKi resistance and enhances the efficacy of an ERK1/2 inhibitor in a model of acquired BRAFi + MEKi resistance. Thus combining ERK1/2 pathway inhibitors with MCL1 antagonists in melanoma could improve therapeutic index and patient outcomes., BRAF or MEK1/2 inhibitors are cytostatic in melanoma and the surviving cells develop drug resistance. This study shows that the pro-survival pool is biased towards MCL1 in melanoma so that BRAF or MEK1/2 inhibitors are synthetic lethal with the MCL1 inhibitor AZD5991, improving tumour growth inhibition.

Here we present a machine learning model, Deep Purity (DePuty) that leverages convolutional neural networks to accurately predict tumor purity from next-generation sequencing data from clinical samples without matched normals. As input, our model utilizes SNP-based copy number and minor allele frequency data formulated as a scatterplot image. With a representation matching that used by expert human annotators, we best an existing algorithm using only ~100 manually curated samples. Our simple, data-efficient approach can serve as a straightforward alternative to traditional, more complex statistical methods, for building performant purity prediction models that enable downstream bioinformatic analysis of tumor variants and absolute copy number alterations relevant to cancer genomics.

Acquired resistance to MEK1/2 inhibitors (MEKi) arises through amplification of BRAFV600E or KRASG13D to reinstate ERK1/2 signalling. Here we show that BRAFV600E amplification and MEKi resistance are reversible following drug withdrawal. Cells with BRAFV600E amplification are addicted to MEKi to maintain a precise level of ERK1/2 signalling that is optimal for cell proliferation and survival, and tumour growth in vivo. Robust ERK1/2 activation following MEKi withdrawal drives a p57KIP2-dependent G1 cell cycle arrest and senescence or expression of NOXA and cell death, selecting against those cells with amplified BRAFV600E. p57KIP2 expression is required for loss of BRAFV600E amplification and reversal of MEKi resistance. Thus, BRAFV600E amplification confers a selective disadvantage during drug withdrawal, validating intermittent dosing to forestall resistance. In contrast, resistance driven by KRASG13D amplification is not reversible; rather ERK1/2 hyperactivation drives ZEB1-dependent epithelial-to-mesenchymal transition and chemoresistance, arguing strongly against the use of drug holidays in cases of KRASG13D amplification., Colorectal cancer cells can acquire resistance to MEK inhibition due to BRAF or KRAS amplification. Here, the authors show that while MEK inhibitor withdrawal in BRAF mutant cells restores sensitivity to the inhibitor through the loss of BRAF amplification mediated by a p57-dependent mechanism, drug withdrawal from KRAS mutant cells does not restore sensitivity but results in EMT and chemoresistance.

Genomic profiling of tumours in patients in clinical trials enables rapid testing of multiple hypotheses to confirm which genomic events determine likely responder groups for targeted agents. A key challenge of this new capability is defining which specific genomic events should be classified as 'actionable' (that is, potentially responsive to a targeted therapy), especially when looking for early indications of patient subgroups likely to be responsive to new drugs. This Opinion article discusses some of the different approaches being taken in early clinical development to define actionable mutations, and describes our strategy to address this challenge in early-stage exploratory clinical trials.

The Cancer Dependency Map Portal enables combined analyses of genomic and phenomic data such as genomic alterations and CRISPR gene dependencies but focuses on single genes or cell lines. This is well suited for verifying previously generated hypotheses such as the dependency of cells on PRMT5 in the setting of loss of function mutations in MTAP. To identify additional actionable de-novo vulnerabilities in-silico, we refined the genomic status of 1000 cell lines by 1. focusing on variants (driver mutations and copy number alterations) that have higher prevalence in a given cancer type in TCGA and 2. reclassifying variants as actionable only if they cause loss of function in known tumour suppressor genes or gain of function in oncogenes. This gives a clear line of sight, a more manageable set of variants and is analogous to Mendelian randomization. This approach was applied to lung squamous cell carcinoma to form a binary event matrix for the shortlisted genes and copy number alterations across a panel of lung squamous cell lines. For each of the variants, we used machine learning approaches to define a dependency signature from CRISPR dependencies for the matching cell lines that distinguish altered vs non-altered cell lines. The top ranking dependencies were further examined for biological relevance and potential toxicity, yielding novel targets to follow up on. The approach generalises well and can be applied to any other cancers with unmet needs. All data in this study was derived from public databases. By combining TCGA cancer specific prevalent genomic alterations in tumor suppressor and oncogenes we drew clear line of sight and associated these alterations to CRISPR based vulnerabilities. This yielded a manageable set of alterations, and for each of these we identified a set of essential genes/pathways to use in machine learning. These signatures were further refined by biological enrichment, targetability and toxicity in order to build a process for rational novel target ID generation. Citation Format: Miika Ahdesmäki, Joshua Armenia, Krishna Bulusu, Ben Sidders, Jonathan Dry, Ultan McDermott. Cancer dependency signatures for line of sight genomic alterations in squamous cell lung cancer yield rational targets [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 3379.

