Creative strategies tackle technical aspects of PET/CT
Users offer tips for limiting common artifacts caused by contrast, timing, and motion
By: Deborah R. Dakins
Growing experience with PET/CT has given users a better handle on the idiosyncrasies of fusion imaging. With some sites reporting more than 8000 cases imaged with PET/CT to date, the finer points of patient preparation and technical specifications are emerging.
In just the past year, a plethora of published papers, along with hands-on familiarity at installed sites, has yielded strategies for limiting common PET/CT artifacts resulting from contrast, exam timing, respiratory motion, and muscle uptake.
Acquiring a PET/CT scan is similar in many ways to imaging with PET alone. A key difference is speed. PET/CT takes less time than a plain PET study, due to CT's ability to more rapidly acquire data for attenuation correction (AC) than the germanium-68 sources built into a PET device. CT attenuation maps can be created in about 38 seconds, compared with up to 30 minutes with Ge-68.1
Patient instructions to fast, to limit carbohydrate intake during their last meal, and to refrain from exercise for at least a day prior to the exam remain the same with PET/CT. And as with PET, patients with diabetes who undergo a fusion study must have good glucose control, verified by a finger stick just prior to exam prep.
One important difference between the two types of studies is the use of contrast in conjunction with the CT portion of the exam. Oral contrast is commonly used with PET/CT but is known to cause artifacts that can be limited or read through. Use of intravenous contrast is more controversial.
Several groups have examined artifacts caused by oral contrast on CT AC images. Researchers led by Dr. Richard Wahl, director of nuclear medicine at Johns Hopkins University, found that high-density oral and intravenous contrast can lead to quantitative alterations in the standardized uptake value (SUV) on CT AC.1 The group concluded that PET/CT is most safely performed without either type of contrast.
Another research group, based in Switzerland, also reported on the effect of oral contrast on PET/CT studies and found only minimally increased attenuation for 511-keV photons, even at high concentrations. The team concluded that oral contrast has a limited influence on SUV uptake.2
At Baptist Hospital in Miami, all patients undergoing PET/CT receive dilute oral contrast, typically Gastrografin or barium, said Dr. Jack Ziffer, medical director of nuclear medicine. The facility uses its PET/CT for oncologic applications in more than 90% of patients, primarily for lymphoma, pulmonary nodules, and breast and lung cancer staging.
Contrast is administered two hours prior to the study to allow time for transit through the small bowel and colon. On rare occasions, rectal contrast is used as well. FDG is subsequently administered an hour before the PET/CT exam.
The dilute nature of the contrast helps avoid potential artifacts, Ziffer said.
"We use dilute contrast because the algorithm for assigning attenuation coefficients can cause us to overcorrect and make scans look hot," he said.
PET/CT readers at Baptist Hospital, who are all board-certified in both nuclear medicine and radiology, can quickly review a study without the AC applied to determine what is artifact and what isn't.
"If there is no increased uptake on the non-AC image, then we have a high degree of confidence that it's not pathologic," Ziffer said. "It's pretty straightforward. After a while, you learn to read straight through these cases with the AC, non-AC, and CT on the same screen."
Ziffer applies the same AC versus non-AC approach to other iffy artifacts. Researchers have found, for example, that metallic structures such as dental fillings and hip implants, among others, can cause increased tracer uptake on CT AC images.3
Despite the need to be alert for artifacts, Ziffer believes that oral contrast is nearly indispensable for oncologic imaging with PET/CT.
"When we need to separate nodes from bowel, the oral contrast does that for us," he said. "It does affect attenuation correction and the intensity a little, and we get a bit more scatter, but what it brings to table far outweighs the mild diminution in quality."
At M.D. Anderson Cancer Center, users have also learned to read through artifacts posed by oral and rectal contrast. The team has concluded, based on its own early animal studies, that both can be used without problem, said Dr. Donald Podoloff, division head of diagnostic imaging.
TO IV, OR NOT TO IV
But use of intravenous contrast in conjunction with PET/CT poses more challenges than its oral counterpart because the influence of attenuation factors is not fully known.
At M.D. Anderson, where PET/CT is most frequently used to evaluate lung, esophageal, head and neck, and colorectal cancer, Podoloff's team is working to determine the best approach when it comes to IV contrast.
"We haven't completed our studies on that yet, but I think we will eventually be able to use IV contrast," he said. "Colleagues here have been conducting a variety of physics experiments to determine if the attenuation effects posed by IV contrast can be overcome by a mathematical algorithm."
Those attenuation effects have also been documented by Wahl's group, which found that high concentrations of IV contrast led to overestimation of tracer activity in phantoms. With the exception of high contrast concentrations, however, the emission bias was modest in in vivo canine studies.4
However slight, the focal artifacts on PET images that can result from incorrectly scaled contrast-enhanced pixels is an undesirable outcome, particularly for tumor imaging.
