Thomas Yu

Community assessment of cancer drug combination screens identifies strategies for synergy prediction

In the last decade advances in genomics, uptake of targeted therapies, and the advent of personalized treatments have fueled a dramatic change in cancer care. However, the effectiveness of most targeted therapies is short lived, as tumors evolve and develop resistance. Combinations of drugs offer the potential to overcome resistance. The space of possible combinations is vast, and significant advances are required to effectively find optimal treatment regimens tailored to a patient’s tumor. DREAM and AstraZeneca hosted a Challenge open to the scientific community aimed at computational prediction of synergistic drug combinations and predictive biomarkers associated to these combinations. We released a data set comprising ~11,500 experimentally tested drug combinations, coupled to deep molecular characterization of the respective 85 cancer cell lines. Among 150 submitted approaches, those that incorporated prior knowledge of putative drug targets showed superior performance predicting drug synergy across independent data. Genomic features of best-performing models revealed putative mechanisms of drug synergy for multiple drugs in combination with PI3K/AKT pathway inhibitors.Abstract The effectiveness of most cancer targeted therapies is short lived since tumors evolve and develop resistance. Combinations of drugs offer the potential to overcome resistance, however the number of possible combinations is vast necessitating data-driven approaches to find optimal treatments tailored to a patient’s tumor. AstraZeneca carried out 11,576 experiments on 910 drug combinations across 85 cancer cell lines, recapitulating in vivo response profiles. These data, the largest openly available screen, were hosted by DREAM alongside deep molecular characterization from the Sanger Institute for a Challenge to computationally predict synergistic drug pairs and associated biomarkers. 160 teams participated to provide the most comprehensive methodological development and subsequent benchmarking to date. Winning methods incorporated prior knowledge of putative drug target interactions. For >60% of drug combinations synergy was reproducibly predicted with an accuracy matching biological replicate experiments, however 20% of drug combinations were poorly predicted by all methods. Genomic rationale for synergy predictions were identified, including antagonism unique to combined PIK3CB/D inhibition with the ADAM17 inhibitor where synergy is seen with other PI3K pathway inhibitors. All data, methods and code are freely available as a resource to the community.

A high-throughput molecular data resource for cutaneous neurofibromas

Neurofibromatosis type 1 (NF1) is a genetic disorder with a range of clinical manifestations such as widespread growth of benign tumours called neurofibromas, pain, learning disorders, bone deformities, vascular abnormalities and even malignant tumours. With the establishment of the Children’s Tumour Foundation biobank, neurofibroma samples can now be collected directly from patients to be analysed by the larger scientific community. This work describes a pilot study to characterize one class of neurofibroma, cutaneous neurofibromas, by molecularly profiling of ~40 cutaneous neurofibromas collected from 11 individual patients. Data collected from each tumour includes (1) SNP Arrays, (2) Whole genome sequencing (WGS) and (3) RNA-Sequencing. These data are now freely available for further analysis at http://www.synapse.org/cutaneousNF.

A DREAM Challenge to Build Prediction Models for Short-Term Discontinuation of Docetaxel in Metastatic Castration-Resistant Prostate Cancer

Purpose Docetaxel has a demonstrated survival benefit for patients with metastatic castration-resistant prostate cancer (mCRPC); however, 10% to 20% of patients discontinue docetaxel prematurely because of toxicity-induced adverse events, and the management of risk factors for toxicity remains a challenge. Patients and Methods The comparator arms of four phase III clinical trials in first-line mCRPC were collected, annotated, and compiled, with a total of 2,070 patients. Early discontinuation was defined as treatment stoppage within 3 months as a result of adverse treatment effects; 10% of patients discontinued treatment. We designed an open-data, crowd-sourced DREAM Challenge for developing models with which to predict early discontinuation of docetaxel treatment. Clinical features for all four trials and outcomes for three of the four trials were made publicly available, with the outcomes of the fourth trial held back for unbiased model evaluation. Challenge participants from around the world trained models and submitted their predictions. Area under the precision-recall curve was the primary metric used for performance assessment. Results In total, 34 separate teams submitted predictions. Seven models with statistically similar area under precision-recall curves (Bayes factor ≤ 3) outperformed all other models. A postchallenge analysis of risk prediction using these seven models revealed three patient subgroups: high risk, low risk, or discordant risk. Early discontinuation events were two times higher in the high-risk subgroup compared with the low-risk subgroup. Simulation studies demonstrated that use of patient discontinuation prediction models could reduce patient enrollment in clinical trials without the loss of statistical power. Conclusion This work represents a successful collaboration between 34 international teams that leveraged open clinical trial data. Our results demonstrate that routinely collected clinical features can be used to identify patients with mCRPC who are likely to discontinue treatment because of adverse events and establishes a robust benchmark with implications for clinical trial design.

