CT, on the other hand, offers excellent spatial resolution and rich anatomical detail. Integrating the two on a common gantry and patient table, in a concept that has only recently been brought to the commercial marketplace, holds the promise of simplifying patient handling, data acquisition, and coregistration of the CT and radionuclide image data.1 Particularly in the field of cardiac imaging, this fused technology also has the potential to improve quantitative functional assessments and correct the problem of attenuation.
For those now using SPECT/CT or PET/CT, the benefits are already making themselves clear. As with any new technology, however, there are hurdles to overcome before such a promising start is extended throughout the radiology community.
SPECT/CT and PET/CT operate on the same basic design principle: the dual modality acquires CT and radionuclide scans by translating the patient from one detector to the other while the patient remains on the table. This allows both images to be taken with a consistent scanner geometry and with minimal delay between the two acquisitions. After both sets of images have been acquired and reconstructed, image-registration software fuses the images while accounting for differences in scanner geometry and image format between the two data sets.1
While the principle may be the same, each modality has its specific benefits. One of the major anticipated uses of SPECT/CT is the production of better attenuation correction. Apparent perfusion defects occur most often in the anterior wall in women and in the inferior wall in men, and soft-tissue attenuation can also shift between resting and stress images. Interpreting these examinations requires clinicians to recognize any attenuation artifacts and allow for them in evaluating the underlying perfusion pattern.2
Randy Hawkins, MD, PhD, chief of nuclear medicine, University of California San Francisco (UCSF) School of Medicine, says, “We have long known that breast artifacts are a major problem in cardiac imaging, and only the most experienced readers feel comfortable reading around those artifacts.” According to Hawkins, the prototype of the SPECT/CT unit has been in the laboratory of UCSF radiology department faculty member Bruce Hasagawa, PhD, for several years. That prototype served as the model for commercial SPECT/CT units, one of which was installed at UCSF several years ago.1
“It is still too early to say what the impact of the combined technology will be in cardiac imaging, but we have already noted improvement in our attenuation correction,” Hawkins says. “Of course, we could do the same thing with separate scans, but there are advantages to having them performed contemporaneously. The results of both studies are available on the same day, so the pathology and the patient’s medical condition obviously have not changed. Combined technology helps because it produces the result right away, with the appropriate technical subtleties. Even though computer methods can be used to superimpose images with fusion techniques based on software, unless that has been planned in advance, it can be awkward in a clinical setting.”
Miami Cardiac and Vascular Institute has had a SPECT/CT unit in place for 3 years, and the modality is used for general nuclear medicine and cardiac imaging. Jack Ziffer, MD, PhD, director of cardiac imaging at the institute, says that the unit has dramatically improved the quality of the facility’s nuclear cardiology studies by solving some of the traditional problems of SPECT. “With a routine nuclear SPECT study, we inject an agent that is taken up by the heart according to blood flow,” Ziffer says. “To assess flow, we then need to image the photon, regardless of where it comes from; absorption of that photon can cause problems in the inferior wall in men and in the anterior wall in women. What frequently happens then is that scans may be appropriately read as abnormal, but falsely interpreted as positive. Patients may be taken to the catheterization laboratory because of that, and we can also start to dismiss defects when they may be real, and thus underdiagnose disease and have infarcts that could have been avoidable.”
An additional benefit of SPECT/CT is its ability to quantitate blood flow in an absolute sense, which is important for the better detection of global balanced ischemia. “If a patient has triple vessel disease, there will be reduced flow in all those vessels,” Ziffer says. “That can look normal because the information is relative, not absolute. Attenuation correction, coupled with scatter correction, can reveal absolute coronary blood flow to diagnose patients better.”
While UCSF has not yet added PET/CT to its facility, Miami Cardiac and Vascular Institute does have the modality, and Ziffer says that it has revolutionized the general nuclear medicine department, in particular. Within the cardiac arena, Ziffer notes, the modality combines the ability of PET to provide nearly perfect perfusion images with the ability of CT to provide a noninvasive coronary angiogram. “PET provides superb images using rubidium 82, but when patients are done with that test, we may not know which vessels are abnormal,” Ziffer says. “That usually requires a diagnostic angiogram. A normal perfusion scan can also potentially underestimate the presence of significant atherosclerosis that is not yet hemodynamically significant. The limitation of cardiac CT, on the other hand, is that the resolution is not sufficient to assess hemodynamic significance, though it can tell you how much hard plaque there is and, potentially, how much soft plaque there is,” he continues. “Now, with one test, we can determine myocardial perfusion using the best test there is to do that, and see the coronary anatomy noninvasively using the best noninvasive test there is for that. Potentially, that will answer all the important clinical questions and allow us to identify and treat patients with non-hemodynamically significant disease.”
PET/CT also has the advantage of shortening overall examination time, thereby increasing throughput in the imaging center; a PET/CT fusion scanner can often image up to 16 cases per day.3
As these advanced modalities come into more frequent use, the question of who will operate the technology is under discussion. Rules vary by state, leaving hospitals to prepare a strategy, on their own, that works within the constraints of both local regulations and staff availability.
Robert E. Henkin, MD, is professor of radiology, vice-chair of the department of radiology, and director of nuclear medicine at Loyola University Stritch, Maywood, Ill. “States individually decide who can do what right now, and there is no national decision,” he says. “A joint meeting of the American Society of Radiologic Technologists and the Society of Nuclear MedicineTechnologist Section, the purpose of which is to come up with training that technologists should pursue to make sure they are competent in both areas, will soon be held. We will recommend a curriculum ensuring that CT and nuclear medicine technologists can be trained to use either modality.”
