Accurate estimations of disease progression and treatment response are crucial in personalized therapy. High-throughput microarray technologies have the potential to allow molecular prediction. Nevertheless, there have been few gene expression-based tests applied in clinics for disease intervention. This fact puts a premium on developing innovative methodologies to embed biological relevance into biomarker identification. Limited research has addressed combined contributions of clinical, pathological, and demographic factors in predicting treatment outcome with large-population data. This project will develop a novel computational methodology to model disease-mediated genome-scale co-expression networks and complex crosstalk with signaling pathways for biomarker identification. The gene expression-based biomarkers identified from microarray data will be validated with quantitative real-time RT-PCR for clinical applications. This framework will be applied to prognostication of lung cancer, and a clinically applicable model will be developed to predict the risk for tumor recurrence in early-stage patients and to select patients for adjuvant chemotherapy. Lung cancer is difficult to manage clinically, and tumor recurrence is the major cause of treatment failure and death. We hypothesize that a combination of network-based genome-wide studies, real-time RT-PCR validation in independent lung cancer tumor samples, and population approaches will improve treatment decision-making.
In Aim 1, a novel network-based methodology will be developed to identify prognostic biomarkers from microarray data. Candidate genes co-expressed with major lung cancer signaling genes will be pinpointed from genome-wide co-expression networks specifically associated with different prognostic patient groups. From these candidate genes, feature selection algorithms will be used to identify the best prognostic gene signatures.
In Aim 2, quantitative real-time RT-PCR microfluidic low-density arrays will be designed to validate and refine the gene signatures identified in Aim 1. A multi-gene assay will be developed based on RT-PCR profiling of tumor samples collected from multiple cancer centers and clinical trials. The gene expression patterns generated in DNA microarrays and real-time RT-PCR assays will be compared.
In Aim 3, a prognostic categorization scheme will be designed by combining clinical, pathological, demographic, and comorbidity factors using large patient population data retrieved from SEER (Surveillance, Epidemiology, and End Results) and linked SEER-Medicare databases. This scheme will be integrated with the multi-gene model constructed in Aim 2, to provide personalized prognosis. This multidisciplinary research will actively engage bioinformaticians, clinicians, and biomedical researchers throughout model development and evaluation, and will develop web- based clinical prediction software with broad applications to diverse treatment modalities. This project will improve personalized clinical care and overall survival of lung cancer patients, as well as advance translational informatics and have important impacts on personalized medicine in general.

Public Health Relevance

This project will develop a clinically applicable model to predict the risk for tumor recurrence in early-stage lung cancer patients and to select patients for adjuvant chemotherapy. The project results will improve personalized clinical care and overall survival of lung cancer patients.

Agency
National Institute of Health (NIH)
Institute
National Library of Medicine (NLM)
Type
High Priority, Short Term Project Award (R56)
Project #
2R56LM009500-04
Application #
8291472
Study Section
Biomedical Library and Informatics Review Committee (BLR)
Program Officer
Ye, Jane
Project Start
2008-09-15
Project End
2014-09-14
Budget Start
2012-09-15
Budget End
2013-09-14
Support Year
4
Fiscal Year
2012
Total Cost
$260,000
Indirect Cost
$84,324
Name
West Virginia University
Department
Public Health & Prev Medicine
Type
Schools of Medicine
DUNS #
191510239
City
Morgantown
State
WV
Country
United States
Zip Code
26506