| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Perspective |
1 All authors: Department of Radiology, Duke University Medical Center, Box 3949 Duke University Medical Center, Durham, NC 27710.
Received June 20, 2006; accepted after revision August 1, 2006.
R. C. Nelson is a consultant for GE Healthcare.
OBJECTIVE. Protocols for PET/CT are not yet standardized. In particular, image quality, utilization, and reporting of the findings of the CT component of PET/CT can vary widely, making it complicated for physicians to request the appropriate information. In an effort to address this problem, we describe a set of four PET/CT protocols that satisfy a broad range of clinical needs among oncology patients. Current technology allows acquisition of diagnostic-quality CT scans as part of PET/CT examinations, and referring physicians are given the option of requesting formal interpretation of the CT findings. In this case, the PET and CT images are interpreted by the corresponding specialists, and equivocal or discordant findings are adjudicated through joint review of the PET/CT images.
CONCLUSIONS. The menu of PET/CT imaging protocols has gained wide acceptance by our referring physicians and have been used successfully in more than 6,000 PET/CT studies. Newer PET/CT protocols will be developed as technology advances. Continued collaboration among oncologists, CT specialists, and nuclear medicine specialists is essential for deriving the maximum clinical benefit from combined PET/CT. Standardization of imaging protocols will become increasingly important as multiple-institution trials are developed for evaluation of present and future applications of PET/CT.
Keywords: abdominal imaging • chest • CT technique • oncologic imaging • PET
Currently available PET/CT scanners combine the capabilities of state-of-the-art PET with MDCT, and a wide range of possibilities exist for PET/CT. In particular, use of the CT images varies widely among institutions. On one end of the spectrum, low-dose CT images are used primarily for attenuation correction of PET images and show limited anatomic localization of the PET-defined abnormalities. On the other end of the spectrum, a single CT examination can be performed with the goal of obtaining diagnostic-quality images (equivalent to a separately obtained CT scan) and attenuation correction. In the latter approach, CT is performed with the appropriate use of IV and oral contrast agents and the understanding that the images will be interpreted in the same manner as any other diagnostic CT scans. There are advantages and disadvantages to each approach. Although many of the issues concerning PET/CT have been summarized [1], controversy remains about the best imaging technique, and no standard protocols have been established.
There are two major algorithms that a referring clinician may consider for PET and CT (Fig. 1A, 1B). The two studies can be obtained serially or be combined. In the serial approach, diagnostic CT is performed first. If the clinical question (e.g., disease progression) is answered with CT, PET is not necessary. If a clinical question remains unanswered, the patient is scheduled for a separate PET scan. The primary advantage of this approach is the potential cost and time savings due to selective utilization of PET. Another advantage of the serial approach is that the CT and PET protocols can be optimized individually without the constraints of combined PET/CT. For example, CT can be performed at full inspiration, and multiphase or dynamic contrast enhancement can be more easily accommodated. Finally, the serial approach can be used by institutions that have a PET-only scanner and this approach does not require a hybrid PET/CT system.
|
|
|
A disadvantage of the combined PET/diagnostic CT approach is that compromises must be made for both PET and CT. These compromises, however, are becoming less of an issue as newer PET/CT scanners are being developed. Our experience has been that the compromises made to perform both studies at the same time either are acceptable or have been adequately addressed with improved technology. For PET/diagnostic CT to be a viable alternative to conventional imaging, both the CT and the PET images should provide at least equivalent diagnostic information as when the two studies are performed separately. Through a joint effort by the nuclear medicine, abdominal imaging, and cardiac and thoracic imaging divisions at Duke University Medical Center, combined PET/diagnostic CT protocols have been developed to meet this objective. This combined approach is used for all patients for whom both PET and CT are indicated.
We give referring clinicians a menu of PET/CT options (Fig. 2), allowing the PET/CT examinations to be tailored to the needs of the patient. When diagnostic CT scans are obtained (with or without contrast material), a separate CT report is provided to the referring clinician, and the patient is billed for both studies. Since we began offering PET/diagnostic CT in January 2004, we have performed more than 6,000 PET/CT examinations and more than 1,500 PET/diagnostic CT studies. We describe our experience developing this integrated program.
Implementation of Combined PET/Diagnostic CT
Equipment and Personnel
Imaging is performed with a PET/CT scanner (Discovery ST, GE Healthcare)
with 16-MDCT. A programmable dual-syringe power injector (Stellant D, Medrad)
is used for IV injection of contrast material. A video camera is in place in
the scanner room for visual monitoring of the patient during injection and
imaging. PET/CT scans are obtained by certified nuclear medicine technologists
specializing in PET who have received additional training in CT. The staff
also includes a full-time nurse to screen and interview patients before PET/CT
and to monitor the patients during and after IV injection of the contrast
agent.
|
Option 1 (PET/CT) is a "PET-alone" PET/CT study in which the referring physician is given a PET report only. In this case, the CT images are reviewed, but the findings are not formally reported. All patients (except patients with chest or head and neck cancer) routinely receive an oral contrast agent. CT findings of potential diagnostic significance are included in the PET report. This option is used most frequently for imaging of patients who have recently undergone diagnostic CT and for assessment of the response to therapy of patients who have undergone PET.