Hawkins says, “Individual hospitals must have a strategy in place to have personnel to perform and interpret CT with PET. When doing diagnostic-quality scans with PET/CT, an appropriate personnel strategy might be a shared effort between radiology and nuclear medicine technologists; however, that may require some cross training.”
Even with a curriculum in place, it will take several years to implement technologist training in both modalities. “A lot of training is currently redundant, though, so this represents an opportunity for economies in training,” Ziffer says. “The combined technology could also be staffed using nuclear and radiographic technologists in tandem, and not all states require certification by a technologist to use this instrumentation.”
Some outcomes data exist for the improved treatment planning offered by PET/CT, including the results of a 300-patient study conducted by Townsend and Meltzer between 1998 and 2001. The study found that the accurate spatial localization offered by PET/CT fusion studies improved the assessment of response to treatment and changed clinical management for 20% to 30% of patients.3 For the most part, however, the technology in practice is too new to have much documented outcomes data.
Ziffer says that his facility is in the process of collecting data as part of a multicenter study of the dual technology, and that effort should be finished sometime in 2003. Hawkins notes that Hasagawi and his group have done a number of research studies on the prototype unit, looking at cardiac and other applications of the technology, but current clinical applications have not been fully defined. “One obvious potential use is better attenuation for SPECT scans done with technetium and other types of radiopharmaceuticals,” Hawkins says. “It may turn out that there is particular advantage to adding CT because of specific attenuation correction with certain pharmaceuticals, but we do not know that yet.”
While it is clear that the improved images and faster throughput offered by fusing imaging technology can change how facilities operate, Henkin adds that other factors should be considered before SPECT/CT and PET/CT technologies are fully disseminated. “Until there is further exploration of software fusion techniques that do not require additional hardware, there is some trepidation about how this will change how we do things,” Henkin says. “Where should this new technology find a homein nuclear medicine or in CT? Should body imagers run the machines, or should the nuclear medicine staff? What studies need to be done on these units and which can be done on single devices? Once that technology is used in the broader community, with less sophisticated users and patients who are less complex, then people start to raise questions. In terms of public policy, we have to be sure that we have answered the right questions.”
The Cost of Technology
Cost represents a significant hurdle for any new technology, and single-photonemission computed tomography (SPECT)/CT and positron-emission tomography (PET)/CT are no exceptions. The combination modalities are more expensive than single-modality imaging units, with some of the latest models costing more than $2 million.
“All machines are significant investments and there has to be a cost-justifiable reason for getting any of them,” Randy Hawkins, MD, PhD, chief of nuclear medicine, University of California San Francisco (UCSF) School of Medicine, says. “Before even dealing with cost, however, we need to decide if this is a better test. If we think we are getting more information with these tests, then there is a reason to develop a strategy to test cost effectiveness.”
Sometimes the initial cost can be justified by down-the-line savings related to improved patient results and averted unnecessary procedures. “The initial experience with the new technology has been that it does not save money, unfortunately,” Henkin says. “It does result in better therapy planning, particularly in radiation oncology; the quality of a treatment plan is better if we use PET/CT.”
Further economic implications fall under the realm of reimbursement, and the combined technology currently has no Current Procedural Terminology (CPT) codes for billing. This places the burden of payment on either the facility or the patient, neither of whom may yet be convinced of the worth of fusion imaging. Hawkins says that most facilities using SPECT/CT still bill for it as a standard nuclear medicine procedure without the addition of CT, but even a SPECT/CT bone scan will still be a bone scan, and will have to be billed as such.
At Miami Cardiac and Vascular Institute, the facility has chosen to take up the cost of the surcharge for attenuation correction using SPECT/CT. “We have been more than willing to do that for a better-quality study, but it is not easy to justify from a purely economic standpoint,” Jack Ziffer, MD, PhD, director of cardiac imaging, says. “I hope that we will have a separate code for SPECT for attenuation correction, but that is under review right now and we are not sure what the recommendations will be.”
Robert E. Henkin, MD, professor of radiology, vice-chair of the department of radiology, and director of nuclear medicine at Loyola University Stritch, Maywood, Ill, says, “Having no CPT codes presents some significant issues from the point of view of reimbursement. If a patient had a diagnostic CT a week ago, and there is a reason we are now doing a PET or SPECT study, then will the third party pay for the CT scan? The PET or SPECT study is really being done for our interest, and not to render a diagnosis.” He adds, “The question goes beyond technology to public policy. It is easy to ask for additional reimbursement for something that costs more to do, but if there is no money to pay for it, the implications are significant.” Hawkins says, “Perhaps as new modalities permeate the market more, there will be some recognition and adjustment of CPT codes to reflect that.”? Elizabeth Finch
Elizabeth Finch is a contributing writer for Decisions in Axis Imaging News.
- Hasegawa B, Hawkins RA. Dual-modality imaging with SPECT/CT. Available at: http://www.radiology. ucsf.edu/research/ 05Dual-Modality_ Imaging.shtml. Accessed February 16, 2003.
- Wallis J. Improving the accuracy of cardiac SPECT perfusion imaging. Available at: http://www.medscape.com /viewarticle/416428?WebLogicSession=PkhsnypukKdgFr1ouh2Yk52ljs2aWeXeMZypbr2iNrCAOk9dSpL1|-50199489 34753861993/ 184161392/6/7001/7001/7002/7002/7001/-1. Accessed February 16, 2003.
- Bradley WG Jr. PET/CT fusion imaging captures center stage. Available at: http://www.medscape.com/viewarticle/448633. Accessed February 16, 2003.