· ARIA BIS System Adds Biopsy Capability
· Sonography Effective in Diagnosis of Carpal Tunnel Syndrome
· Ultrasound System Takes Trip Around the World

Reflections on Radioisotope Supply

Alexander McEwan, MD
Alexander McEwan, MD

Alexander McEwan, MD, president of the Society of Nuclear Medicine and director of oncologic imaging at the Cross Cancer Institute in Edmonton, Alberta, recently spoke to Medical Imaging about last November’s shutdown of a nuclear reactor in Chalk River, Ontario, and its subsequent impact on US hospitals and patients. Here, McEwan offers his perspective of the crisis and discusses how to prevent a future shortage.

MI: Why did the closing of one plant in Canada impact us here in the United States?

McEwan: Generators produce the radioisotope called molybdenum-99, which decays to Technetium 99, and that is used in molecular imaging. The United States gets its supply from two main manufacturers, Covidien and BMS (Bristol-Myers Squibb Medical Imaging). They receive it raw from the reactor from Chalk River, process it, and then it is sent to hospitals and radiopharmacies. It’s a fairly complex supply chain. So if the reactor at the front of this chain breaks down, then you lose your supply of molybdenum and Technetium.

MI: What solutions were hospitals forced to undertake?

McEwan: Some of them just had to shut down. Studies and data suggest that 20% of hospitals were running close to the limit, and many others would have shut down within a week or two if the supply wasn’t restored.

About 40% of hospitals were imaging at 50%. Because Covidien gets some of its supply from Europe, it was able to supply European molybdenum. Pharmacies that BMS supplied were the ones that were hit the hardest.

MI: What was the impact on patients?

McEwan: We don’t have an exact number of patients who were affected, but you can come up with an estimate through extrapolation. For example, we know that in the United States, there are about 19 million imaging facilities a year using molybdenum, so about a million and a half facilities use it per month.

MI: Is it common for one plant to churn out most of the continent’s supply of tracers?

McEwan: It’s not that it is common or uncommon. It’s just a fact. There are three reactors in the world that make molybdenum; the one in Chalk River has typically made more than three quarters of the world’s supply. The others are in Belgium (The Belgian Institut National des Radioelements) and South Africa (the Nuclear Energy Corporation of South Africa).

MI: Although the facility has since reopened, are health care analysts concerned about relying on limited sources?

McEwan: The Society of Nuclear Medicine has identified for 10 years that this is a problem. There is no US supplier of Technetium, and we as a society are working very hard to get a US reactor into supply. But obviously, that is expensive.

One of the reasons I think there has been reluctance is because the AECL (Atomic Energy of Canada Ltd, which owns the Chalk River reactor) has been making two other reactors, Maple 1 and Maple 2. Both of these, 7 to 9 nine years later, are still being commissioned, and there is no clear identification of when these two reactors are coming upstream. Because of this most recent problem, there is much more interest from the States in developing an internal supply. However, it would probably have to take a brave investor who would put a lot of money into building a reactor.

MI: How should hospitals and centers prepare for future emergency shortages? What is the society doing to avoid future shortages?

McEwan: The society is urging the United States to build its own medical reactor. That is pretty much the ultimate solution. I am certain that those people who are getting their molybdenum from a sole supplier will now be looking at getting it from alternative suppliers. Manufacturers will be looking at Europe and South Africa for partial supplies.

—Elaine Sanchez

FDA Clears New Varian Algorithm for Eclipse Software

On December 14, 2007, Varian Medical Systems Inc, Palo Alto, Calif, announced it had received FDA 510(k) clearance for a new proton scanning dose algorithm available now on the company’s Eclipse treatment-planning software. The algorithm makes it possible to plan for intensity modulated proton therapy by optimizing dose distribution in three dimensions, protecting critical structures near the treatment target.

“No one else has a pencil beam scanning algorithm that’s been 510(k) cleared,” explained Armin Langenegger, a researcher with Varian’s treatment planning products team. “It’s a different way of doing proton therapy. This clearance covers three new techniques: double-single scattering, spot scanning, and uniform scanning.”

The new techniques make intensity modulated proton therapy possible by optimizing dose distribution in three dimensions. “The algorithm optimizes dose in the x and y directions as well as the depth direction,” said Langenegger. “Protons don’t exit the body; they stop after a certain distance. So when you deliver spots, you do so in an x-y matrix and also to a specified depth. And you can vary that intentionally to get different dose distribution.”

