Quantification and expert evaluation of evidence for chemopredictive biomarkers to personalize cancer treatment
Metrics: PDF 2279 views | HTML 2738 views | ?
Shruti Rao1,*, Robert A. Beckman1,2,3,*, Shahla Riazi1, Cinthya S. Yabar4,5, Simina M. Boca1,2,3, John L. Marshall2,6, Michael J. Pishvaian2,6, Jonathan R. Brody4 and Subha Madhavan1,2
1Innovation Center for Biomedical Informatics, Georgetown University Medical Center, Washington, D.C., USA
2Department of Oncology, Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, Washington, D.C., USA
3Department of Biostatistics, Bioinformatics, and Biomathematics, Georgetown University Medical Center, Washington, D.C., USA
4Pancreas, Biliary and Related Cancer Center, Department of Surgery, Thomas Jefferson University, Philadelphia, PA, USA
5Department of Surgery, Albert Einstein Medical Center, Philadelphia, PA, USA
6Otto J. Ruesch Center for the Cure of Gastrointestinal Cancer, Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, Washington, D.C., USA
*These authors have contributed equally to this work
Subha Madhavan, email: [email protected]
Keywords: precision medicine, predictive biomarkers, clinical utility, biocuration, evidence framework
Received: June 07, 2016 Accepted: November 12, 2016 Published: November 24, 2016
Predictive biomarkers have the potential to facilitate cancer precision medicine by guiding the optimal choice of therapies for patients. However, clinicians are faced with an enormous volume of often-contradictory evidence regarding the therapeutic context of chemopredictive biomarkers.
We extensively surveyed public literature to systematically review the predictive effect of 7 biomarkers claimed to predict response to various chemotherapy drugs: ERCC1-platinums, RRM1-gemcitabine, TYMS-5-fluorouracil/Capecitabine, TUBB3-taxanes, MGMT-temozolomide, TOP1-irinotecan/topotecan, and TOP2A-anthracyclines. We focused on studies that investigated changes in gene or protein expression as predictors of drug sensitivity or resistance. We considered an evidence framework that ranked studies from high level I evidence for randomized controlled trials to low level IV evidence for pre-clinical studies and patient case studies.
We found that further in-depth analysis will be required to explore methodological issues, inconsistencies between studies, and tumor specific effects present even within high evidence level studies. Some of these nuances will lend themselves to automation, others will require manual curation. However, the comprehensive cataloging and analysis of dispersed public data utilizing an evidence framework provides a high level perspective on clinical actionability of these protein biomarkers. This framework and perspective will ultimately facilitate clinical trial design as well as therapeutic decision-making for individual patients.
All site content, except where otherwise noted, is licensed under a Creative Commons Attribution 4.0 License.