Resistance to targeted EGFR inhibitors is likely to develop in EGFR-mutant lung cancers. Early identification of innate or acquired resistance mechanisms to these agents is essential to direct development of future therapies. We describe the detection of heterogeneous mechanisms of resistance within populations of EGFR-mutant cells (PC9 and/or NCI-H1975) with acquired resistance to current and newly developed EGFR tyrosine kinase inhibitors, including AZD9291. We report the detection of NRAS mutations, including a novel E63K mutation, and a gain of copy number of WT NRAS or WT KRAS in cell populations resistant to gefitinib, afatinib, WZ4002, or AZD9291. Compared with parental cells, a number of resistant cell populations were more sensitive to inhibition by the MEK inhibitor selumetinib (AZD6244; ARRY-142886) when treated in combination with the originating EGFR inhibitor. In vitro, a combination of AZD9291 with selumetinib prevented emergence of resistance in PC9 cells and delayed resistance in NCI-H1975 cells. In vivo, concomitant dosing of AZD9291 with selumetinib caused regression of AZD9291-resistant tumors in an EGFRm/T790M transgenic model. Our data support the use of a combination of AZD9291 with a MEK inhibitor to delay or prevent resistance to AZD9291 in EGFRm and/or EGFRm/T790M tumors. Furthermore, these findings suggest that NRAS modifications in tumor samples from patients who have progressed on current or EGFR inhibitors in development may support subsequent treatment with a combination of EGFR and MEK inhibition. Cancer Res; 75(12); 2489–500. ©2015 AACR.

PARP inhibitors (PARPi) are promising novel therapies for tumors harboring defects in homologous recombination (HR) DNA repair. Loss of function (LoF) mutations in the tumor suppressors BRCA1 and BRCA2, two critical components in HR pathway, are the most common cause of HR deficiency in tumors. Complete loss of BRCA1/2 function requires bi-allelic loss, and it is hypothesized to be needed for PARPi sensitivity. However, the degrees of bi-allelic loss in breast and ovarian tumors are still lacking consensus, especially when both LoF events are somatic. Recent publications shown a high percentage of germline BRCA2 lacking LOH, including TCGA’s breast and ovarian cohorts. To better understand the loss of wild type BRCA genes, we reanalyzed raw data from TCGA using a novel pipeline to identify both germline and somatic BRCA mutations in breast and ovarian. We then applied mathematical models to determine LOH or bi-allelic loss of those mutations. Our analysis shows a higher degree of bi-allelic loss of BRCA mutations than previously reported, especially germline BRCA2 in breast cancer, which shows nearly 80%. Furthermore, somatic BRCA mutations, which account for about 1/3 of all BRCA mutations, showed a similar degree of bi-allelic loss as germline BRCA mutations. In addition, somatic and germline BRCA were shown to have comparable HRD scores, a measure of tumor genomic instability. Germline BRCA patients are typically diagnosed at younger ages in breast and ovarian cancer, whereas somatic show no difference to non-BRCA patients. In ovarian cancer, patients with somatic or germline BRCA had similar progression-free and overall survival with comparable treatment journeys, suggesting a phenotypic equivalence. Interestingly, BRCA2 tumors showed lower HRD scores than BRCA1 tumors in both breast and ovarian. In breast cancer BRCA1 mutation is significantly enriched in TNBC, but the degree of bi-allelic loss did not show any significant difference between ER+ and TNBC subtypes. To validate our findings, we examined a much larger cohort of breast (n > 5,000) and ovarian (n > 2,500) cancers sequenced by Foundation Medicine, with germline/somatic and LOH status determined computationally using validated methodology. We found similar prevalence of somatic BRCA mutations and the degree of bi-allelic loss and HRD scores in both germline and somatic BRCA mutations. In addition, potential reactivating reversion mutations were discovered in both somatic and germline mutated BRCA in both breast and ovarian cancers, suggesting they are driver mutations. In summary, large scale analysis of BRCA mutations in breast and ovarian cancers suggest that about 1/3 of BRCA mutations are somatic, and both germline and somatic BRCA mutations show high degree of bi-allelic loss and similar genomic instability and may benefit from PARPi therapies. Citation Format: Zhongwu Lai, Darren Hodgson, James Sun, Adrienne Johnson, Carl Barrett, Jonathan Dry. Both somatic and germline BRCA mutations show high degree of bi-allelic loss in breast and ovarian cancers [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 1747.

