One of the constant barriers that physicians, and their patients, seem to encounter is the battle against time. In the case of tumor treatments and diagnosis, researchers have published a study that could save time to recovery if a treatment isn't working for a patient – via a new way of observing a standard MRI.
As early as one week after beginning treatment for brain tumors, a new imaging analysis method was able to predict which patients would live longer, researchers from the University of Michigan Comprehensive Cancer Center (Ann Arbor) have found.
The method uses a standard magnetic resonance imaging (MRI) protocol to monitor changes over time in tumor blood volume within individual voxels of the image, rather than a composite view of average change within the tumor. This parametric response map (PRM) allowed researchers to see specific areas in which tumor blood volume increased or decreased, that may have canceled each other out when looking at the changes as an average.
Results of the study appear in the advance online edition of Nature Medicine.
Craig Galban, PhD, assistant professor of radiology at the U-M Medical School and the study author, told Medical Device Daily that this could be a "huge paradigm shift" if clinicians go in this new direction. Galban describes the potential for this imaging analysis as "adaptive therapy."
The researchers looked at 44 people with high-grade glioma, a type of brain tumor, who were treated with chemotherapy and radiation. Each participant underwent MRIs before treatment, and one week and three weeks after starting treatment. The researchers then looked at the relative cerebral blood volume and the relative cerebral blood flow of the tumor to analyze voxel-wise changes among the serial scans.
Looking at standard comparisons using averages, the scans indicated no change one week and three weeks into treatment. But, using the parametric response map approach, the researchers were able to show changes in the tumor's blood volume and blood flow after one week that corresponded to the patient's overall survival.
"Right now, physicians have to wait until the treatment is over to determine if it was effective. We are proposing that we can determine right away if the treatment is effective. If it is not working, then we can tailor the therapy rapidly to the individual. We are trying to personalize the care and get this valuable and accurate information to the clinician so that they can make a decision on the proper course of treatment for the patient," Galban told MDD.
High-grade gliomas have a high-mortality rate, with people surviving only an average of 12 months after diagnosis. Typically, patients receive six to seven weeks of treatment, followed by a traditional MRI scan six weeks after completing therapy to determine if the tumor shrank. If the cancer did not respond to the treatment, a new approach may be tried.
A voxel is a volume element, representing a value on a regular grid in 3-D space, as compared to a pixel which is 2-D. As with pixels, voxels themselves typically do not contain their position in space (their coordinates) — but rather, it is inferred based on their position relative to other voxels (i.e., their position in the data structure that makes up a single volume image).
Galban explained the differences in standard imaging analysis compared to PRM for MDD. "It's really looking at voxel-wide changes," he said. "PRM is measuring the changes in blood volume in these individual voxels. The standard way to look at the functional data [such as an MRI on a tumor] is to take all the information and gather it into a mean, or average. We are proposing not just to look at the statistical average of all the voxels and blood volume, but to examine them discreetly and individually, and determine spatially how much of the tumor did respond to treatment. These are highly heterogeneous tumors. Some of the tumor might have responded well to the treatment, but another part of the tumor may not. So that's how BPM differs — it is really looking at changes in the blood volume in the individual voxels. We're interested in how much of that tumor volume changes. "
The researchers believe this approach might also be useful with other imaging techniques such as PET and CT scans.