Poster viewing and lunch

76P - Using Bioinformatics and Artificial Intelligence (AI) to Map the Cyclin-dependent Kinase 4/6 Inhibitor (CDK4/6i) Translational Biomarker Landscape (ID 295)

Lecture Time
12:15 - 12:15
Session Name
Poster viewing and lunch
Room
Exhibition area
Date
Fri, 12.05.2023
Time
12:15 - 13:00
Speakers
  • Kim Wager (Tubney, United Kingdom)
Authors
  • Kim Wager (Tubney, United Kingdom)
  • Yao Wang (New York, NY, United States of America)
  • Andrew Liew (Melbourne, Australia)
  • Dean Campbell (New York, NY, United States of America)
  • Lynsey M. Fettig (Newtown, PA, United States of America)
  • Feng Liu (San Diego, CA, United States of America)
  • Jean-Francois Martini (San Diego, CA, United States of America)
  • Niusha Ziaee (New York, NY, United States of America)
  • Yuan Liu (San Diego, CA, United States of America)

Abstract

Background

CDK4/6i plus endocrine therapy is standard of care for patients with hormone receptor-positive/human epidermal growth factor receptor-negative advanced or metastatic breast cancer; however, patients frequently develop resistance to therapy. Identifying biomarkers that are associated with CDK4/6i resistance or response is key to selecting the most appropriate patients for treatment and could be used to develop novel agents or combination therapies. The aim of this study was to comprehensively describe the CDK4/6i biomarker landscape to guide future research.

Methods

We used a combination of bioinformatics and AI techniques to examine and collate putative biomarkers. Data were mined from Uniprot to identify proteins related to CDK4/6. Biomarkers used in clinical trials were mined from the PharmaProjects database. Protein interaction networks were generated using Signor, STRING, and Bioplex 3.0. To explore relevant publications, we performed a Geneshot analysis and used the output to conduct an enrichment analysis using the MSigDB database. To find non-protein biomarkers such as non-coding RNA species or molecular signatures, we used deep learning text analytics to interrogate more than 50 million documents across several platforms including PubMed, medRxiv, and bioRxiv. Results were prioritized by manual curation and target tractability analysis.

Results

Bioinformatic database analyses identified 876 genes with known Uniprot functions potentially related to CDK4/6i response in breast cancer. Additionally, genes identified by Geneshot were contained within 42 oncogenic gene sets; 5 of these gene sets contained > 60 of the identified genes, with the most significant dysregulated pathways being TBK1, Rb, PDGF, SNF5, cyclin D1, Raf, and p53. Following target tractability analysis and literature review, we prioritized ∼20 promising candidates, including ERBB2, AKT1, and PIK3CA.

Conclusions

The compendium developed using our multi-approach, AI-assisted review of existing knowledge of the CDK4/6i biomarker landscape will be central in supporting next generation research and drug development for breast cancer.

Editorial acknowledgement

Editorial support, conducted in accordance with Good Publication Practice (GPP3) and the International Committee of Medical Journal Editors (ICMJE) guidelines, was provided by Oxford PharmaGenesis, Inc., Newtown, PA with funding provided by Pfizer Inc.

Legal entity responsible for the study

Pfizer Inc.

Funding

Pfizer Inc., USA.

Disclosure

K. Wager: Financial Interests, Personal, Writing Engagements, are employees of Oxford PharmaGenesis, Inc. who was funded by Pfizer Inc. to conduct the study.: Oxford PharmaGenesis. Y. Wang: Financial Interests, Personal, Stocks/Shares, are employees and stockholders of Pfizer Inc.: Pfizer Inc. A. Liew: Financial Interests, Personal, Writing Engagements, are employees of Oxford PharmaGenesis, Inc. who was funded by Pfizer Inc. to conduct the study.: Oxford PharmaGenesis. D. Campbell: Financial Interests, Personal, Stocks/Shares, are employees and stockholders of Pfizer Inc.: Pfizer Inc. L.M. Fettig: Financial Interests, Personal, Writing Engagements, are employees of Oxford PharmaGenesis, Inc. who was funded by Pfizer Inc. to conduct the study.: Oxford PharmaGenesis. F. Liu: Financial Interests, Personal, Stocks/Shares, are employees and stockholders of Pfizer Inc.: Pfizer Inc. J. Martini: Financial Interests, Personal, Stocks/Shares, are employees and stockholders of Pfizer Inc.: Pfizer Inc. N. Ziaee: Financial Interests, Personal, Stocks/Shares, are employees and stockholders of Pfizer Inc.: Pfizer Inc. Y. Liu: Financial Interests, Personal, Stocks/Shares, are employees and stockholders of Pfizer Inc.: Pfizer Inc.

Collapse