Keith Ching

Evaluation of the antitumor effects and mechanisms of PF00299804, a pan-HER inhibitor, alone or in combination with chemotherapy or targeted agents in gastric cancer

Recently, HER2-directed treatment, such as trastuzumab, has shown clinical benefit in HER2-amplified gastric cancer. On the basis of recent studies about epidermal growth factor receptor (EGFR) or HER2-targeting agents (including gefitinib, lapatinib, and trastuzumab) in gastric cancer, the potent effects of pan-HER inhibitors targeting the HER family are anticipated. In this study, we evaluated the activity and mechanisms of PF00299804, an irreversible pan-HER inhibitor, in gastric cancer in vitro and in vivo models. PF00299804 showed significant growth-inhibitory effects in HER2-amplified gastric cancer cells (SNU216, N87), and it had lower 50% inhibitory concentration values compared with other EGFR tyrosine kinase inhibitors, including gefitinib, lapatinib, BIBW-2992, and CI-1033. PF00299804 induced apoptosis and G1 arrest and inhibited phosphorylation of receptors in the HER family and downstream signaling pathways including STAT3, AKT, and extracellular signal–regulated kinases (ERK) in HER2-amplified gastric cancer cells. PF00299804 also blocked EGFR/HER2, HER2/HER3, and HER3/HER4 heterodimer formation as well as the association of HER3 with p85α in SNU216 cells. The combination of PF00299804 with clinically relevant chemotherapeutic agents or molecular-targeted agents including trastuzumab (an anti-HER2 monoclonal antibody), CP751871 (an IGF1R inhibitor), PD0325901 (an ERK1/2 inhibitor), and PF04691502 (a PI3K/mTOR inhibitor) produced synergistic effects. These findings indicate that PF00299804 can be used as a targeted therapy for the treatment of HER2-amplified gastric cancer through inhibition of HER family heterodimer formation and may augment antitumor efficacy of chemotherapeutic and/or molecular-targeted agents. Mol Cancer Ther; 11(2); 439–51. ©2011 AACR.

An integrated genomic approach to identify predictive biomarkers of response to the aurora kinase inhibitor PF-03814735.

PF-03814735 is a novel, reversible inhibitor of Aurora kinases A and B that finished a phase I clinical trial for the treatment of advanced solid tumors. To find predictive biomarkers of drug sensitivity, we screened a diverse panel of 87 cancer cell lines for growth inhibition upon PF-03814735 treatment. Small cell lung cancer (SCLC) and, to a lesser extent, colon cancer lines were very sensitive to PF-03814735. The status of the Myc gene family and retinoblastoma pathway members significantly correlated with the efficacy of PF-03814735. Whereas RB1 inactivation, intact CDKN2A/p16, and normal CCND1/Cyclin D1 status are hallmarks of SCLC, activation or amplification of any of the three Myc genes (MYC, MYCL1, and MYCN) clearly differentiated cell line sensitivity within the SCLC panel. By contrast, we found that expression of Aurora A and B were weak predictors of response. We observed a decrease in histone H3 phosphorylation and polyploidization of sensitive lines, consistent with the phenotype of Aurora B inhibition. In vivo experiments with two SCLC xenograft models confirmed the sensitivity of Myc gene-driven models to PF-03814735 and a possible schedule dependence of MYC/c-Myc–driven tumors. Altogether our results suggest that SCLC and other malignancies driven by the Myc family genes may be suitable indications for treatment by Aurora B kinase inhibitors. Mol Cancer Ther; 11(3); 710–9. ©2012 AACR.

