Research news and notes

1. Carotid endarterectomy or stenting? 

Endarterectomy is commonly performed for severe carotid artery stenosis, whether symptomatic or not. As part of a general trend toward less invasive therapies and the development of endovascular technology, carotid angioplasty/stenting has also been tried. It has the obvious benefits of avoiding neck incision, and maybe it can shorten the hospital stay. It is now considered for patients who have extremely high surgical risk. The overall efficacy and risk of stenting is not well known, especially in direct comparison with endarterectomy. In the October 19 issue of the New England Journal of Medicine, Mas et al [1] present the results of a prospective randomized comparison between carotid endarterectomy and stenting in 527 patients conducted between 2000 and 2005. The study was stopped early because in an interim analysis, the risk of stroke or death after stenting was double that of endarterectomy at 30 days (3.9% after endarterectomy and 9.6% after stenting). At 6 months, the difference persisted, with 6.1% risk after endarterectomy and 11.7% after stenting.

Endovascular technology continues to evolve, but at this time, endarterectomy appears to be the treatment of choice for symptomatic severe carotid stenosis.

2. A mini preview of the future 

One day in the future, we will have endoscopes thinner that human hair but have images of higher quality than current endoscopes. It will show not only the surface it inspects but also its composition. One other thing—the endoscope of the future will show a 3-dimensional image. Only imagination is the limit to what can be done with such tools, if and when that day comes. How about an endoscope that is inserted through a thin-needle lumbar puncture and navigates through the entire cerebrospinal fluid space? Certainly sounds like science fiction.

The best part—a working prototype of such a fantastic device was reported in the October 19 issue of Nature. Yelin et al [3] presented images of metastatic ovarian tumor on a mouse peritoneum obtained with a novel miniature endoscope. They used spectrally encoded endoscopy, a method that does away with the classical bundles of optical fibers. Instead, a single optical fiber transmits light at a spectrum of wavelengths, configured so that each wavelength is projected to a different location on the surface. Light is reflected from the surface and decoded by a spectrometer. Rapid scanning movement of the tip by an external motor or galvanometer provides the second dimension, and optical interferometry performed on the spectral data generates information about depth. Although a single fiber is used, the number of pixels in the image can be very large—it depends on the bandwidth of the light source and the ability of the equipment to separate and analyze spectral data. The prototype used in the experiment can obtain volumetric images with 400,000 resolvable points (pixels) at 30 frames per second. The whole endoscope with the equipment at the tip of the fiber was introduced into the mouse abdominal cavity through a modified 23-gauge needle. It demonstrated raised tumor nodules on the peritoneum.

The authors suggest that color can be introduced by using 3 separate bands of wavelength, centered at each color. Spectrally encoded endoscopy can also be used to image labeled molecules, blurring the line between gross observation and pathological analysis.

3. Molecular imaging 

Magnetic resonance imaging (MRI) is great for anatomical images. However, although nuclear magnetic resonance has been used extensively in chemistry for identification of molecules, molecular imaging with MRI is not yet possible. For that, we still use modalities like proton emission tomography. Current MRI is limited by sensitivity. The radiofrequency signal polarizes a small proportion of the nuclei. To generate a high-resolution image, each pixel has to contain a very large number of nuclei of the same type. Therefore, abundant nuclei such as the protons of water or fat are detected in conventional MRI. In the October 20 issue of Science, Schroder et al [2] described a method that can allow true molecular imaging with MRI at high sensitivity. First, they use hyperpolarized xenon nuclei (129Xe) that are more magnetized by a factor of 10,000. Thus, even a small concentration of these nuclei produces a strong MRI signal. Then, they introduced a novel biosensor molecule for xenon that binds it and depolarizes it, making the xenon undetectable to MRI. Each biosensor can make a large number of xenon nuclei depolarized, amplifying the effect. Attaching the biosensor to a target protein, the authors were able to detect the protein with MRI at micromolar concentration. They claim a 3300-fold increase in sensitivity compared with direct detection of the same protein with existing methods.

Many technical issues remain, but it is not difficult to imagine xenon biosensors bound to specific antibodies used as a molecule-specific contrast material for MRI. True high-resolution molecular imaging using the regular MRI equipment is another way to get a specific rather than merely descriptive diagnosis from imaging.