In a marathon-like final session of the RSNA meeting, a five-physician panel offered insight into the PET boom of the past decade, discussed clinical applications for which metabolic imaging has become almost routine, and ventured predictions about future PET and hybrid imaging applications. The two-and-a-half hour symposium took place at McCormick's Crown Theater Friday afternoon.

According to Dr. Ronald L. Van Heertum, director of nuclear medicine at Columbia University College of Physicians and Surgeons, recent studies have validated PET's usefulness in the early diagnosis of Alzheimer's disease and in diagnosis and treatment planning for epilepsy. PET is also leading the way in new diagnostic approaches, such as plaque imaging, to assess the risk of stroke, he said.

Dr. Steven R. Bergman, director of nuclear cardiology at Columbia, spoke of the growing evidence of PET's advantages for detection and treatment of coronary artery disease. Although PET takes more time to perform and interpret than SPECT, it provides an edge in quantifying diffusion and predicting viability. PET surpasses SPECT with thallium 21 in the assessment of patients with infarction, he said.

"Patients with myocardial infarction have better survival rates if revascularization is done early," Bergman said. "Making the right imaging choice could also make the difference in correctly prescribing surgery or transplant."

Dr. R. Edward Coleman, director of nuclear medicine at Duke University Medical Center, described PET's latest achievements in oncology. Encouraging results have been recorded in preliminary studies of breast cancer diagnosis with PET, and clinical tests could be available to patients as soon as May, he said.

Coleman noted that applications have been filed with the Centers for Medicare and Medicaid Services for reimbursement of PET procedures for brain, cervical, small-cell lung, testicular, and pancreatic cancer.

"I'm optimistic that we'll see better reimbursement rates for oncology PET in 2003," he said.

Dr. Barry Siegel, director of nuclear medicine at the Mallinckrodt Institute of Radiology, discussed PET's ability to predict response to therapy. Until now, he said, chemotherapy has been used empirically and has not been individualized to each patient's case, which may explain why current response rates are low.

According to Siegel, PET can predict response to treatment in advanced breast carcinoma and lymphoma. In other cases, such as lung cancer, PET has found residual tumors and distinguished them from post-treatment fibrosis.

"Many molecular agents amenable to PET may make it more accurate to predict treatment response," he said.

Because FDG-PET provides a reliable, noninvasive way to assess cancer phenotypes, it could be valuable not only for evaluating treatment response, but for therapy planning as well, he said.

According to Dr. Richard Wall, director of nuclear medicine at the Johns Hopkins University, better spatial localization, faster scans, reduced patient motion, consolidation of visits to imaging specialists, and overall cost-effectiveness are some of the advantages of PET/CT.

Hybrid imaging also helps reduce the number of incorrectly identified lesions and increases the certainty level for less experienced readers. Localizing certain lesions, such as lung metastases from breast cancer, neck and head lesions, ovarian cancer, and bone lesions, would have been harder or nearly impossible without PET/CT, Wall said.

Dr. Harry Agress Jr., director of PET imaging at Hackensack Radiology in River Edge, NJ, closed the session by outlining the requirements for establishing PET in clinical practice.

After the first 3000 cases, he said, it is necessary to educate both partners and patients about PET, organize staff to mainstream the process from referral to billing, coordinate images digitally through PACS, correlate original imaging studies with PET, and consult with imaging specialists, technicians, and administration staff to provide PET as a part of integral services.