Neuroimaging is progressing and new techniques are continually being developed and refined; providing neuroradiologists with a premier window and clinically relevant insight to many complex diseases of the brain.

Neuroimaging is progressing and new techniques are continually being developed and refined; providing neuroradiologists with a premier window and clinically relevant insight to many complex diseases of the brain.

Proving to be of great importance in the diagnostic and therapeutic evaluation of brain tumors, this rapidly evolving methodology not only provides conventional structural imaging, it allows neuroradiologists to further their expertise in functional neurobiology.

Nevertheless, "local practice customs and clinical questions will dominate how you do this," says Dr. Mark Mullins, Assistant Professor of Radiology at Emory University.

Presenting a learning centered keynote lecture at the 2008 American Roentgen Ray Society meeting, Dr. Mullins took those in attendance through the pearls and pitfalls of advanced MR imaging techniques in distinguishing treatment effects from recurrent brain tumors.

Distinguishing treatment effects in brain tumors is one of the most important aspects in a patient's treatment. Due to its physiological and biochemical basis, the utilization of MR imaging for radiotherapy planning and identifying post treatment effects is the preferred method at this time.

Glioblastoma multiforme is one of the most common and most malignant of the glial tumors. Studies show that of an estimated 17,000 primary brain tumors diagnosed in the United States each year, approximately 60 per cent are gliomas.

When utilizing the many tools for assessing and calculating post treatment effects in primary high grade brain tumors, Mullins says, "one of the most underutilized things in our clinical tool kit is morphology." Although, he adds, "as there is no cure for WHO grade classification III and IV primary high grade brain tumors, stability is a definite win."

Modern MRI techniques in neuroimaging have revolutionized the practice of neuroradiology. While many advanced methods may still exist predominantly within the research arena, the imaging modalities we can expect to see within our future imaging arsenal include MR perfusion weighted imaging, MR spectroscopy, and diffusion tractgraphy imaging. In the very advanced imaging arena, Mullins states, "by the time we meet again in Boston for ARRS 2009 we can expect to see molecular neuroimaging."

As we approach the new horizons in advanced MR imaging, before using any modality, Mullins advises, "the question you should always ask yourself is, is it technically adequate?"

Some of the most remarkable advancements in neuroimaging research have been fueled by increasingly sophisticated neuroimaging technology. Recently, these innovative techniques are allowing researchers to develop extraordinary investigations.

How much variability do we see in measurements of tumor size in brain tumors? This is cause for significant concern if we take into account that bi-dimensional tumor measurements representing a greater than 25 per cent increase in tumor size are generally accepted as indicating tumor progression by neuro-oncologists. Dr. James Provenzale, Professor of Radiology and Division Chief at Duke University Medical School, assessed the interobserver variability of such measurements.

"Measurements are critical to patient management as well as patient lives," states Provenzale.

Under institutional review board protocol, Dr. Provenzale et al enrolled 22 patients with contrast-enhancing high-grade brain tumors. Using five mm slice thickness images, all patients underwent contrast-enhanced MR imaging in three orthogonal planes. For each patient, eight board certified radiologists measured tumor diameters in each plane on digital images on two MR scans. So that the radiologists would forget the initial measurement, the scans were measured a month apart. No patient had a change in therapy during the one month interval. Variance components models were run on the diameter products and the ratio of consecutive maximal diameter products. These components included patient examination effect, reader effect and residual effect.

Researchers found complete agreement among readers in 10 cases, all indicating stable disease. The remaining 12 patients were judged by at least one reader to have progressive disease. Results found that the variance component due to reader was small however; the residual variance component was large indicating that repeat measurements of the same case are variable even with the same reader.

"Previous studies have only shown moderate agreement in measurement," says Provenzale.

Dr. Provenzale concluded that substantial inter-reader disagreement, as well as the very nature of the measurement method, is associated with a moderately high rate of false positive readings. "The Macdonald criteria, as we use them, are not robust or reliable," states Provenzale. He also suggests, "these findings should be considered when using these widely accepted tumor progression criteria."