A community-based collaboration to build prediction models for short-term discontinuation of docetaxel in metastatic castration-resistant prostate cancer patients

Background Docetaxel has a demonstrated survival benefit for metastatic castration-resistant prostate cancer (mCRPC). However, 10-20% of patients discontinue docetaxel prematurely because of toxicity-induced adverse events, and managing risk factors for toxicity remains an ongoing challenge for health care providers and patients. Prospective identification of high-risk patients for early discontinuation has the potential to assist clinical decision-making and can improve the design of more efficient clinical trials. In partnership with Project Data Sphere (PDS), a non-profit initiative facilitating clinical trial data-sharing, we designed an open-data, crowdsourced DREAM (Dialogue for Reverse Engineering Assessments and Methods) Challenge for developing models to predict early discontinuation of docetaxel Methods Data from the comparator arms of four phase III clinical trials in first-line mCRPC were obtained from PDS, including 476 patients treated with docetaxel and prednisone from the ASCENT2 trial, 598 patients treated with docetaxel, prednisone/prednisolone, and placebo in the VENICE trial, 526 patients treated with docetaxel, prednisone, and placebo in the MAINSAIL trial, and 528 patients treated with docetaxel and placebo in the ENTHUSE 33 trial. Early discontinuation was defined as treatment stoppage within three months due to adverse treatment effects. Over 150 clinical features including laboratory values, medical history, lesion measures, prior treatment, and demographic variables were curated and made freely available for model building for all four trials. The ASCENT2, VENICE, and MAINSAIL trial data sets formed the training set that also included patient discontinuation status. The ENTHUSE 33 trial, with patient discontinuation status hidden, was used as an independent validation set to evaluate model performance. Prediction performance was assessed using area under the precision-recall curve (AUPRC) and the Bayes factor was used to compare the performance between prediction models. Results The frequency of early discontinuation was similar between training (ASCENT2, VENICE, and MAINSAIL) and validation (ENTHUSE 33) sets, 12.3% versus 10.4% of docetaxel-treated patients, respectively. In total, 34 independent teams submitted predictions from 61 different models. AUPRC ranged from 0.088 to 0.178 across submissions with a random model performance of 0.104. Seven models with comparable AUPRC scores (Bayes factor ≤; 3) were observed to outperform all other models. A post-challenge analysis of risk predictions generated by these seven models revealed three distinct patient subgroups: patients consistently predicted to be at high-risk or low-risk for early discontinuation and those with discordant risk predictions. Early discontinuation events were two-times higher in the high-versus low-risk subgroup and baseline clinical features such as presence/absence of metastatic liver lesions, and prior treatment with analgesics and ACE inhibitors exhibited statistically significant differences between the high- and low-risk subgroups (adjusted P < 0.05). An ensemble-based model constructed from a post-Challenge community collaboration resulted in the best overall prediction performance (AUPRC = 0.230) and represented a marked improvement over any individual Challenge submission. A Findings Our results demonstrate that routinely collected clinical features can be used to prospectively inform clinicians of mCRPC patients’ risk to discontinue docetaxel treatment early due to adverse events and to the best of our knowledge is the first to establish performance benchmarks in this area. This work also underscores the “wisdom of crowds” approach by demonstrating that improved prediction of patient outcomes is obtainable by combining methods across an extended community. These findings were made possible because data from separate trials were made publicly available and centrally compiled through PDS.

Abstract 772: The molecular landscape of dermal neurofibromatosis

Background: Neurofibromatosis type I (NF1) is a genetic disorder that disrupts neurological tissue growth and can lead to a diverse set of symptoms including systematic growth of benign tumors, learning disorders and bone deformities. It is a rare disease occurring in only 1 in 3,000 people worldwide. While the disease has been linked to loss of function in the NF1 gene - a known tumor suppressor - there is a high degree of phenotypic diversity in the NF1 patient population, making it difficult to identify the underlying cause of the disease and treat it effectively. In this work we seek to improve overall knowledge of dermal NF1 through global molecular characterization of the disease. Methods: We have collected four dermal neurofibromas and peripheral blood from each of 11 NF1 patients. We analyzed each sample using (1) Whole genome sequencing (WGS) on the Illumina HiSeq X platform, (2) Illumina OMNI2.5 Arrays (3) RNA-Sequencing on an Illumina HiSeq v4 machine and (4) iTRAQ-labeled proteomics. WGS data for both tumor and blood samples from each patient were used to identify patient-specific germ-line mutations as well as tumor-specific somatic mutations in each sample. Single nucleotide polymorphisms identified by the OMNI Arrays were used to identify copy number alterations in both blood and tumor samples. RNA-Seq data and proteomics data were mapped to transcripts and proteins respectively. Results: Preliminary analysis of this data illustrates a diverse genomic landscape of NF1. Hierarchical clustering of copy number alterations largely show samples clustering by tissue, suggesting that most copy number alterations are somatic and not shared across the germline. However, there are two patients that show germline copy number alterations, including one patient with loss in the NF1 region. WGS analysis suggests similar diversity with each patient possessing a distinct combination of germline and somatic mutations of NF1 and other cancer-related genes. Cluster analysis of the RNA-Seq data shows no patient-specific clusters, suggesting that that each tumor executes a unique transcriptional program. Conclusion: This work represents a first-ever attempt to profile the diversity of dermal neurofibromatosis at a molecular level. Preliminary analysis of the data underscores the complexity of this disease and explains, in part, previous difficulty in identifying effective treatments. Ongoing work includes expanding the analysis to include more patient samples and other types of NF1-derived tumors. As an orphan disease, NF1 has been poorly characterized compared to more common cancers. To rectify this, the Children9s Tumor Foundation and Sage Bionetworks are collaborating to make NF1 data available to the public to accelerate research and the drug discovery pipeline. We expect that this data will be a resource for other NF1 researchers to assist in the study of this disease at the molecular level. All data and preliminary results are publicly available at http://www.synapse.org/dermalNF Citation Format: Sara JC Gosline, Pamela Knight, Thomas Yu, Nripesh Prasad, Angela Jones, Shristi Shrestha, Braden Boone, Shawn E. Levy, Andrew J. Link, Allison C. Galassie, Hubert Weinberg, Stephen Friend, Salvatore La Rosa, Justin Guinney, Annette Bakker. The molecular landscape of dermal neurofibromatosis. [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 772.