PEST Domain Mutations in Notch Receptors Comprise an Oncogenic Driver Segment in Triple-Negative Breast Cancer Sensitive to a γ-Secretase Inhibitor

Purpose: To identify and characterize novel, activating mutations in Notch receptors in breast cancer and to determine response to the gamma secretase inhibitor (GSI) PF-03084014. Experimental Design: We used several computational approaches, including novel algorithms, to analyze next-generation sequencing data and related omic datasets from The Cancer Genome Atlas (TCGA) breast cancer cohort. Patient-derived xenograft (PDX) models were sequenced, and Notch-mutant models were treated with PF-03084014. Gene-expression and functional analyses were performed to study the mechanism of activation through mutation and inhibition by PF-03084014. Results: We identified mutations within and upstream of the PEST domains of NOTCH1, NOTCH2, and NOTCH3 in the TCGA dataset. Mutations occurred via several genetic mechanisms and compromised the function of the PEST domain, a negative regulatory domain commonly mutated in other cancers. Focal amplifications of NOTCH2 and NOTCH3 were also observed, as were heterodimerization or extracellular domain mutations at lower incidence. Mutations and amplifications often activated the Notch pathway as evidenced by increased expression of canonical Notch target genes, and functional mutations were significantly enriched in the triple-negative breast cancer subtype (TNBC). PDX models were also identified that harbored PEST domain mutations, and these models were highly sensitive to PF-03084014. Conclusions: This work suggests that Notch-altered breast cancer constitutes a bona fide oncogenic driver segment with the most common alteration being PEST domain mutations present in multiple Notch receptors. Importantly, functional studies suggest that this newly identified class can be targeted with Notch inhibitors, including GSIs. Clin Cancer Res; 21(6); 1487–96. ©2015 AACR.

A Novel SND1-BRAF Fusion Confers Resistance to c-Met Inhibitor PF-04217903 in GTL16 Cells though MAPK Activation

Targeting cancers with amplified or abnormally activated c-Met (hepatocyte growth factor receptor) may have therapeutic benefit based on nonclinical and emerging clinical findings. However, the eventual emergence of drug resistant tumors motivates the pre-emptive identification of potential mechanisms of clinical resistance. We rendered a MET amplified gastric cancer cell line, GTL16, resistant to c-Met inhibition with prolonged exposure to a c-Met inhibitor, PF-04217903 (METi). Characterization of surviving cells identified an amplified chromosomal rearrangement between 7q32 and 7q34 which overexpresses a constitutively active SND1-BRAF fusion protein. In the resistant clones, hyperactivation of the downstream MAPK pathway via SND1-BRAF conferred resistance to c-Met receptor tyrosine kinase inhibition. Combination treatment with METi and a RAF inhibitor, PF-04880594 (RAFi) inhibited ERK activation and circumvented resistance to either single agent. Alternatively, treatment with a MEK inhibitor, PD-0325901 (MEKi) alone effectively blocked ERK phosphorylation and inhibited cell growth. Our results suggest that combination of a c-Met tyrosine kinase inhibitor with a BRAF or a MEK inhibitor may be effective in treating resistant tumors that use activated BRAF to escape suppression of c-Met signaling.

Molecular Predictors of Sensitivity to the Insulin-like Growth Factor 1 Receptor Inhibitor Figitumumab (CP-751,871)

Figitumumab (CP-751,871), a potent and fully human monoclonal anti–insulin-like growth factor 1 receptor (IGF1R) antibody, has been investigated in clinical trials of several solid tumors. To identify biomarkers of sensitivity and resistance to figitumumab, its in vitro antiproliferative activity was analyzed in a panel of 93 cancer cell lines by combining in vitro screens with extensive molecular profiling of genomic aberrations. Overall response was bimodal and the majority of cell lines were resistant to figitumumab. Nine of 15 sensitive cell lines were derived from colon cancers. Correlations between genomic characteristics of cancer cell lines with figitumumab antiproliferative activity revealed that components of the IGF pathway, including IRS2 (insulin receptor substrate 2) and IGFBP5 (IGF-binding protein 5), played a pivotal role in determining the sensitivity of tumors to single-agent figitumumab. Tissue-specific differences among the top predictive genes highlight the need for tumor-specific patient selection strategies. For the first time, we report that alteration or expression of the MYB oncogene is associated with sensitivity to IGF1R inhibitors. MYB is dysregulated in hematologic and epithelial tumors, and IGF1R inhibition may represent a novel therapeutic opportunity. Although growth inhibitory activity with single-agent figitumumab was relatively rare, nine combinations comprising figitumumab plus chemotherapeutic agents or other targeted agents exhibited properties of synergy. Inhibitors of the ERBB family were frequently synergistic and potential biomarkers of drug synergy were identified. Several biomarkers of antiproliferative activity of figitumumab both alone and in combination with other therapies may inform the design of clinical trials evaluating IGF1R inhibitors. Mol Cancer Ther; 12(12); 2929–39. ©2013 AACR.

