MR diffusion maps track response to tumor therapy

December 3, 2004

With brain cancer, time isn't money. Time is death. A new diffusion-based MR scanning protocol that speeds up the assessment of therapeutic strategies could dramatically improve the odds of success.

With brain cancer, time isn't money. Time is death. A new diffusion-based MR scanning protocol that speeds up the assessment of therapeutic strategies could dramatically improve the odds of success.

Half of the 20,000 people in the U.S. diagnosed with gliomas each year die within 18 months. Using accepted CT imaging practices, an oncologist has no more than two opportunities to make a correct therapeutic choice. Appropriate therapy or dosage defines the difference between death and survival, Dr. Bradford A. Moffat, an assistant professor of radiology at the University of Michigan, said in a study presentation at the Society for Molecular Imaging meeting in September.

Although glioma treatment will always be a race against time, success with a new MR diffusion-based test of therapeutic response may improve the odds of winning. The 10-minute protocol, first proposed in 1994 by the study's principal investigator, Brian Ross, Ph.D., uses apparent diffusion coefficient (ADC) mapping to determine whether chemotherapy or radiation therapy is effective. Unlike anatomic MR or CT, which can't reliably show therapeutic efficacy until at least a full month after completion of a four-week regime of therapy, ADC mapping produces results a week before the end of a therapeutic regime.

The diffusion approach works off changes in the density and permeability of brain tumors as cells die and deteriorate during a positive response to therapy, Moffat said. The ADC map of the entire brain is acquired on a 1.5T scanner, using a single-shot, echo-planar sequence, and is coregistered with FLAIR, T1-, and T2-weighted imaging, which provide anatomic landmarks.

While adjusting for therapy-induced shrinkage, software automatically performs a voxel-by-voxel comparison of diffusion data acquired in studies performed before and three weeks after the start of treatment. Tissue showing a significant diffusion increase, corresponding with a positive response, is displayed in red. Tissue with a significant diffusion decrease is presented in blue, and tissue with relatively little change is shown in green. Images can be configured in either 2D slices or 3D volume renderings for interpretation.

In his presentation, Moffat reported results from 20 patients with various types of glioma. Treatment involved chemotherapy or radiation therapy or both.

A comparison of results of baseline and diffusion mapping performed three weeks after the initiation of therapy indicated that six patients partially responded to therapy, six had stable disease, and eight had progressive disease.

The results were highly accurate, Moffat said. Diffusion mapping correctly singled out all partial responders and differentiated them from patients with stable and progressive disease. The sensitivity and specificity rates for depicting the therapeutic response were both 100%.

Looking at the individual cases, Moffat found that nonresponders showed few changes in diffusion anywhere in their tumors. In the responders, little change was observed in some regions of heterogeneous tumor, while other areas showed substantial response. In most cases, the response was heterogeneous, indicating that some sections responded more than others. Survival data were not collected, however.

Diffusion maps could potentially allow an oncologist to modify the treatment regime, according to Moffat. Chemotherapy dosages could be increased in some cases; in others, different chemotherapies might be prescribed. For patients undergoing radiation therapy, the 3D maps could also be used to redirect radiation to the nonresponding regions.

Over the entire course of the disease, oncologists could have more opportunities than with conventional CT to consider treatment options, Moffat said.

"With diffusion imaging, we can accelerate the patient from protocol one to protocol two and perhaps still have time to try one, two, or three additional types of therapy," he said. "Imaging will not only play a critical role in determining which patients should move on to these therapies, but it will influence dose scheduling as well."

The ADC mapping technique may also help track the therapeutic response of head and neck, prostate, and breast cancer, Moffat said, but larger trials are needed to confirm these early results.