Purpose: KRAS wild-type status is an imperfect predictor of sensitivity to anti-EGF receptor (EGFR) monoclonal antibodies in colorectal cancer, motivating efforts to identify novel molecular aberrations driving RAS. This study aimed to build a quantitative readout of RAS pathway activity to (i) uncover molecular surrogates of RAS activity specific to colorectal cancer, (ii) improve the prediction of cetuximab response in patients, and (iii) suggest new treatment strategies. Experimental Design: A model of RAS pathway activity was trained in a large colorectal cancer dataset and validated in three independent colorectal cancer patient datasets. Novel molecular traits were inferred from The Cancer Genome Atlas colorectal cancer data. The ability of the RAS model to predict resistance to cetuximab was tested in mouse xenografts and three independent patient cohorts. Drug sensitivity correlations between our model and large cell line compendiums were performed. Results: The performance of the RAS model was remarkably robust across three validation datasets. (i) Our model confirmed the heterogeneity of the RAS phenotype in KRAS wild-type patients, and suggests novel molecular traits driving its phenotype (e.g., MED12 loss, FBXW7 mutation, MAP2K4 mutation). (ii) It improved the prediction of response and progression-free survival (HR, 2.0; P < 0.01) to cetuximab compared with KRAS mutation (xenograft and patient cohorts). (iii) Our model consistently predicted sensitivity to MAP–ERK kinase (MEK) inhibitors (P < 0.01) in two cell panel screens. Conclusions: Modeling the RAS phenotype in colorectal cancer allows for the robust interrogation of RAS pathway activity across cell lines, xenografts, and patient cohorts. It demonstrates clinical utility in predicting response to anti-EGFR agents and MEK inhibitors. Clin Cancer Res; 20(1); 265–72. ©2013 AACR.

MEK inhibitor (selumetinib) is a potent, orally active inhibitor of MAPK/ERK pathway. It is important to develop an accurate and robust method indicative of RAS pathway activity to stratify potential patients who can benefit from selumetinib treatment in gastric cancer (GC). First, we surveyed the sensitivity to selumetinib in a panel of 22 GC cell lines and correlated with MEK signature to selumetinib sensitivity. Next, we analyzed MEK signature via nanostring assay in two Asian cohorts using clinical samples (n = 218) and then performed a correlative analysis with MEK signature status and KRAS genotype in GC. MEK signature was predictive of response of selumetinib in GC cell lines regardless of KRAS mutation status. The proportion of high MEK signature by nanostring assay was 6.9% and the proportion of high MEK signature was significantly higher in KRAS altered group in a Korean cohort. None of PIK3CA altered cases belonged to high MEK signature group. MEK high signature was more prevalent in intestinal type by Lauren classification. The correlation between MEK signature, KRAS alteration and treatment response to selumetinib should be validated in prospective clinical trials.

Accurate variant calling in next generation sequencing (NGS) is critical to understand cancer genomes better. Here we present VarDict, a novel and versatile variant caller for both DNA- and RNA-sequencing data. VarDict simultaneously calls SNV, MNV, InDels, complex and structural variants, expanding the detected genetic driver landscape of tumors. It performs local realignments on the fly for more accurate allele frequency estimation. VarDict performance scales linearly to sequencing depth, enabling ultra-deep sequencing used to explore tumor evolution or detect tumor DNA circulating in blood. In addition, VarDict performs amplicon aware variant calling for polymerase chain reaction (PCR)-based targeted sequencing often used in diagnostic settings, and is able to detect PCR artifacts. Finally, VarDict also detects differences in somatic and loss of heterozygosity variants between paired samples. VarDict reprocessing of The Cancer Genome Atlas (TCGA) Lung Adenocarcinoma dataset called known driver mutations in KRAS, EGFR, BRAF, PIK3CA and MET in 16% more patients than previously published variant calls. We believe VarDict will greatly facilitate application of NGS in clinical cancer research.

AZD8055 is a small-molecule inhibitor of mTOR (mammalian target of rapamycin) kinase activity. The present review highlights molecular and phenotypic differences between AZD8055 and allosteric inhibitors of mTOR such as rapamycin. Biomarkers, some of which are applicable to clinical studies, as well as biological effects such as autophagy, growth inhibition and cell death are compared between AZD8055 and rapamycin. Potential ways to develop rational combinations with mTOR kinase inhibitors are also discussed. Overall, AZD8055 may provide a better therapeutic strategy than rapamycin and analogues.

This abstract has been withheld from publication due to its inclusion in the AACR Annual Meeting 2018 Official Press Program. It will be posted online following its presentation. Citation Format: Jonathan R. Dry, Michael P. Menden, Krishna Bulusu, Justin Guinney, Julio Saez-Rodriguez. A large cancer pharmacogenomics combination screen powering crowd-sourced advancement of computational drug synergy predictions [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 3886.

Exciting clinical benefit has been seen from both immune checkpoint inhibitors and PARP inhibitors in several cancer types. Initial reports have suggested the tumor mutational burden (TMB) may be a tumor agnostic marker for predicting immunotherapy (IO) response. Deficiencies in DNA damage repair (DDR) resulting from BRCA1 and BRCA2 mutations can lead to an accumulation of tumor mutations, and sensitivity to PARP inhibitors. We hypothesized that patients with germline or somatic mutations in BRCA1/2 or other homologous recombination repair (HRR) genes, termed “HRRm”, will have higher TMB, higher immune infiltration and activity, and greater benefit from the combination of DDR and IO therapies. We investigated the relationship between DDR deficiencies, TMB, PDL1, transcriptomic immunoscores and drug response across The Cancer Genome Atlas (TCGA) disease cohorts, data from AstraZeneca's clinical studies and a linked NSCLC genomic and EHR dataset from Foundation Medicine and FlatIron Health. HRRm cancer samples show higher TMB in several cancer types. This does trend with higher PDL1 but does not associate with transcriptomic immunoscores. HRRm could be a practical clinical biomarker to select patients with DDR defects likely to benefit from combination of DDR and IO therapies. Citation Format: Jonathan R. Dry, Zhongwu Lai, Brian Dougherty, Carl Barrett, Steven Criscione. Homologous recombination repair gene mutations for DNA repair and immune oncology drug combinations [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 5724.