Options proposed by researchers at the University of Pittsburgh, one of the first sites to foster development of fusion scanning, include avoiding CT contrast administration entirely.5 But performing a CT scan without contrast may push some radiologists out of their comfort zone.
At Holy Name Hospital in Teaneck, NJ, IV contrast is used only if referring physicians specifically request it. The contrast-enhanced scan is performed after the PET/CT study is completed.
Many patients who undergo a PET/CT at Holy Name have already received a contrast CT scan, so performing another is not necessary, said Dr. Jacqueline Brunetti, assistant medical director of radiology.
"There are a lot of differing opinions as to whether you can use IV contrast, and there are data suggesting that you may be able to," she said. "But my feeling is that you don't really need it. The FDG acts as a contrast agent and really provides a lot of information."
The percentage of patients who receive IV contrast at Baptist is steadily decreasing, due to a similar realization by Ziffer and his colleagues.
Although about 60% of patients historically scanned by PET/CT at the site have been administered IV contrast, that percentage has dropped in recent months to between 30% and 40%, he said.
As is the case at Holy Name, contrast administration and the enhanced scan are performed after the PET scan and CT attenuation phase have been completed. The strategy helps avert focal artifacts on PET caused by the IV contrast, Ziffer said.
When preparing to conduct the second CT, it helps if patients remain on the table in the same position they were in for the other two studies, he said. The strategy abets fusion between the contrast CT and the PET/CT study.
"As PET becomes the accepted standard for staging patients, I think we will be doing even fewer contrast CTs," Ziffer said. "The oncologists don't need both sets of information. FDG is a superlative contrast agent, and the contrast in the CT isn't providing incremental diagnostic information in most patients."
TIMING IS EVERYTHING
Bringing out the best in PET/CT by maximizing the contribution of FDG is all about timing. Early users experimented with a wide range of wait times-anywhere from 30 minutes to three hours-between FDG administration and scanning.
Through practical experience, users have learned that scanning too soon (30 minutes or less) yields poor target-to-background ratios. The isotope has less chance of infiltrating the tumor, so signal to noise suffers. If they wait too long (four hours or more), readers must cope with the consequences of low count rates.
"Waiting time is very important," Podoloff said. "You should not do the scans any sooner than 60 to 90 minutes after injection."
Brunetti has also found value in extending wait times, particularly for breast cancer and pancreatic cancer patients. Typical uptake time in these cases is about 90 minutes. In cases of low-grade lymphoma, the wait might stretch a little longer, she said.
"You do better in identifying abnormalities if you have a little more washout from your backgrounds. And some of these tumors may take longer to actually accumulate. So you have better uptake later than in those tumors that are more active," Brunetti said.
BREATHTAKING PET/CT
Another trial-and-error exercise with PET/CT is establishing a breathing protocol that results in the best fusion image. Respiratory differences between the rapid CT AC acquisition and the longer PET scan can result in deterioration of the final image.5
The mismatch is most apparent when the CT is acquired with breath-holding at full inspiration and the PET is acquired while the patient breathes normally.
The artifacts, which include bright spots and inaccurate measurement of FDG concentrations, are typically seen in the diaphragm, base of the lung, and upper part of the liver. Alternative protocols range from partial breath-holding during the CT scan to shallow breathing throughout both the PET and the CT acquisition.
The most recent studies, conducted by researchers at the University of Zurich, suggest that a normal expiration protocol is superior to shallow breathing during CT scanning. Normal expiration is defined as the respiration level achieved when patients, after inspiration and expiration, hold their breath without forcefully exhaling.6
Although other options were tried, the normal expiration protocol is used at Baptist Hospital, Ziffer said.
"The problem is that the CT scan is so fast that when you do sample the lung at different phases of respiration, the attenuation correction map is less accurate," he said. "We have no doubt that acquiring the AC image by scanning during mid-to-end expiration in a fixed-state breath-hold has brought enhanced quality to our images."
BROWN FAT ARTIFACTS
Another learn-by-doing tip for PET/CT involves recognizing and learning to read through artifacts posed by uptake in supraclavicular area fat (USA-fat), known colloquially as brown fat.
PET users have previously noted the phenomenon of FDG uptake in the supraclavicular region, which is a common site for lymph node metastases. It was not until the advent of PET/CT that researchers were able to correlate the PET finding with CT anatomy.7
Users hypothesized that on PET the abnormal uptake in the neck was muscular, due to swallowing, neck tension, or patients not being in a resting state during FDG uptake. As researchers at Johns Hopkins discovered, the culprit instead turned out to be an unrelated process called the USA-fat finding-with FDG uptake in tissues of low-Hounsfield density. They hypothesize that the finding is caused by an underlying nonpathologic process that lights up brown fat.
