New Horizons lecturer touts opportunities for imaging biomarkers

Academic radiologists usually cut to the chase in describing how imaging increasingly factors into drug discovery and research. They display cellular metabolism FDG-PET images or dynamic-contrast MR images quantifying changes in tumor vascularity. These glamor children of radiological research reflect the potential of medical imaging in measuring therapeutic response.

In his Pendergrass New Horizons lecture at the RSNA meeting Monday, Dr. Lawrence H. Schwartz focused on the larger context of imaging and drug discovery, extending the discussion to imaging's influence on clinical practice. For Schwartz, director of MRI at Sloan-Kettering Institute, diagnostic imaging in drug testing is more than a measurement of therapeutic response. It also functions as a disruptive technology that promises to rewrite the rules for tracking whether drugs produce desired effects.

This role should not intimidate radiologists, for imaging has been used to assess the response to therapy for years, Schwartz said. Chest radiography reveals the effect of antibacterial therapy upon pneumonia. Radiologists use MRI to measure the elimination of white matter plaques indicative of a positive response in multiple sclerosis patients to the drug Betaseron. MR angiography or CTA can be used to measure changes in the degree of cardiovascular stenosis following the administration of lipid-lowering drugs.

Additional demands are placed on the efficacy of medical imaging, however, when it is applied to the testing of investigational drugs. Imaging in this context must directly visualize physiological processes associated with a disease and quantify the changes in those processes brought about by the investigational therapy, Schwartz said.

In this way, imaging serves as a biomarker of responses. Traditionally, anatomic changes have served as biomarkers. In oncology, for example, the Response Evaluation Criteria in Solid Tumors and the World Health Organization have defined how reductions in the diameter of tumors, measured with CT, correspond with the complete, partial, stable, and progressive responses to cancer therapies.

Combining functional and anatomic measurement boosts the predictive power of individual biomarkers applied individually. Looking at investigational applications, F-18 fluorothymidine PET can provide a quantifiable measurement of changes in cell proliferation as a measure of response to cancer therapies. Dr. Malik E. Juweid demonstrated that F-18 FDG-PET performed with CT generates a more accurate assessment of the response of lymphoma to chemotherapy than does CT alone (J Clin Oncol 2005;23(21):4577-4580). Volumetric imaging, facilitated with multislice CT and 3D reconstruction postprocessing, has been shown to be a more powerful predictive instrument than simple measurement of tumor diameter for assessing tumor response in some instances, Schwartz said.

Anatomic measurement has proven inadequate for assessing response to molecularly targeted therapies, according to Schwartz. The adoption in drug design of genetic and proteomic pathways to disease treatment has produced a new generation of cytostatic agents that kill cancers without necessarily affecting volume. Drugs targeted to specific protein kinases require surrogate imaging markers tuned to cell metabolism or proliferation to measure response and calculate optimal dosages.

This line of thinking brought Schwartz full circle. Some FDG-PET and dynamic contrast-enhanced MRI strategies offer a readout of response within 24 hours after the initiation of therapy, he said. Several problems must be addressed, however, if imaging is to realize its potential in the discovery of new drugs and their clinical testing. New biomarkers are needed, especially to complement targeted therapies. Existing biomarkers can be improved by producing more robust imaging techniques and refining image analysis. Standardized imaging techniques, robust image analysis tools to create databases, image repositories, and tools for data analysis are all necessary.

No individual, department, or corporation can completely validate an agent on its own, Schwartz said. Careful collaboration is required among academia, clinical practices, government, and industry.

"Together with engagement of the entire radiology community, we can make a reality of a rapid progress of biomarkers utilizing imaging," he said.

For more online information, visit Diagnostic Imaging's RSNA Webcast.