As promising as the current crop of PET/CT scanners is, researchers have already begun work on the next generation. Dr. David Townsend, the senior physicist at the University of Pittsburgh in the U.S., who developed the prototype PET/CT, has received a $3 million grant from the National Cancer Institute to create even faster hybrids. He will apply new panel detector technology to improve the sensitivity of future scanners by increasing their axial length as much as 50 cm, covering a larger part of the body in a single shot.
Townsend estimates that the new panel detectors will cut PET imaging time of the thorax and upper abdomen to two to three minutes. His concept is to merge the panel detectors with multislice CT to produce up to 16 slices and perform whole-body imaging in five minutes.
The advantages will be quite evident, he said. Faster imaging translates into higher patient throughput because it solves the logistical problem of injecting four patients an hour with FDG. It also greatly enhances image quality by reducing motion artifact. In the research domain, next-generation PET/CTs will allow investigators to explore faster processes with other tracers, such as carbon-11, which can provide a whole-body snapshot in three minutes rather than 30 to 40 minutes.
The anticipated advancements will help PET/CT solidify its position as the premier modality for oncology imaging. They also will open the door to new clinical applications in imaging inflammation, atherosclerosis, myocardial function, and stenosis. In the process, PET/CT may challenge MRI of malignancies, except in the brain or bones, and replace nuclear medicine techniques for imaging infection throughout the body.
MAYBE EVEN CHEAPER
Reduced scan time with new PET/CT technology will greatly improve imaging in the chest by reducing the artifact caused by respiration and movement of the diaphragm, which compromises scans of the upper dome of the liver and the base of the lung. Faster scanners therefore will eliminate the need for gating or breathing protocols such as the one being used at the University of Essen in Germany.
Townsend explained that if patients hold their breath during scans of the lungs and upper abdomen, CT will match the average position of PET imaging. For clear imaging while the CT passes over the diaphragm, however, patients need to hold their breath on partial expiration, which is difficult to do for prolonged periods. But patients could hold their breath without taking a deep breath if they were scanned by a CT faster than today's dual-slice PET/CT instruments that could cover the whole region in a few seconds.
Gating subdivides image acquisition to different parts of the breathing cycle so each part corresponds to a frame of a few seconds of breathing to match PET and CT frames during partial breathing. Other PET and CT frames do not match, however, and must be discarded.
"This is not acceptable for PET because you are count-limited," Townsend said. "Maybe with the new high-speed systems it will be possible to throw away 80% of the data. But at the moment, you don't have enough counts on the PET side."
Prof. Gustav von Schulthess, director of nuclear medicine at the University of Zurich in Switzerland, is sure that manufacturers of future editions of PET/CT will produce less expensive units because they can be designed without duplicating PET's transmission scanning system with CT.
"Manufacturers will build the PET scanning component by itself because the addition of CT makes PET's transmission correction system obsolete," he said. "The PET portion of the scanner will become cheaper because you are essentially getting transmission scanning free with CT."
PET/CT scanning may become cost-effective because of patient throughput as well. It can increase throughput by a factor of two, which in itself makes the technology economically sensible, according to Dr. Jorg Debatin, director of radiology at the University of Essen.
PET/CT IN ONCOLOGY
PET/CT may end up rivaling MRI in oncology, at least in staging cancers in the body, von Schulthess said. An effective clinical imaging modality must be robust enough to produce high-quality images of virtually every patient, and MRI is less robust than PET/CT, particularly in tumor staging, because it provides much less information.
MRI is also less effective than PET/CT in differentiating between increased contrast associated with a residual tumor and a thermally ablated lesion in the liver, Debatin said. He and his associate, Dr. Andreas Bockisch, director of nuclear medicine, have performed a few PET/CT studies in patients who have undergone laser or radio-frequency ablation of small tumors in the liver and found that PET identifies active sites and CT gives their precise location.
MRI gathers superior information in the brain, however, according to von Schulthess. Since PET/MR software fusion has none of the drawbacks in the brain that it has in the body, he predicts that PET alone or PET/MR software fusion would capture brain tumor applications but that MR might be preferable for identifying bone metastases. In imaging a few patients with cancer spread to the bone, von Schulthess did not see metastatic lesions in the CT part of PET/CT, but he did find them with MR.
