Researchers from the University of Georgia College of Public Health have received a two-year, $358,875 grant from the National Cancer Institute at the National Institutes of Health to develop novel methods for converting promising predictive cancer biomarkers into clinical tests that can be used to guide patient therapeutic decisions.

Kevin Dobbin and Xiao Song, both associate professors in the Department of Epidemiology and Biostatistics, are principal investigators on the grant.

Recent advances in biotechnology and bioinformatics are producing a wide array of novel targeted anti-cancer therapies,” Dobbin said. “Many of these therapies target specific biological pathways in the tumor, and predictive biomarkers can be used to identify a tumor’s weak spots and which therapies will work best. But developing and validating these biomarkers often requires rare and valuable patient specimens with associated clinical follow-up data.”

Dobbin and Song will develop statistical methods for validating biomarkers that do not require these valuable patient specimens. Their aim is to clear the current research logjam of potential biomarkers and allow more of these valuable clinical decision-making tools to be made available to patients.

The challenges faced by researchers developing cancer biomarkers are many, Dobbin explained. For some studies, it is often desirable to perform a “platform migration” and move an assay from a technologically complex platform to a process that can be used in the clinic.  Another common occurrence is that a test originally developed in one laboratory needs to be exported to multiple labs because a single lab cannot handle the throughput needed for a clinical trial.

In addition, researchers often must select from a number of competing technologies that measure the same biomarker, some of which are more expensive than others, he said.

“All these problems can potentially hinder or halt the process of biomarker development, because the measurement process has to be changed,” said Dobbin. “This means the modified biomarker has not been validated and any previous associations between the biomarker and clinical outcomes must be re-investigated.”

The biostatistical methods developed by Dobbin and Song will allow the clinical performance of the modified biomarker to be estimated using data from an assay reproducibility study. The reproducibility study will compare the original biomarker and the modified biomarker on a set of inexpensive and widely available patient samples that don’t require costly clinical follow-up.

“If this application is successful, cancer researchers will no longer need to feel locked into an assay as it was originally developed, or be forced to abandon an assay because modifications are required,” said Dobbin.  “Predictive biomarker assays will therefore be developed more quickly and easily, allowing the benefits they provide to public health, patients and physicians to be realized.”

The research is being supported by the National Cancer Institute of the National Institutes of Health under award number 1R21CA201207-01.

Posted January 20, 2016.