Selumetinib (AZD6244, ARRY-142886) is a selective, non–ATP-competitive inhibitor of mitogen-activated protein/extracellular signal–regulated kinase kinase (MEK)-1/2. The range of antitumor activity seen preclinically and in patients highlights the importance of identifying determinants of response to this drug. In large tumor cell panels of diverse lineage, we show that MEK inhibitor response does not have an absolute correlation with mutational or phospho-protein markers of BRAF/MEK, RAS, or phosphoinositide 3-kinase (PI3K) activity. We aimed to enhance predictivity by measuring pathway output through coregulated gene networks displaying differential mRNA expression exclusive to resistant cell subsets and correlated to mutational or dynamic pathway activity. We discovered an 18-gene signature enabling measurement of MEK functional output independent of tumor genotype. Where the MEK pathway is activated but the cells remain resistant to selumetinib, we identified a 13-gene signature that implicates the existence of compensatory signaling from RAS effectors other than PI3K. The ability of these signatures to stratify samples according to functional activation of MEK and/or selumetinib sensitivity was shown in multiple independent melanoma, colon, breast, and lung tumor cell lines and in xenograft models. Furthermore, we were able to measure these signatures in fixed archival melanoma tumor samples using a single RT-qPCR–based test and found intergene correlations and associations with genetic markers of pathway activity to be preserved. These signatures offer useful tools for the study of MEK biology and clinical application of MEK inhibitors, and the novel approaches taken may benefit other targeted therapies. Cancer Res; 70(6); 2264–73

Purpose: To develop a clinically viable gene expression assay to measure RAS/RAF/MEK/ERK (RAS–ERK) pathway output suitable for hypothesis testing in non–small cell lung cancer (NSCLC) clinical studies. Experimental Design: A published MEK functional activation signature (MEK signature) that measures RAS–ERK functional output was optimized for NSCLC in silico. NanoString assays were developed for the NSCLC optimized MEK signature and the 147-gene RAS signature. First, platform transfer from Affymetrix to NanoString, and signature modulation following treatment with KRAS siRNA and MEK inhibitor, were investigated in cell lines. Second, the association of the signatures with KRAS mutation status, dynamic range, technical reproducibility, and spatial and temporal variation was investigated in NSCLC formalin-fixed paraffin-embedded tissue (FFPET) samples. Results: We observed a strong cross-platform correlation and modulation of signatures in vitro. Technical and biological replicates showed consistent signature scores that were robust to variation in input total RNA; conservation of scores between primary and metastatic tumor was statistically significant. There were statistically significant associations between high MEK (P = 0.028) and RAS (P = 0.003) signature scores and KRAS mutation in 50 NSCLC samples. The signatures identify overlapping but distinct candidate patient populations from each other and from KRAS mutation testing. Conclusions: We developed a technically and biologically robust NanoString gene expression assay of MEK pathway output, compatible with the quantities of FFPET routinely available. The gene signatures identified a different patient population for MEK inhibitor treatment compared with KRAS mutation testing. The predictive power of the MEK signature should be studied further in clinical trials. Clin Cancer Res; 23(6); 1471–80. ©2016 AACR. See related commentary by Xue and Lito, p. 1365

Background Identification of synthetic lethal interactions in cancer cells could offer promising new therapeutic targets. Large-scale functional genomic screening presents an opportunity to test large numbers of cancer synthetic lethal hypotheses. Methods enriching for candidate synthetic lethal targets in molecularly defined cancer cell lines can steer effective design of screening efforts. Loss of one partner of a synthetic lethal gene pair creates a dependency on the other, thus synthetic lethal gene pairs should never show simultaneous loss-of-function. We have developed a computational approach to mine large multi-omic cancer data sets and identify gene pairs with mutually exclusive loss-of-function. Since loss-of-function may not always be genetic, we look for deleterious mutations, gene deletion and/or loss of mRNA expression by bimodality defined with a novel algorithm BiSEp. Results Applying this toolkit to both tumour cell line and patient data, we achieve statistically significant enrichment for experimentally validated tumour suppressor genes and synthetic lethal gene pairings. Notably non-reliance on genetic loss reveals a number of known synthetic lethal relationships otherwise missed, resulting in marked improvement over genetic-only predictions. We go on to establish biological rationale surrounding a number of novel candidate synthetic lethal gene pairs with demonstrated dependencies in published cancer cell line shRNA screens. Conclusions This work introduces a multi-omic approach to define gene loss-of-function, and enrich for candidate synthetic lethal gene pairs in cell lines testable through functional screens. In doing so, we offer an additional resource to generate new cancer drug target and combination hypotheses. Algorithms discussed are freely available in the BiSEp CRAN package at http://cran.r-project.org/web/packages/BiSEp/index.html. Electronic supplementary material The online version of this article (doi:10.1186/s12864-016-2375-1) contains supplementary material, which is available to authorized users.