Compared with typical fat, brown fat is metabolically active, residing in several areas of the body, Ziffer said.
"It's the same type of fat that bears have that is influenced by their state of hibernation," he said. "It can be found around the coronary arteries, and it may be there to keep those arteries warm. In addition to being active, brown fat uses glucose intensely."
The brown fat artifact is more commonly seen in women; it becomes more active when patients are in a cold or shivering state.
"It can interfere with reading the scan in the neck, so head and neck cancers can be exceedingly challenging to read," he said.
The artifact is especially troubling when evaluating young women for breast cancer. A key aspect of that assessment is detection of supraclavicular nodes, and brown fat can get in the way of a clear view.
Both Ziffer and Brunetti have found that brown fat artifact can be tempered by keeping patients warm.
"Anecdotally, we're potentially cutting down on nonspecific uptake and brown fat by keeping patients toasty during and after FDG administration," Ziffer said.
In addition, cold IVs are proscribed, Brunetti said, as they tend to induce metabolic activity in brown fat.
PET/CT SURPRISE
As the clinical prowess of PET/CT becomes more accepted, finding enough hours in the day to scan patients may become the biggest challenge users face.
At Baptist Hospital, where about 8000 PET/CT studies have been performed, the scanner runs from 6 a.m. to 11 p.m. six days a week. A seventh day is scheduled to be added; the site will also install two additional PET/CT devices by year's end.
"The biggest complaint that the clinicians have is about the backlog in trying to get their patients evaluated," Ziffer said.
In the minds of clinicians, he said, PET/CT is emerging as the sole test for oncologic evaluation. In the past, at a typical tumor board gathering, Ziffer would be peppered with queries about what tests should be performed in addition to PET/CT. Now clinicians want to know whether other tests such as bone and gallium scans can be safely dropped in favor of simply performing a PET/CT study.
Even early adopters of PET/CT are surprised at just how much they've come to value fusion imaging. They have found that the CT part of the PET/CT equation adds far more than just attenuation correction. The Johns Hopkins group has reported that CT provides unique and useful diagnostic information in more than 10% of patients.
Podoloff initially expected PET/CT to be most valuable in increasing patient throughput because of the rapidity with which attenuation scans could be acquired with CT. But he did not expect CT to add clinical information that he now believes is important to patient assessment.
"My philosophy about how this machine should be used has changed. We now feel strongly that there is diagnostic information in these noncontrast CTs," he said.
In fact, radiologists at M.D. Anderson have changed their report format to include a line explaining that prior to the PET study, a noncontrast-enhanced CT was acquired "to obtain data for attenuation correction and diagnostic purposes."
Some say that the speed with which users master PET/CT's technical issues may depend on whether they hail from nuclear medicine or radiology. But current users caution neophytes that regardless of where they hang their hats, PET/CT needs to be approached as an entirely new modality, not as a combination of two known quantities.
"I'm a proponent of true fusion imaging," Brunetti said. "You really need to look at a PET/CT as a whole instead of trying to separate it into different exams. That's not easy for everyone to do."
REFERENCES
1. Wahl R. PET-CT in clinical practice: experience in 1500 patients. Proceedings of the Japanese Society of Nuclear Medicine, 2003.
2. Dizendorf E, Hany T, Buck A, et al. Cause and magnitude of error induced by oral CT contrast agent in CT-based attenuation correction of PET emission studies. J Nucl Med 2003;44(5):732-738.
3. Goerres G, Ziegler S, Burger C, et al. Artifacts at PET and PET/CT caused by metallic hip prosthetic material. Radiology 2003;226:577-584.
4. Nakamoto Y, Chin B, Kraitchman D, et al. Effects of nonionic intravenous contrast agents at PET/CT imaging: phantom and canine studies. Radiology 2003;227:817-824.
5. Townsend D, Beyer T, Blodgett T. PET/CT scanners: a hardware approach to image fusion. Semin Nucl Med 2003;3:193-204.
6. Goerres G, Kamel E, Seifert B, et al. Accuracy of image coregistration of pulmonary lesions in patients with non-small cell lung carcinoma using an integrated PET/CT system. Radiology 2003;226:906-910.
7. Cohade C, Osman M, Pannu H, et al. Uptake in supraclavicular area fat ("USA-fat"): description on 18FDG PET/CT. J Nucl Med 2003;44:170-176.
---
TECHNICAL TIPS FOR PET/CT
Use dilute oral contrast to avoid attenuation artifacts.
Check attenuation-corrected (AC) images against non-AC images to verify artifacts versus pathology.
Consider reducing or eliminating use of IV contrast as experience grows.
PET/CT scans are best performed between 60 and 90 minutes after FDG administration.
Normal expiration protocol is superior to shallow breathing during CT AC scan to avoid artifacts.
Keep patients warm and comfortable to avoid abnormal uptake in brown fat.
|
