Half of the patients with brain tumours who were enrolled in the fMRI study had their treatment strategies altered after the critical motor and language regions of the brain were identified prior to surgery. "In patients who underwent fMRI, neurosurgeons estimated that more tumour was removed at surgery, operations were shorter and skull incisions were smaller, compared with what they would have had if the fMRI data were not available," said Jeffrey Petrella, associate professor of radiology in the neuroradiology division at Duke.
In the study, 39 patients diagnosed with potentially operable brain tumours were evaluated with fMRI of the brain (Radiology 240 793). During the imaging exam, they performed sentence-completion tasks to map areas of the brain involved in language function, and hand-squeezing tasks to detect sensory motor areas. The entire procedure, including the time it took to train patients and acquire the images, took less than one hour.
The researchers found that in 19 of 39 patients, the surgeons significantly changed their treatment plans after receiving the fMRI findings. Of the 19 patients, 18 underwent more aggressive approaches than originally planned.
None of the patients in the study showed post-operative speech or motor deficits and Petrella feels that fMRI offers real benefits to patient outcome. He told medicalphysicsweb: "Though we have not directly proved that this new information is of benefit to patients, the fact that our fMRI request volume continues to increase over time indicates that the technology has gained acceptance - which suggests a benefit on some level."
Despite its apparent benefits, however, fMRI is still not widely used. "In the US, I would estimate that about 50 centres are doing it routinely in the clinical environment, yet there are likely well over a thousand centres with the capability to do it," said Petrella. He believes that the major hurdle to wider take-up is not the availability of hardware or software but a lack of experience of how to administer a task, and then process and interpret the images.
Petrella acknowledges that future studies focusing on the outcome of patients who undergo preoperative fMRI imaging are still needed, though such investigations will be difficult to perform for ethical reasons. The next step in his research would be to assess whether patient outcomes are actually improved in a large multicentre trial with a control group. "However, this sort of trial raises ethical concerns, especially for neurosurgeons who believe that fMRI could have a beneficial effect on their patients," he said. "Such neurosurgeons may feel uncomfortable denying the technology to a given patient solely on the basis of randomization to the control group who does not undergo fMRI."
Briefing: What is fMRI?
While MRI is a technique for examining brain anatomy, fMRI allows clinicians to examine the brain in action (i.e. while performing a particular task). fMRI can be performed on a 1.5 T MR scanner. However, the scanner must be equipped with strong and fast gradients capable of performing echo-planar imaging.
Many, if not most, state-of-the-art commercial systems are capable of this. In addition to the scanner, it must be possible to present stimuli to patients (e.g. using a within-scanner projection system or a set of MR-compatible audiovisual goggles).
fMRI uses rapid echoplanar or spiral imaging - of the order of 20 brain slices per second - to image the entire brain repeatedly every 1–2 s. The images are sensitive to changes in blood oxygenation, which reflects the level of brain activity in a particular region while performing a task.