Advanced scans could distinguish brain tumors from other damage

By: Deanna Chieco

Posted: 3/6/08

For a person with a brain injury, invasive and risky biopsies are usually needed in order to determine the type of lesion. However, with advances in neuroradiology, there are imaging techniques that are now being tested for their accuracy in detecting certain types of brain lesions.

In a recent article from the American Journal of Neuroradiology, researchers at Hopkins Hospital studied the effectiveness of different imaging methods in distinguishing between types of brain injuries. This study could allow for non-invasive diagnosis of brain tumors.

The group used two similar diagnostic tests, proton magnetic resonance spectroscopic imaging (MRSI) and perfusion magnetic resonance (MR), to differentiate between high-grade and low-grade tumors.

High-grade tumors are considered neoplastic, meaning they exhibit new, uncontrolled tissue growth - they are potentially cancerous. There are other types of brain lesions that are considered nonneoplastic, such as injuries from strokes or demyelination, the loss of neurons' insulation that is related to conditions such as multiple sclerosis.

Conventional magnetic resonance imaging (MRI) can be used to diagnose brain tumors. However, MRI generally gives a structural picture of the brain without information about metabolism and blood flow, which are also important for tumor diagnosis. MRI is unable to accurately differentiate between neoplastic and nonneoplastic masses.

MRSI is a technique that combines the structural data from conventional MRI with chemical information from nuclear magnetic resonance (NMR) spectroscopy.

The researchers hypothesized that MRSI should provide a high-resolution analysis of brain lesion anatomy, chemistry, and blood flow patterns, leading to a better ability to distinguish neoplastic from nonneoplastic lesions.

Perfusion MR, which provides information on cerebral blood flow and volume, would be used to improve or confirm a diagnosis made on the MRSI data.

In this study, 36 patients with high and low-grade brain tumors were studied as part of the group with neoplastic lesions. Additionally, 33 patients with nonneoplastic lesions, including stroke and multiple sclerosis, were studied. The diagnosis of all of these patients had already been made.

The researchers wanted to see how their new method would compare to the standard. They used conventional MR imaging to identify the region in which the lesion was located. Then perfusion MR and MRSI were performed on patients.

Researchers compared various chemical ratios and blood volumes to better characterize the previous diagnosis. They also tried to establish cutoff values from MRSI and perfusion MR data that would indicate the type of lesion and could be used as diagnostic standards.

MRSI was most successful in distinguishing between high-grade and low-grade tumors. Perfusion MR was also effective in tumor discrimination, yet it could not distinguish between low-grade and benign tumors. Using these methods, the researchers were able to confirm the original diagnosis 84 percent of the time.

Also, MRSI proved effective in diagnosing patients with demyelination, often the result of multiple sclerosis. If these types of lesions cannot be detected with conventional MRI, MRSI provides important metabolite data that can suggest a diagnosis.

Future research into the new applications of these imaging techniques could lead to better diagnostic tools for brain lesions, perhaps minimizing the need for invasive and dangerous biopsies.