Genome medicine 8(1), 125 (2016). doi:10.1186/s13073-016-0379-8, Published by BioMed Central, London

This paper presents an algorithm which is able to extract discriminant rules from oligopeptides for protease proteolytic cleavage activity prediction. The algorithm is developed using genetic programming. Three important components in the algorithm are a min-max scoring function, the reverse Polish notation (RPN) and the use of minimum description length. The min-max scoring function is developed using amino acid similarity matrices for measuring the similarity between an oligopeptide and a rule, which is a complex algebraic equation of amino acids rather than a simple pattern sequence. The Fisher ratio is then calculated on the scoring values using the class label associated with the oligopeptides. The discriminant ability of each rule can therefore be evaluated. The use of RPN makes the evolutionary operations simpler and therefore reduces the computational cost. To prevent overfitting, the concept of minimum description length is used to penalize over-complicated rules. A fitness function is therefore composed of the Fisher ratio and the use of minimum description length for an efficient evolutionary process. In the application to four protease datasets (Trypsin, Factor Xa, Hepatitis C Virus and HIV protease cleavage site prediction), our algorithm is superior to C5, a conventional method for deriving decision trees.

Selecting the right drug combination for the right patient is a complex problem. Some success has been seen from preclinical screening and computational approaches combining molecular, drug and tumor properties to predict combination benefit within the tumor cell. However the context specificity and heterogeneity of drug resistance is often overlooked, leading to inappropriate assumptions of statistical power and resulting in poor translatability. Knowledge-aware approaches will be needed to augment pre-clinical statistical discovery but a huge variety of techniques exist, each offering different strengths. To test this world of methods, AstraZeneca partnered with DREAM, Sage Bionetworks and the Sanger Institute. We released over 11,000 test results from drug combinations in cancer cell lines for an incentivized crowd-sourcing challenge to develop computational models predicting synergistic drug combinations and biomarkers determining response. This became the highest participated of any DREAM challenge, with >700 participants from diverse geographies and knowledge/skill specialties collectively contributing >50,000 hours to the problem. Over 70 models were submitted, testing most machine learning algorithms alongside knowledge-driven approaches. The best models out-performed the state of the art, consistently predicting at an equivalent level to biological replication. Network techniques to select molecular markers most relevant to drug target biology improved prediction the most. Amongst results that could be validated in independent data, ERBB2/EGFR/EP300 variants were predictive for combinations inhibiting AKT+EGFR/ERBB, KRAS for BRAF+MAP2K, AR for AKT/SGK+MAP2K, and ESR1 for ESR1+AKT/PI3K. Alongside revealing interesting biology in our data and providing predictive models, the challenge has given clear direction to our future investment in computational methods for combination pharmacogenomics. Citation Format: Jonathan R. Dry, Michael Menden, Julio Saez Rodriguez, Justin Guinney. A DREAM-world of approaches for predicting combination synergies in cancer [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2017 Oct 26-30; Philadelphia, PA. Philadelphia (PA): AACR; Mol Cancer Ther 2018;17(1 Suppl):Abstract nr B076.

Targeted sequencing is used increasingly in clinics to guide therapeutic decisions by measuring mutations, small indels and/or rearrangements in cancer genes. An ability to use the same platform to detect additional oncogene activation (or tumor suppressor loss) through copy number changes could significantly expand the number of patients able to benefit from targeted therapies. Many currently available tools either requires a matched normal, works only for whole genome or exome sequencing, or don't work for PCR-based targeted sequencing. In addition, no tools available to detect breakpoints within a gene from targeted sequencing. A versatile copy number analysis tool from sequencing is needed to maximize the value of the sequencing routinely applied in clinics. Here we presented a novel computational tool, Seq2C (Sequencing To Copy Number), which is versatile to handle various situations and reports aberrations at gene level ready for interpretation. Seq2C works at cohort level and does not require a matched ‘normal’, though it can optionally use one or more normal samples for small ( Another distinct feature differentiating Seq2C from other currently available tools is that Seq2C identifies breakpoints and detects one or more exon deletion or duplication rearrangements within a gene, which is common in tumor suppressors as a mechanism to lose function. In addition, it can also detect potential fusions in genes such as ALK and ERG, where a copy number change is often accompanied with the fusion and a breakpoint can thus be called by Seq2C. Furthermore, it can predict gender from exome or whole genome sequencing. We applied Seq2C to exome sequencing of CCLE cell lines, and showed that it produced gene level copy number data highly correlated to those derived from microarrays, the current gold standard. Interestingly, Seq2C identified one or more exon deletions in several common tumor suppressors, such as TP53, PTEN, CDKN2A, NF1, STK11 and RB1, in cell lines with no known aberrations. They were supported by RNA-Seq data. It also correctly called known ERG fusion in prostate cell line NCI-H660 and identified a previously un-reported EML4-ALK fusion in a pancreatic cancer cell line SNU-324, which is confirmed by RNA-Seq data, suggesting EML4-ALK is not limited to lung cancer. In conclusion, Seq2C is a versatile copy number analysis tool for sequencing and will be useful for cancer research. Seq2C is freely available in GitHub. Citation Format: Zhongwu Lai, Aleksandra Markovets, Jonathan Dry. Seq2C: from sequence to copy number for cancer samples. [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 5268.