Initial Clinical Sensitivity and Acquired Resistance to MET Inhibition in MET-Mutated Papillary Renal Cell Carcinoma

Introduction Papillary renal cell carcinoma (RCC) is the most prevalent nonclear cell histologic subtype of renal carcinoma and constitutes approximately 10% of renal cancers, affecting 5,400 patients per year in the United States. Activating MET mutations are present in the majority of hereditary papillary RCCs and up to 13% of sporadic papillary RCCs. Here we describe a patient with MET-mutated papillary RCC that responded to MET inhibition with a small molecule kinase inhibitor, PF-04217903, for 26 months. At the time of acquired resistance to treatment, we identified a genomic duplication that encompassed the mutated kinase domain of MET.

Abstract 4874: OASIS: A centralized portal for cancer omics data analysis

Advances in cancer research and sequencing technologies have contributed to the proliferation of large-scale cancer omics data sets from both public consortia and privately funded collaborations. There are increasing demands for broadly accessible tools that enable scientists to perform ad hoc analyses of cancer omics data. Here we present OASIS (http://oasis.pfizer.com), an open-access web portal that enables complex analytical queries across somatic mutations, copy number changes (CNV) and gene expression data from public domain and Pfizer funded omics studies. OASIS was designed to perform multidimensional data integration, allowing users to analyze correlations among multiple data types, as well as visualize and compare alterations across different cancers. Users can browse alteration summary reports at the gene and sample level and perform analyses using interactive visualizations. These include the Pan-Cancer Report, a unique tool that provides a high-level overview of genetic variation across multiple cancer types. Through the use of the interactive visualizations researchers can explore differentially expressed genes, identify tumor samples for over expression, investigate the correlation between CNV and expression or survey alteration patterns for a list of genes. Users can explore molecular profiles from the Cancer Cell Line Encyclopedia (CCLE) and the Catalog of Somatic Mutations in Cancer (COSMIC), to identify cell line models of interest. Researchers can also construct complex queries through a user-friendly web interface supported by the Biomart query engine. Programmatic access is also available through web services. OASIS is a powerful tool that enables cancer researchers to perform integrative analyses of large-scale cancer omics data sets, thereby facilitating key steps in oncology drug discovery ranging from target identification to model selection. Citation Format: Julio Fernandez-Banet, Anthony Esposito, Scott Coffin, Sabine Schefzick, Ying Ding, Keith Ching, Istvan Horvath, Peter Roberts, Paul Rejto, Zhengyan Kan. OASIS: A centralized portal for cancer omics data analysis. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4874. doi:10.1158/1538-7445.AM2015-4874

Cell Index Database (CELLX): a web tool for cancer precision medicine.

The Cell Index Database, (CELLX) (http://cellx.sourceforge.net) provides a computational framework for integrating expression, copy number variation, mutation, compound activity, and meta data from cancer cells. CELLX provides the computational biologist a quick way to perform routine analyses as well as the means to rapidly integrate data for offline analysis. Data is accessible through a web interface which utilizes R to generate plots and perform clustering, correlations, and statistical tests for associations within and between data types for ~20,000 samples from TCGA, CCLE, Sanger, GSK, GEO, GTEx, and other public sources. We show how CELLX supports precision oncology through indications discovery, biomarker evaluation, and cell line screening analysis.

