ANN ARBOR—Two new studies at the University of Michigan Medical Center demonstrate that a technology, previously considered experimental in the United States, is significantly more accurate than other imaging methods used to detect cancer and can also find areas of infection in the body far quicker than conventional means.

PET, or positron emission tomography, is widely used clinically in Europe, but has seen relatively limited clinical use in the United States. To produce a PET image, patients are injected with a glucose that contains a radioactive tracer. Cells that are the site of higher metabolic activity, which is usually the case with cancer and infection, take up more glucose. Radioactivity emitted by the glucose is recorded on the PET camera and reconstructed by a computer to form an image similar to a CT (computed tomography) scan. Unlike CT, which shows anatomy, a PET scan highlights tissue metabolism. These areas show up as bright color on the computer-generated image.

The U-M studies were presented Dec. 3 at the 83rd Scientific Assembly and Annual Meeting of the Radiological Society of North America. They come in the wake of an announcement by the U.S. Health and Human Services Department that it will accelerate its review of PET technology.

In the two U-M studies, researchers found: Dedicated PET scanners outperformed newer imaging devices that were created as a lower-cost alternative to PET.

Researchers found that the newer devices, called dual-head coincidence cameras, missed from 10 percent to 77 percent of the cancers successfully detected by standard PET cameras.

Researchers at U-M performed both PET and dual-head camera imaging on 35 patients with known or suspected cancers. PET detected 121 cancer sites, compared with 70 sites found by dual-head imaging.

Compared with cancer types detected by PET imaging, dual-head cameras only found the following percentage of sites: neck—90 percent; lung—86 percent; metastatic cancers in the rib or spine—77 percent; lymph nodes in the chest—58 percent; axillary lymph nodes—50 percent; abdomen—23 percent.

“This study demonstrates that current dual-head coincidence camera technology is considerably less sensitive than PET in detection of small cancers,” says lead author, Paul Shreve, M.D., assistant professor of radiology and internal medicine. “The newer technique may compare to PET in detecting some lung cancers.” Shreve adds there are significant improvements to the dual-head imaging devices now under development.

The dual-head cameras were created as a low-cost alternative to PET, but according to Shreve, the price of PET scanners has fallen dramatically. They now compare favorably to the cost of some CT scanners. A new study shows PET can quickly diagnose infections.

U-M researchers studied 11 patients who were believed to be harboring various bacterial infections. PET successfully identified infection in eight of nine patients and correctly ruled out infection in the other two—and did so much faster than other conventional diagnostic methods. Current tests can take anywhere from 24 hours to a few days to diagnose infections, while PET can do so in an hour or less.

“The speed with which the site of an infection can be pinpointed and appropriate treatment started can have a significant effect on outcomes and cost of care,” says Richard Wahl, M.D., professor of radiology and internal medicine, and director of general nuclear imaging at U-M. Wahl cautions that the study results are preliminary, but says they indicate PET shows great promise in rapidly identifying sites of infection.

“More extensive clinical evaluation is warranted to determine the accuracy of this method,” Wahl says, “including its application for imaging viral and fungal infections.”

Previous studies have shown that in up to 30 percent of patients with cancers, PET can detect cancer spread not found by other imaging methods. PET has been shown in other studies to change treatment decisions in 25 to 40 percent of cases.

Wahl’s co-authors are Y. Sugawara, M.D., who presented the study at RSNA; S. Fisher, B.S.; P. Kison, B.S.; and K. Zasadny, Ph.D. Co-authors on the paper presented by Shreve at RSNA are R. Steventon, C.N.M.T; E. Deters, C.N.M.T.; M. Gross, M.D.; and R. Wahl, M.D.