Our current understanding of the mutation spectrum of relapse/refractory patients is limited. Public sequencing efforts have focused by design, on primary tumors, while re-biopsy of patients who are refractory to first line therapy or who relapse on treatment is not standard protocol. We have performed whole exome sequencing on post standard of care biopsies from 3 patient cohorts, 47 DLBCL patients post-relapse following R-CHOP, 49 triple negative breast (TNBC) following taxol alone or in combination and 41 lung adenocarcinoma (LUAD) following platinum alone or in combination. Comparative analysis of the mutation spectrum of these cohorts was performed against available clinically matched primary tumor cohorts. Comparison of the mutational signatures, somatic variant frequencies and copy number variation between matched pre- and post-treatment samples, and treated-primary tumor cohorts were performed. Significant differences were identified in the frequency of oncogenic drivers including PIK3CA in TNBC and KRAS in LUAD. We also observed mutational signature differences between cohorts that could impact treatment regimens post standard of care. Collectively these data further highlight the value of re-assessing the tumor genetic landscape at the time of treatment rather than relying on primary tumor samples, which may not be representative of the cohort to be treated. Citation Format: Danielle Greenawalt, Austin Dulak, Manasa Ramakrishna, Zhongwu Lai, Justin Johnson, Hongyue Dai, Melissa Mitchell, Carl Barrett, Stephen Fawell, Jonathan Dry. Your targeted population might not be what you predict: Changes in tumor genetic landscapes post standard of care. [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 3168.

The acquisition of resistance to protein kinase inhibitors is a growing problem in cancer treatment. We modeled acquired resistance to the MEK1/2 (mitogen-activated or extracellular signal-regulated protein kinase kinases 1 and 2) inhibitor selumetinib (AZD6244) in colorectal cancer cell lines harboring mutations in BRAF (COLO205 and HT29 lines) or KRAS (HCT116 and LoVo lines). AZD6244-resistant derivatives were refractory to AZD6244-induced cell cycle arrest and death and exhibited a marked increase in ERK1/2 (extracellular signal-regulated kinases 1 and 2) pathway signaling and cyclin D1 abundance when assessed in the absence of inhibitor. Genomic sequencing revealed no acquired mutations in MEK1 or MEK2, the primary target of AZD6244. Rather, resistant lines showed a marked up-regulation of their respective driving oncogenes, BRAF(600E) or KRAS(13D), due to intrachromosomal amplification. Inhibition of BRAF reversed resistance to AZD6244 in COLO205 cells, which suggested that combined inhibition of MEK1/2 and BRAF may reduce the likelihood of acquired resistance in tumors with BRAF(600E). Knockdown of KRAS reversed AZD6244 resistance in HCT116 cells as well as reduced the activation of ERK1/2 and protein kinase B; however, the combined inhibition of ERK1/2 and phosphatidylinositol 3-kinase signaling had little effect on AZD6244 resistance, suggesting that additional KRAS effector pathways contribute to this process. Microarray analysis identified increased expression of an 18-gene signature previously identified as reflecting MEK1/2 pathway output in resistant cells. Thus, amplification of the driving oncogene (BRAF(600E) or KRAS(13D)) can drive acquired resistance to MEK1/2 inhibitors by increasing signaling through the ERK1/2 pathway. However, up-regulation of KRAS(13D) leads to activation of multiple KRAS effector pathways, underlining the therapeutic challenge posed by KRAS mutations. These results may have implications for the use of combination therapies.

Tumors are characterized by properties of genetic instability, heterogeneity, and significant oligoclonality. Elucidating this intratumoral heterogeneity is challenging but important. In this study, we propose a framework, BubbleTree, to characterize the tumor clonality using next generation sequencing (NGS) data. BubbleTree simultaneously elucidates the complexity of a tumor biopsy, estimating cancerous cell purity, tumor ploidy, allele-specific copy number, and clonality and represents this in an intuitive graph. We further developed a three-step heuristic method to automate the interpretation of the BubbleTree graph, using a divide-and-conquer strategy. In this study, we demonstrated the performance of BubbleTree with comparisons to similar commonly used tools such as THetA2, ABSOLUTE, AbsCN-seq and ASCAT, using both simulated and patient-derived data. BubbleTree outperformed these tools, particularly in identifying tumor subclonal populations and polyploidy. We further demonstrated BubbleTree's utility in tracking clonality changes from patients’ primary to metastatic tumor and dating somatic single nucleotide and copy number variants along the tumor clonal evolution. Overall, the BubbleTree graph and corresponding model is a powerful approach to provide a comprehensive spectrum of the heterogeneous tumor karyotype in human tumors. BubbleTree is R-based and freely available to the research community (https://www.bioconductor.org/packages/release/bioc/html/BubbleTree.html).