Abstract 2615: Amplification and/or high expression of Myc family genes sensitizes tumor cells to aurora kinase inhibitors

The identification of patient populations predicted to derive clinical benefit from systemic treatment regimens is of critical importance for development of targeted therapies. Collective clinical data available for the Aurora kinase inhibitor class of targeted therapies has indicated that broad single agent clinical activity is not readily apparent. We screened a panel of established lung (mostly non-small cell) and colon cell lines for growth inhibitory sensitivity to PF-03814735, an Aurora family kinase inhibitor, revealing a potential correlation with Myc family amplification or expression. The Aurora kinases A and B have been shown to be functionally linked with Myc family oncoproteins in a number of tumor types. Myc family gene amplification events have been reported in about 30% of small cell lung cancer (SCLC) primary tumors and around 50% of cell lines established from SCLC. We next screened a panel of around 20 SCLC lines for sensitivity to PF-03814735 and its relationship to Myc family gene copy number and gene expression levels. Sensitivity to PF-03814735 in vitro was strongly correlated with amplification events in at least one of the Myc family genes (c-Myc, L-Myc, N-Myc) as well as mRNA expression levels of those genes. Myc family expression demonstrated a significant correlation with gene copy number. Follow up studies to evaluate antitumor efficacy in two SCLC xenograft models, H82 (c-Myc) and H69 (N-Myc), indicated significant tumor growth inhibition by PF-03814735. However, the SCLC models were not appreciably more sensitive to PF-03814735 than several non-Myc amplified tumors studied previously (HCT-116, COLO-205, MDA-MB-231) indicating that further study of optimal dose schedule and/or the molecular basis of sensitivity is warranted. 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 2615.

Abstract 2278: Next-generation sequencing (NGS) analysis from a phase II single-agent gedatolisib study in patients with endometrial cancer

Genetic alterations in the PI3K pathway are abundant in endometrial cancer. Hence it is hypothesized that PI3K/mTOR inhibitors will have utility for treatment of endometrial cancer. Gedatolisib, also known as PF-384, is a potent and selective dual PI3K-mTOR inhibitor with broad anti-tumor activity in preclinical studies. Gedatolisib delivered weekly intravenously was investigated in a Phase II single-agent study of advanced endometrial cancer. The cumulative clinical benefit rate from 38 response evaluable patients was 39.5%, comparable to historical rates with mTOR inhibitors. A retrospective NGS analysis of archival biopsies from patients in the study was conducted using the Foundation Medicine Inc (FMI) Foundation One Test to explore potential predictive biomarkers for clinical benefit. Of the 26 patient samples submitted for analysis, data was obtained from 19 samples of which 17 had evaluable response data. Of these 17 samples, 6 were from patients who exhibited progressive disease (PD), 6 were from patients with stable disease (SD) and 5 were from patients that showed a partial response (PR) to treatment. In general, the best responders had a low mutation load while the worst responders had a higher mutation load. Whether large mutation burden was due to mismatch-repair defects determined by MSI status of the tumors is currently under investigation. PTEN alterations were found in tumors from patients that exhibited PD and SD, but not PR. Multiple genetic alterations in ARID1A were observed in 3 of 6 tumors from patients with PD. While PIK3CA alterations were frequently present, 2 of 5 patients with PR and 1 with SD exhibited an activating PIK3CA mutation at H1047R. Among 175 genes that were observed to be mutated in the analysis of 17 independent tumor samples, the top ranking genes associated with a reduction in tumor size (% change from baseline by RECIST 1.0; Wilcoxon rank sum test) were MAP3K1 (p = 0.027) and CTNNB1 (p = 0.050). Activating mutations in β-catenin were observed in tumor from 3 of 5 patients with PR and were not present in tumor from patients with PD or SD. Tumor from a patient who exhibited an outlier clinical response (stayed on study after 2 years on treatment) had an activating mutation in Akt at E17K and also a novel mutation in mTOR at F2184L. Computational modeling analysis revealed that the mutation did not have a significant impact on kinase structure or ligand binding. KRAS mutations did not appear to correlate with clinical response. This study highlights both the wealth of information provided by NGS analysis, but also the complexity of NGS data analysis and interpretation. Citation Format: Nuzhat Pathan, Patricia English, Keith Ching, Stephen Huang, Mehran Jalaie, Kristen Pierce, Jennifer Vermette. Next-generation sequencing (NGS) analysis from a phase II single-agent gedatolisib study in patients with endometrial cancer. [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 2278.