SPECT strengths hold up against PET for long term

Given the high quality of FDG-PET imaging, the likelihood that other useful PET tracers will be approved for clinical applications, and the enthusiasm with which the larger radiology community has embraced PET/CT, the future of single-photon scintigraphy in diagnostic imaging is a relevant discussion for nuclear medicine and radiology departments. Decisions have to be made about the allocation of funds, space, and physician training.

It is not reasonable, however, to make decisions based on anticipation of clinical needs beyond a single instrument's life-cycle, such as five to 10 years, as there are too many variables that could affect the practice of diagnostic imaging beyond that time. For purposes of this discussion, I will address the issue as if I were equipping a department today for the next five to 10 years.

Should I invest in SPECT cameras to the same degree that I would have in the pre-PET era? Or, if my existing cameras are reaching the end of their life cycles, should I replace them on a one-for-one basis? I actually had occasion to make this decision three years ago, and my conclusion was "yes," despite the fact that we have a viable PET imaging program and are delighted with the way it has augmented patient diagnosis and management. I would make the same decision today, based on two lines of thought:

- What procedures that are still performed with single-photon tracers remain important clinically and are likely to continue to be important for at least five to 10 years?

- What procedures has PET imaging replaced?

The answer to the first question includes dynamic studies such as renal function and scintigraphy (Tc-99m mercaptoacetyltriglycine [MAG3]); hepatobiliary imaging for evaluation of patients with suspected acute cholecystitis, detection of bile leaks, and assessment of hepatobiliary function and gastrointestinal bleeding; and GI motility studies (gastric emptying, reflux) that typically involve planar imaging.

Other clinically relevant static examinations, acquired either as planar or tomographic studies, include thyroid imaging (both the thyroid gland and whole-body iodine-131 or I-123 imaging of thyroid carcinoma); parathyroid and adrenal imaging (both medullary and cortical); labeled peptide (indium-111-DTPA-pentetreotide, Octreoscan) and antibody (In-111-capromab pendetide, ProstaScint) imaging; cisternography and evaluation of shunts; brain death; ventilation/perfusion scintigraphy; and the majority of nuclear cardiology studies (technetium-99m MIBI/tetrafosmin and thallium-201).

Although fluorine-18 FDG can be used to identify infectious foci, it is less specific than In-111 white blood cell (WBC) scintigraphy, which can also be used in a dual-tracer study if it is necessary to compare bone marrow distribution with Tc-99m sulfa colloid and the In-111 WBC images.

Even skeletal scintigraphy is likely to continue to be performed with Tc-99m tracers (rather than F-18) for at least the next five years. Convenient access to Tc-99m at any time, the resultant flexible scheduling, and likely competition for time on costly PET scanners will account for its continued use.

IMPACT OF FDG-PET

Most of the F-18 FDG-PET imaging in our department is done in patients with tumor to evaluate extent of disease and response to therapy, as well as to monitor for recurrence. Except for the previous use of gallium-67 citrate in Hodgkin's and non-Hodgkin's lymphoma, nuclear medicine had not had much success in these areas.

To be sure, FDG-PET has replaced Ga-67 scintigraphy for this purpose. It has also had a negative impact on the use of other single-photon tracer-labeled tumor diagnostic imaging agents, such as CEA-Scan (Immunomedics) for staging colorectal carcinoma and NeoTect (Berlex) for differential diagnosis of solitary pulmonary nodules. Both agents were just being introduced when FDG imaging became available to clinical departments, initially with dual-detector coincident imaging, and, for the past three years, with dedicated ring detector imaging systems.

The decreased number of requests to our department for Ga-67 imaging has been compensated for by the growth in utilization of other planar and SPECT imaging. With the recent extension of reimbursement to include FDG-PET imaging of the brain for limited applications, I expect that the use of SPECT for assessment of cerebral imaging will decline, but this procedure accounted for only a few studies per week. Cardiology patients are well served with SPECT imaging, and only a small subset require further assessment with PET perfusion or FDG-PET imaging.

Rather than see PET replace SPECT and become the exclusive method for performing functional imaging with radionuclides, I would hope that instrument companies recognize the unique advantages of SPECT radiotracers. Their convenience and availability should encourage those companies to develop better imaging devices for both planar and tomographic imaging of single-photon radionuclides.