Inhibition of the RAS-RAF-MEK axis holds significant promise for targeted cancer therapy, highlighted by the recent clinical approval of BRAF (vemurafenib) and MEK (trametinib) inhibitors, and ongoing late-stage clinical trials for several other MEK inhibitors. All genetic aberrations resulting in pathway activation and dependency through which optimum patient populations may be selected are as yet not fully determined. Previously we published an 18-gene mRNA expression signature reproducibly predictive of MEK pathway output and response to MEK inhibition with selumetinib. Here we demonstrate that a 6-gene sub-signature shows enhanced reproducibility in cell lines, primary tumor explant in vivo models and fresh/fixed tissue samples, crossing tumor type and gene expression platforms. In published next-generation sequencing datasets we assessed pathway-linked genetic aberrations for their impact on MEK signature output. We show the potential for this signature to dissect MEK dependency amongst RAS mutant tumors. Furthermore we identify a number of genetic aberrations outside of BRAF and KRAS driving high MEK signature expression, including a synthetic lethal-like profile with NF1 deletion. The signature provides a robust and dynamic marker of MEK output applicable to clinical samples, and highlights a number of genetic opportunities to expand the clinical scope for MEK inhibitor based therapies. Citation Information: Mol Cancer Ther 2013;12(11 Suppl):A209. Citation Format: Jonathan R. Dry, Darren Hodgson, Roz Brant, Tom Liptrot, Elizabeth Harrington, Carl Barrett, Paul D. Smith. Predicting genetic drivers of MEK dependency with gene expression. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2013 Oct 19-23; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2013;12(11 Suppl):Abstract nr A209.

Building upon the exciting precedent exemplified by olaparib in the PARP-BRCA axis, we aim to identify and exploit further synthetic lethal (SL) interactions within and beyond the DNA damage response (DDR) area. To facilitate with this search, we have developed a method implemented as an in silico algorithm that uses gene expression and mutation data to detect putative SL interactions in cell lines and cancer tissue across The Cancer Genome Atlas (TCGA). This approach has detected relatively rare signatures including those described in the literature, for example ATM and ATR. The algorithm has also revealed potentially novel in silico SL gene pairs outside of DDR involving frequently mutated tumor suppressor genes linked to targets that AstraZeneca has previously developed selective compounds against. We are validating an attractive selection of these in vitro using appropriate cell lines, siRNA and small molecule inhibitors. Citation Format: Mark Wappett, Richard Ward, Paul Fisher, James Bradford, Jonathan Dry, Mark O'Connor, Alan Lau, Steve Durant. Exploiting synthetic lethal interactions in the DNA damage response area and beyond in drug discovery. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Synthetic Lethal Approaches to Cancer Vulnerabilities; May 17-20, 2013; Bellevue, WA. Philadelphia (PA): AACR; Mol Cancer Ther 2013;12(5 Suppl):Abstract nr A41.

Primary tumor xenograft models (PTX) are established in immunodeficient mice by implantation of material directly from patient tumors. These models are propagated in vivo so that they are not subjected to additional selection pressures within tissue culture. The models reputedly reflect a greater genetic diversity of disease than can be recapitulated within tumor cell line derived xenografts and may have morphological differences. Consequently, such models may be of greater relevance for the evaluation of drug efficacy. To verify differences versus cell line derived xenograft models, we have examined samples of 42 CRC PTX models, derived from Dukes stage A-D tumor samples, which were obtained from European contract research organisations) Characterisation included analysis of common genetic mutations (KRas, BRAF, PI3Ka, PTEN, P53 and APC), mRNA expression via microarray platforms (Affymetrics HG_U133_plus_2), high-throughput RT-PCR of specific probes (mouse and human) to examine stromal genes, and the activation of particular proteins using reverse phase antibody arrays and immunohistochemistry. Comparisons were made with 6 commonly used CRC cell line derived xenografts. The PTX models overall represent a broader range of molecular pathology - for example the inclusion of wild-type Kras tumors. Histopathological characterisation also confirmed that the PTX models have a higher stromal content that is routinely observed in cell line derived xenografts, and a more complex morphology with higher tumor cell differentiation. PTX models may potentially complement drug discovery activities by providing a wider platform in which to test preclinical hypotheses - based upon a defined biological mechanism and/or a genetic determinant of sensitivity. 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 A13.

Patient-derived explant models (PTX) show a higher genetic and pathological heterogeneity than cell line derived xenograft models, and may have greater clinical relevance in cancer drug development and personalised oncology treatments than cell line derived xenograft models. Therefore PTX models can potentially be used to provide additional confidence in patient stratification hypotheses for novel therapies or combination approaches. To examine comparative pharmacological responses in PTX models, the monotherapy activity of the MEK inhibitor selumetinib was determined in colorectal cancer PTX models that were stratified according to a MEK pathway “functional activation” mRNA signature, originally derived using a broad human tumor cell line panel (1). Tumor passage, immunohistochemical and genetic markers of target pathway activity were also assessed to collectively select 6 models of high interest for examining response to treatment. In addition, to monotherapy treatment, 3 models were also used to examine the combination of selumetinib with the mTORC1/2 kinase inhibitor AZD8055. The results suggest that the MEK pathway mRNA signature is applicable in colorectal PTX models and enables selumetinib responsive versus non-responsive models to be identified. Combination therapy of selumetinib and AZD8055 has also shown additional therapeutic benefit, similar to that observed in selected cell line derived xenografts. These data illustrate that use of PTX panels can be applied to substantiate in-vitro derived biomarker-led hypotheses for a given therapeutic approach. In addition, biomarker stratification has focused experimental design to test novel:novel combinations and may enable stratification for prospective patient subsets. Collectively this approach to disease modelling could significantly reduce wide screening in vivo studies and steer more refined clinically relevant hypotheses, whilst allowing for a reduction in animal usage. 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 C2.