Dr. Steven Larson, chief of nuclear medicine at Memorial Sloan-Kettering Cancer Center in New York City, believes PET/CT will have a major impact on imaging cancers that are not yet reimbursed by Medicare, such as tumors in the thyroid and prostate glands and in children. Physicians will likely refine the use of PET/CT, however, to focus on biologically aggressive malignancies that are highly metabolically active, endocrine responsive, or histologically advanced, including prostate cancers with high Gleason scores or tall cell thyroid cancer.
The properties of FDG make PET/CT a strong modality for imaging inflammatory processes. FDG collects not only in tumors but in activated white blood cells such as granulocytes and macrophages that consume large amounts of sugar when fighting infection. Unlike standard nuclear medicine methods of imaging infection, however, FDG-PET does not accumulate in normal bone marrow, von Schulthess said.
PET/CT has the potential to replace activated indium-labeled white cells and gallium for detecting infectious processes in the body, said Dr. Richard Wahl, director of nuclear medicine at Johns Hopkins Medical Center in Baltimore, Maryland. He envisions a time when patients will be injected with FDG and scanned an hour later with PET/CT to find infection anywhere in the body.
Advanced Radiology Services, a private radiology practice in Grand Rapids, Michigan, is beginning to use PET/CT to check for inflammation and infection in small numbers of patients because it is much quicker and easier to do than imaging with labeled leukocytes, said Dr. Paul Shreve, a radiologist with the practice.
Wahl also believes PET/CT may identify small, focal areas of high glucose uptake representing inflamed areas in major arteries or vulnerable plaque. His associate, Dr. Mitsuaki Tatsumi, a postdoctoral nuclear medicine fellow, reported at the 2002 meeting of the Society of Nuclear Medicine that 50 of 85 patients who had been scanned with PET/CT exhibited at least one positive FDG focus in the thoracic aorta and 45 had signs of calcification at other sites.
It's too early to know for certain, but PET/CT may detect atherosclerosis at an earlier stage than standard methods, Wahl said. Standard CT spots areas of calcification, which are manifestations of hard or mature plaque. The soft and inflamed plaque identified by PET/CT may be vulnerable and susceptible to rupture.
"We are still fairly early in our cardiac evaluative phase, but we think PET/CT may be useful for providing a combined assessment of glycolysis and calcification of blood vessels," Wahl said.
Further delineation of PET/CT's role in imaging atherosclerosis most likely will have to wait for faster machines. At the present time, PET/CT scanners can achieve a spatial resolution that is questionable for imaging plaque, von Schulthess said, but plaque may become visible as new technology improves resolution.
MYOCARDIAL FUNCTION AND ARTERIAL STENOSIS
Dr. Bennett Chin, clinical director of nuclear medicine at Johns Hopkins, has been conducting animal experiments to evaluate PET/CT's potential for quantifying ventricular function. The idea is to obtain a series of CT scans in an extremely short period of time (e.g., 100 msec), look at different parts of the cardiac cycle, analyze heart contraction, and calculate functional indices to answer such questions as, How much blood is the heart pumping on each beat? How effective is heart function when there are regional or local wall motion abnormalities?
If that kind of information can be coupled with metabolic indications of myocardial tissue viability, PET/CT imaging may be able to help clinicians decide whether to revascularize a patient who has poor wall motion, Chin said.
Fast PET/CT scanning may influence clinical decision-making by combining imaging data about coronary artery perfusion and myocardial tissue health. Fast CT scanning of coronary vessels and metabolic data about ischemia might not only identify which patients should have bypass surgery but which areas of heart muscle should be bypassed.
"If a patient has myocardial ischemia that is so severe it has caused myocyte death in one region of the heart, then obviously surgeons wouldn't want to revascularize that area," he said.
Chin is exploring PET/CT for coronary stenosis by injecting the perfusion tracer rubidium-82 at high and resting blood flow rates and looking at the difference. At least theoretically, a very tight stenosis may still allow enough blood flow to distribute the tracer evenly throughout the heart at rest. But in response to increased demand from a pharmacologic or exercise stress, a stenosed blood vessel may not dilate to the same extent as other coronary arteries. If PET/CT angiography proves effective, tissue distal to the stenosis will have less tracer uptake over areas of ischemia, and the PET/CT image of the heart will not be uniform.
"There are many aspects along the spectrum of cardiovascular disease that you may be able to study with PET/CT, such as large-vessel disease," Chin said. "But there also are competing modalities, such as MRI, which has high spatial and temporal resolution but lacks metabolic information. It's yet to be proven which is going to be the most accurate."