The new algorithm uses both an aperture and a compensator to shape dose distribution, meaning clinicians no longer need to enter and exit the treatment room repeatedly during therapy to adjust physical compensators for processing proton beams. “With proton scanning and intensity-modulation, this cumbersome process can be completely eliminated,” commented Jeff Amacker, business manager for treatment planning products at Varian. And because the algorithm offers increased precision in treatment planning, dosing becomes less hazardous to the patient.

“Clinical applications for pencil beam scanning are in the head and neck region, specifically in the brain and brain stem, where structures are close to other critical structures,” said Langenegger. “You can place these spots so closely that you can avoid nearby organs entirely. And you don’t get the whole integrated dose scenario where other structures get extra dose that shouldn’t. It doesn’t make treatment planning faster, but it makes the dose cleaner to the patient, which means less side effects. It’s particularly useful for treatment in pediatrics or in difficult cases with very close critical structures.”

Even if planning isn’t quicker, says Langenegger, treatment certainly is. “You have increased biological effectiveness because of the biological dose deposition of protons compared to photons,” he said.

The new clearance adds two additional modules—one for uniform scanning and one for pencil beam scanning—to the Eclipse software, bringing the total number of available modules to five. “We’re currently the most advanced system in the world for doing proton distribution,” Langenegger said. “We’re the only system in the world that offers five modules.”

Additionally, Langenegger notes that the Eclipse system currently supports all major proton vendors, including Ion Beam Applications Solutions (IBA), Louvain-la-Neuve, Belgium, and Hitachi Medical Systems, Twinsburg, Ohio. “Our models are flexible enough to incorporate any other new vendors coming up in the market,” he said.

The new algorithm is available now on Varian’s Eclipse treatment-planning software.

—Cat Vasko

FDG-PET with SSP Is Superior for Alzheimer?s Diagnosis

A recent article in the journal Brain confirms that FDG-PET can significantly improve the accuracy of diagnoses of frontotemporal dementia (FTD), which is commonly mistaken for Alzheimer’s. In the study, six neurologists correctly distinguished FTD from Alzheimer’s 90% of the time using FDG-PET imaging, an improvement of 14% from traditional clinical diagnostic methods.

“The gold standard was pathologic diagnosis,” explained Norman Foster, MD, professor of neurology and director of the Center for Alzheimer’s Care, Imaging and Research at the University of Utah School of Medicine, Salt Lake City. “But we found that PET imaging had greater diagnostic accuracy than clinical information alone. The bottom line is that PET is better than clinical for both accuracy and reliability. We went on to see whether adding PET to clinical information changed physicians’ minds about the diagnosis. And the answer is, it makes physicians more accurate and confident in their diagnoses. When it changed diagnosis, it changed it in the correct direction.”

In their early stages, FTD and Alzheimer’s resemble each other closely in terms of their impact on patient’s memory and behavior. But FDG-PET reveals a key difference between the two: low activity is centered in completely different parts of the brain depending on the type of disease.

“A lot of literature from the past 20 years shows that there are distinctive abnormalities in glucose absorption by patients with Alzheimer’s, but it had never been tested in a systematic way,” said Foster. “Often these reports were based on a small series of cases with no pathological information. We looked at several methods physicians can use to distinguish the two. One is just by doing an exam, one is to use a diagnostic checklist, and the third is with PET. We compared the standard transaxial image display using a color scale with a statistical method that summarized the images.”

Foster and his team looked at the medical records and PET scans of 45 patients who later had autopsies and determined that 31 had Alzheimer’s, while 14 had FTD. Then, National Institutes of Health neurologists were asked to decide which form of dementia the patients had suffered from using clinical information alone, FDG-PET alone, and the two combined.

“The reason these studies weren’t read by radiologists is that it’s important to interpret scans in the context of the clinical patient,” noted Foster. “It’s important that these individuals had different levels of experience. Training was part of this study, a brief but important part. We know that many radiologists and even nuclear medicine physicians have very little experience interpreting PET images for dementia. It’s very different from other studies, where you look for a hot spot or a cold spot. We look for a pattern.”

In the study, the neurologists correctly distinguished FTD and Alzheimer’s using clinical information alone in 76% to 79% of the cases; using FDG-PET alone, they correctly diagnosed the two dementias 85% to 89% of the time. Using both together, they distinguished the two correctly 90% of the time. And the highest accuracy in diagnosis was achieved using stereotactic surface projections (SSP) displays, which use a statistical test to show significant areas of damage.

“One of the issues this study brings up is that physicians’ interpretation of PET scans was aided by a statistical assessment,” said Foster. “These are things that weren’t previously done. My contention is that we ought to be using advanced image analysis methods to enhance the accuracy and reliability of clinical interpretations of images. This was a great example of taking scientific advances that are more or less everyday in research settings and getting them to make a difference in patient care.”

—C. Vasko