Case Studies

Biostatistics in Oncology Research

Significant Findings in the Field of Oncology Research

Goal

Enable oncology researchers to quickly conduct deep statistical analysis on large datasets.

Featured Organizations

Mary Babb Randolph Cancer Center, National Cancer Institute, and West Virginia University

Industry

Medical Research

Background

For nearly a decade, Parabon has assisted many medical researchers with groundbreaking scientific, medical and psychosocial oncology research projects by enabling them to rapidly and comprehensively analyze immense datasets. The areas of research have included such diverse fields as nutritional oncology, immunotherapy, gene therapy, angiostatics, pharmacokinetics, protein structure prediction, and new uses of traditional treatments.

Below are two examples of how researchers use Parabon's Computation on Demand® service and the Compute Against Cancer® Network were to analyze billions of combinations of data.

-----

Project: Expression of Repair Genes in Ovarian Cancer Tumors

Mary Babb Randolph Cancer Center

In prior research studies, Mary Babb researchers had observed that certain types of ovarian cancer were particularly resistant to chemotherapy treatment. Researchers suspected the tumor cells had an ability to quickly repair themselves at the DNA level. Having identified two genes known to play a key role in such repair, called ERCC1 and XPB, the research team systematically tested for an abundance of these genes in cancerous ovarian tissues. After carefully measuring each specimen, the team performed the standard suite of statistical inquiries. No significant results were identified until they partnered with Parabon and took their analysis to the next level.

Analysis and Results

Data was statistically analyzed using thousands of computers participating in the Compute Against Cancer Network. With so much computation available, the calculations were concluded in a matter of days. Using traditional computing methods the analysis would have taken months or even years. The results, according to the principal investigator, Eddie Reed, MD, were "delightfully eye-opening. For the first time we programmatically examined each of the billions of possible combinations of factors to find the most [statistically] powerful models. As a result, we found relationships that we had not anticipated."1

-----

Project: Patient Quality of Life

West Virginia University's (WVU) Blood and Marrow Transplantation Program and WVU's School of Pharmacy

Research teams collected data through clinical studies to explore how factors such as combinations of medications, routes and methods of drug administration affected the quality of life of chemotherapy patients. With over 34 billion relationships possible within the data, the researchers didn't think it was possible to analyze the entire dataset.

Analysis and Results

Using the Computation on Demand service, via the Compute Against Cancer program, WVU researchers teamed with Parabon to analyze the data. According to Dr. Karen Balzer, one of the researchers, access to the massive computation helped find correlations that would not have been found otherwise.

Conventional wisdom about drug treatment holds that intravenous (IV) administration of drugs is generally more effective than oral administration, however the data from this study shows there is strong evidence that oral antiemetic drugs are associated with a higher quality of life than similar IV drugs. This information may lead to less suffering for chemotherapy patients and reduce the overall cost of patient care, because the cost of oral drugs is lower than the cost of IV drugs.

-----

Future Endeavors

Professor Peter Boyle from the International Agency for Research on Cancer (IARC) recently predicted that, in the year 2030, there would be 27 million new cancer cases diagnosed, 17 million cancer deaths, and 75 million people alive with cancer throughout the World2 if current trends continue. Clearly there is a need to accelerate discovery in the field of oncology research.

Parabon Computation, via its Computation on Demand service and the Compute Against Cancer Network, provides an enabling technology to speed discovery and boost oncology research endeavors in a wide array of research fields including, but not limited to: drug discovery, pharmacokinetics, therapeutics, large-scale genomic sequencing, protein folding simulations, microarray gene expression patterns, data mining and biostatistics.

Parabon is dedicated to aiding oncology researchers in their quest for a cure. If you are an oncology researcher and would like to learn more about how Parabon can assist you, please contact us.

-----

1 Eddie Reed, Jing Jie Yu, Antony Davies, James Gannon and Steven Armentrout. Clear Cell Tumors Have Higher mRNA Levels of ERCC1 and XPB Than Other Histological Types of Epithelial Ovarian Cancer, Clinical Cancer Research Vol. 9, 5299-5305, 1 Nov 2003.

2 eCancerMedicalScience. Oncology Times — European Cancer Conference published by Lippincott Williams & Wilkins, Inc, 11 December 2007.

Who Uses Frontier? (Widget)

Who Uses Frontier?

Grid Computing for Scientists & Researchers

"Frontier speeds discovery. Simulations finish in hours versus months, transforming the way we conduct research."

Learn More

Grid Computing for ISVs & Developers

"Getting my application to run on Frontier was easy! Frontier's computational power differentiates my products in the market."

Learn More

Grid Computing for Systems Integrators

"Frontier is a proven COTS solution. It's secure, easy to deploy and requires little administration."

Learn More

Grid Computing for Government Agencies

"Frontier Enterprise increased our HPC capability without impacting the power, space, and cooling issues in our data center."

Learn More

Grid Computing for Commercial Enterprises

"Frontier greatly improved the ROI of our existing IT infrastructure. It's a smart and cost-effective choice."

Learn More

Parabon's computational grid allowed us, for the first time, to [programmatically] examine each of the billions of possible combinations of factors to find the most [statistically] powerful models. As a result, we found relationships that we had not anticipated." Eddie Reed, MD, Oncology Researcher

Access to massive computations helped find correlations that we were not otherwise able to consider." Dr. Karen Balzer, Medical Researcher at WVU