Background: Early results from Phase II trials have shown that treatment with the PARP inhibitor olaparib (AZD2281, KU-0059436) can induce tumour specific synthetic lethality in patients with BRCA-mutated breast and ovarian cancer, with little effect on normal tissues. Pre-clinical studies have indicated that sensitivity can also result from other perturbations affecting homologous recombination (HR), suggesting broader patient populations may benefit from this drug. We aimed to generate pre-clinical hypotheses as to the key biological mechanisms regulating response to this drug; and identify associated biomarkers by which responding patient subsets could be stratified. Methods: A panel of 95 cell lines representing breast, ovarian, colorectal, lung, head & neck and pancreatic cancers was tested for sensitivity to olaparib using 2D-clonogenic survival assays. Baseline (untreated) gene expression profiles (Affymetrix U133A Plus2) were determined for each cell line, alongside protein expression and mutational status of core HR genes. Dynamic pathway gene expression signatures were collated from the literature, and additional signatures generated using RNAi against core HR components. Statistical and bioinformatic approaches were then applied to prioritise correlated networks of genes displaying consistent pathway overlay and predictive of response in cell line subsets. Results: 30 cell lines were highly sensitive to olaparib treatment (IC50 4 µM). Deleterious mutations in BRCA1 or BRCA2 genes were associated with only a small subset of highly sensitive cell lines, highlighting the presence of other factors able to modulate olaparib responsiveness. Transcriptome analysis revealed DNA repair and proliferation associated genes to be most consistently correlated with olaparib response. Based on these results, a candidate predictive baseline gene expression profile was established. Conclusions: By profiling a large panel of cell lines we have determined that factors in addition to BRCA mutation can be linked with olaparib sensitivity. A list of candidate gene transcripts predictive of in vitro sensitivity to olaparib were identified which may have utility as predictive biomarkers in the clinic. Although primarily linked to mechanisms expected to influence response (DNA repair, cell-cycle checkpoints and proliferation), novel roles were suggested for other pathways such as Aurora A kinase signalling. Analyses also highlighted that some classical HR pathway components may prove to be more robust mRNA markers of pathway activity than others. Studies are already underway to determine whether this baseline transcript tumour profile can be correlated with patient responses to olaparib. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 3497.

Background: The oral PARP inhibitor olaparib (AZD2281, KU-0059436) is in Phase II clinical trials in patients with BRCA-mutated breast and ovarian cancer. Early results in these trials have provided positive clinical evidence for the targeted therapy concept of synthetic lethality using PARP inhibitors in molecularly-defined tumors with little effect on normal tissues. However, olaparib offers the potential to treat a broader range of patient tumors and pre-clinical studies have indicated that homologous recombination deficiency (HRD) due to loss of other DNA repair genes or through impaired cell-cycle checkpoints can also lead to sensitivity to PARP inhibitors. The ability to stratify patients whose tumors will respond to treatment will be critical to realise the potential of this targeted therapy approach. In order to identify predictive biomarkers of olaparib response and to provide further insights into mechanisms of olaparib sensitivity a study was undertaken using a panel of cell lines correlating response with molecular profiles. Methods: A cross tumor-type panel of 95 cell lines (KU-95 panel) representing breast, ovarian, colorectal, lung, head & neck and pancreatic cancers was tested for sensitivity to olaparib using 2D-clonogenic survival assays. Baseline (untreated) gene expression profiles using Affymetrix genome-wide U133A 2.0 arrays together with protein expression and mutational status of genes, such as BRCA1, known to modify of olaparib response were determined for each cell line. Novel statistical, bioinformatics and pathway analysis approaches were used to identify genes that are predictive for olaparib response. Results: Across the KU-95 cell-line panel, 30 cell lines were highly sensitive to olaparib (IC50 4 M). Deleterious mutations in BRCA1 or BRCA2 genes were associated with only a small subset of highly sensitive cell lines indicating the presence of other factors able to modulate olaparib responsiveness. Directed analysis of several other DNA repair genes showed correlations with response consistent with previous data. Genome-wide transcriptome and pathway mapping analysis revealed DNA repair and proliferation associated genes to be most consistently correlated with olaparib sensitivity. Based on results from all these approaches a candidate baseline gene expression profile predictive of olaparib response was established. Conclusions: By profiling a large panel of cell lines we have determined that factors in addition to BRCA mutation can be linked with olaparib sensitivity. A list of candidate gene transcripts was identified that predicts sensitivity to olaparib across a broad range of tumor types in vitro and may have utility as predictive biomarkers in the clinic. Studies are already underway to determine whether this baseline transcript tumor profile can be correlated with patient responses to olaparib. Citation Information: Mol Cancer Ther 2009;8(12 Suppl):C119.