MR Imaging of Brain Tumors - Recent Developments

Gilbert Vezina, M.D., Director of Neuroradiology - Children’s National Medical Center, Washington D.C.




Imaging techniques to diagnose, stage, and follow patients with brain tumors are central to their clinical management. Magnetic resonance imaging (MRI) is the most commonly utilized technique for lesion detection, definition of extent, detection of spread and in evaluation of either residual or recurrent disease. This article will review the basic principles and recent developments in MR technology pertinent to patients with brain neoplasms.


MR images result from the excitation of hydrogen protons by radiofrequency (RF) pulses. The MRI machine generates very brief (about 1 millisecond in duration) RF pulses; these RF pulses excite hydrogen protons, and elevate them to a higher energy state. As the protons return to a lower ("resting") energy state, they release electromagnetic energy. (This process, called "relaxation", takes anywhere from a few milliseconds to a few seconds, depending on the kind of molecule (tissue) where the protons are located.) This energy is picked up and amplified by the magnet’s antennae (coils), and turned into visual display (images).


Until recently MRI was used to characterize cerebral neoplasms by:


1. Demonstration of anatomy in various planes. 2. Display of differences in relaxation times (so-called T1 and T2 relaxation times) between normal and abnormal tissues. 3. Detection of breakdown in the blood brain barrier on T1 weighted images with use of intravenous paramagnetic contrast enhancement (gadolinium accumulates in central nervous tissues that lack an intact blood brain barrier, a situation found in many neoplastic lesions).


In recent years, technological advances (improved hardware configuration and novel pulse sequences) have opened new windows in our ability to detect and characterize tumors. These new advances include rapid/ultrafast imaging, metabolic and functional imaging. Metabolic imaging aims to identify molecular biological factors that are potentially able to guide clinical management decisions; these techniques include MR spectroscopy and cerebral perfusion mapping. Functional MRI concerns the changes in cerebral hemodynamics that accompany brain function (activation); this latter topic lies outside the scope of this paper.


Rapid imaging and improved resolution are now possible using more powerful magnetic gradients. Techniques that utilize either strings of spin echoes (Turbo or Fast spin echo), or hybrid combinations of gradient and spin echoes allow faster image acquisition. The cranium can be imaged in to one or two minutes per imaging series, compared to 6 to 8 minutes for conventional spin echo acquisitions. Greater spatial resolution can be achieved without compromise in signal to noise. MR images can also be acquired individually in as little as half a second per image (Single Slice or HASTE techniques), very much like CT, at some loss of resolution and tissue characterization; less cooperative patients can thus be imaged with reduced sedation needs.


The latest innovations in hardware include magnetic field gradients that are strong enough and switch rapidly enough so that an entire imaging plane can be acquired in a single radiofrequency excitation. This technique, called echo planar imaging (EPI), allows image acquisition in a fraction of a second (as short as 50 milliseconds), or repeated multislice acquisitions on the order one or two seconds. The decreased imaging time allows for acquisition of images with higher temporal resolution. This improved temporal resolution permits the evaluation of cortical physiological events such as cerebral tissue perfusion and cortical activation.


Proton MR Spectroscopy provides information about the presence and amount of hydrogen protons attached to different cerebral molecular compounds. These protons possess intrinsic differences in resonant frequencies (or chemical shift) due to their differing molecular environment. A spectrum can be generated that corresponds to a scale of resonant frequencies vs amplitude (concentration). Molecular compounds identified within cerebral tissue include N-acetyl-aspartate (NAA, a neuronal marker), choline (a cell membrane marker), creatinine and phosphocreatinine (energy metabolites), and lactate (a by-product of cerebral metabolism).


MR spectroscopy is useful in characterization of brain tumors. Compared to more benign tumors, malignant tumors have an increased rate of membrane turnover (increased level of choline) and a decreased concentration of neurons (decreased NAA). Spectroscopy has had some success in the pre-operative differentiation of various tumor types. More importantly, spectroscopy allows for the non-invasive monitoring of the response of residual tumor to therapy. Finally, spectroscopy can be utilized to differentiate tumor recurrence from tissue necrosis.


Perfusion MR can demonstrate the microscopic vascular proliferation ("neovascularization") associated with tumor growth. Cerebral tissue perfusion can be assessed following a dynamic injection of Gadolinium. During the first pass transit through the cerebrovascular bed (which lasts only 5 to 15 seconds), gadolinium is restricted to the intravascular space. The restricted intravascular presence of highly paramagnetic contrast molecules (gadolinium) creates microscopic field gradients around the cerebral microvasculature, resulting in a change (shortening) of T2 relaxation and signal loss. From the amount of signal loss, the concentration of gadolinium in each pixel can be calculated, and a pixel by pixel relative estimate of blood volume can be inferred. Maps of cerebral blood volume (CBV) and cerebral blood flow (CBF) can be generated. As a rule, high-grade tumors have higher CBV values than low-grade tumors; and CBV values correlate with the grade of vascularity and mitotic activity.


Knowledge of tumoral vascularity is helpful to improve tumor grading, to identify optimal biopsy site in tumors with heterogeneous vascularity, to monitor for malignant degeneration and treatment efficacy, and to differentiate tumor recurrence from radiation necrosis.


In summary, recent advances in imaging techniques allow for improved tumor detection and biological characterization with MR imaging. This can be performed at the cost of added time (pulse sequences) to a MR exam. Alternatively, MR exam times can be greatly reduced (along with sedation needs), at the cost of poorer image quality and decreased tissue characterization.

Dr. Gilbert Vezina graduated from McGill Medical School, Montreal, Canada in 1983. He is presently the Director of Pediatric Neuroradiology at Children's National Medical Center (CNMC) in Washington, DC and his is an Associate Professor of Pediatric and Radiology at the George Washington University School of Medicine. Dr. Vezina is a member of the Tumor Imaging Committee of the Children's Oncology Group and from 1997-2002 was an Integration Panel Member of the Neurofibromatosis Research Program, United States Army Medical Research and Material Command. 2015



Thanks to Our Sponsors