Rebuttal to Dr. Goldsmith's Counterpoint

Most metabolic studies that Dr. Goldsmith lists as being justifiable for SPECT in the foreseeable future can and will be performed using PET tracers. Iodine-124, for example, a positron-emitting radioiodine with a half-life of four days, is more effective than either I-123 or I-131 for evaluating patients with thyroid malignancy. MIBG and other radioiodine-labeled compounds can be successfully employed with I-124 rather than with I-131. We have decided not to use I-131 MIBG at all because of its poor performance in assessing patients with pheochromocytoma. I-124 MIBG will be a preferable compound for this purpose.

Gallium 68- or yttrium-86-labeled pentetreotide can be more effective than In-111 octreotide in examining neuroendocrine tumors. In addition, F-18 FDOPA is quite effective in identifying certain neuroendocrine lesions. Metastatic prostate cancer can be successfully examined with radiolabeled amino acids or acetate with positron labeling preparations.

The future of radiolabeled antibodies with single-photon-emitting radionuclides is questionable at this time. I believe that Y-86-labeled anti-CD20 antibody will be the agent of choice in detecting the sites of diseases for treatment with Y-90-labeled antibody. This will be applicable to most antibodies that can be used for therapeutic purposes with Y-90 or other treatment radionuclides.

Rubidium generators or copper-62 ATSM will be used frequently for myocardial perfusion imaging, providing more accurate results than Tl-201- or Tc-99-labeled perfusion compounds.

Studies that determine physiological processes such as blood flow and certain other functions can be successfully performed with single-photon-emitting compounds. But in a few years, FDG will be the agent of choice to detect infection and inflammation, particularly sarcoidosis, regional ileitis, and rheumatoid arthritis.

And it is only a matter of time before we begin employing F-18 fluoride for routine bone imaging. Bone scanning with single-photon-emitting phosphates is insensitive and nonspecific for detecting early disease, determining response to treatment, and detecting recurrence. F-18 fluoride may overcome some of these deficiencies.

I am pleased that Dr. Goldsmith projects a bright future for PET brain imaging in indications for which SPECT is being used at this time. I believe that the demand for PET imaging in central nervous system disorders will be substantially larger than that for SPECT and will include a multitude of diseases, such as movement and seizure disorders and a variety of neuropsychotic diseases. -A. Alavi

Rebuttal to Dr. Alavi's Point

I am also enthusiastic about the high-quality images possible with PET tracers and instrumentation, but it is unrealistic to conclude that PET will replace single-photon imaging within a decade, if ever. One of the reasons is cost. If a satisfactory study can be performed on a $500,000 to $700,000 instrument, why use an instrument that costs up to five times as much? In addition, several studies that are performed with planar imaging of single-photon-emitting radionuclides could not conveniently be performed with the current generation of PET scanners even if the positron-emitting radionuclides were available.

Dr. Alavi's example of the failure of Tc-99m pertechnetate to provide satisfactory brain images ignores the advances made with Tc-99m-labeled brain perfusion agents. Tc-99m HMPAO and Tc-99m ECD have been used to assess stroke size and location, seizure foci, and changes in regional perfusion following skull injury, as well as to differentially diagnose dementia. There is no doubt that FDG-PET provides superior images of cerebral perfusion-but at much greater cost. What would SPECT imaging look like on an instrument that would appropriately be marketed in the $2 million range?

Although it is feasible to produce and distribute F-18-labeled radiopharmaceuticals to many clinical facilities, C-11 labels are simply too short-lived to be clinically relevant. We hear about advances in cyclotron design that will lead to the development of tabletop cyclotrons. But it is not likely that this will occur any time soon, and even if it did, would it be a suitable way for a community hospital to access diagnostic tracers?

It is not reasonable to use a PET scanner (soon likely to be universally PET/CT) for clinical studies such as gastrointestinal bleeding (Figure 1), gallbladder filling and emptying (Figure 2), thyroid imaging, or renal function assessment. These studies do not need the remarkable quantitative assessment that Dr. Alavi cites as justification for universal use of PET imaging. -S.J. Goldsmith

Dr. Goldsmith is chief of nuclear medicine at New York-Presbyterian Hospital/Weill Cornell Medical Center in New York City.