Although CT results are not formally reported, we obtain diagnostic-quality CT scans of these patients. We have found that diagnostic-quality scans are needed for accurate anatomic localization. More important, use of diagnostic-quality scans leads to more accurate assessment of incidental findings, such as pulmonary nodules and low-attenuation renal, adrenal, and hepatic lesions. In many cases, the morphologic features and attenuation values on CT scans help reviewers differentiate lesions that are clearly benign from those that are indeterminate or suspected of being malignant. The information obtained with CT leads to appropriate recommendations and reduces the number of unnecessary follow-up studies.
Option 2 is PET combined with thoracic, abdominal, and pelvic CT performed with oral and IV contrast enhancement. In this case, the referring clinician is given separate PET and CT reports, and the patient is billed for both studies. This option is most appropriate when both PET and diagnostic CT studies are indicated. Referring oncologists frequently request PET/diagnostic CT with contrast enhancement for initial staging studies and for evaluation of diseases and conditions such as liver lesions in which IV contrast material is needed for accurate assessment.
Option 3 is identical to option 2 except that IV contrast material is not administered. A separate report for unenhanced thoracic, abdominal, and pelvic CT is provided, and the patient is charged for it. This option may come to be used more frequently because we and referring physicians are recognizing that IV contrast material may not be necessary when PET is performed concurrently with CT. For example, IV contrast material may not be necessary if the primary pathologic finding is in the chest or if previous CT scans are available. Certain clinical indications for which PET alone has high sensitivity, such as melanoma, also may not necessitate use of IV contrast material, but these indications need to be defined. Option 3 also provides an alternative for patients for whom IV contrast material is contraindicated. The PET and CT studies are otherwise rendered formal interpretations in the same way as the PET/diagnostic CT studies in option 2.
Option 4 is identical to option 3, except that the chest CT findings are reported by a thoracic radiologist. At our institution, chest CT is routinely performed without IV contrast material, so for this protocol IV contrast material is not administered. This option is frequently used for staging or for follow-up of patients with lung or esophageal cancer. The abdominal and pelvic CT images are reviewed, as in option 1, but the findings are not formally reported.
This simple menu greatly facilitates communication between the ordering physicians and our staff in determining the type of PET/CT to meet the specific clinical need. At the same time, the options provide flexibility with respect to the type of CT study desired. Utilization of the imaging options over time is shown graphically in Figure 3. Referring clinicians have found combined PET/diagnostic CT particularly valuable, and the proportion of these studies has increased over time. At our institution, combined PET/diagnostic CT has effectively replaced separate PET and CT studies. Combined PET/diagnostic CT has also gained uniform acceptance from patients, who prefer the convenient one-stop approach to these imaging studies.
Potential Limitations of Combined PET/Diagnostic CT
There are two fundamental constraints to ideal combined PET/CT. First, CT
scans are typically obtained with the patient in suspended respiration (full
inspiration), whereas PET images are obtained with the patient breathing
quietly, which can result in image misregistration between the PET and CT
images. Second, CT images obtained at lower energies (typically 120-140 kVp)
are used for attenuation correction of PET images (511 keV), and this
technique can cause inaccuracies in attenuation correction. The
attenuation-correction problem is further complicated when IV contrast
material is used, because vascular and tissue enhancement present during the
CT scan can be cleared (and the attenuation properties different) at PET image
acquisition.
|
|
|
|
Technical Considerations for Combined PET/Diagnostic CT
CT-based attenuation correction—CT-based attenuation
correction is subject to error because it requires conversion of the X-ray
attenuation factors obtained at diagnostic CT energy (120-140 kVp) to
attenuation at PET energy (511 keV). The problem is further complicated by the
fact that the spatial distribution of high-density oral and IV contrast agents
can be quite different during the CT and PET acquisitions. The presence of
high-density contrast material can result in overestimation of the attenuation
factors and falsely increase activity on the PET images
[6].
Similar overcorrection artifacts are caused by pacemakers or metallic implants. These artifacts can be clinically insignificant [7]. If not, they can be suppressed with multiphase IV injection techniques (contrast agent followed by saline flush) to reduce the effects of high contrast density [8]. The use of negative oral contrast agents, such as water and dilute barium preparation (VoLumen, E-Z-EM), can also reduce these effects. The extent to which contrast material causes artifacts on PET/CT and thus necessitates use of dual-phase IV injection or low-density gastrointestinal contrast material is highly dependent on the scanner and the associated CT attenuation-correction software.
Software techniques have been developed to reduce the error attributable to CT attenuation correction. For example, one premise is that iodinated contrast agents have higher density than any physiologically encountered material. The attenuation-correction factor can then be restricted to avoid overcorrection at high radiodensity [9]. We have found that when these newer attenuation-correction algorithms are applied, oral and IV contrast agents no longer cause significant artifacts on PET images (Fig. 4A, 4B, 4C, 4D). Other authors [10, 11] have found that contrast materials do not cause significant artifacts when newer CT attenuation-correction techniques are used.
Image registration between PET and CT— PET images are acquired over several minutes per table position with the patient quietly breathing. To minimize motion artifacts, CT scans should be obtained with the patient in suspended respiration. The best image registration between CT and PET images is obtained when the patient suspends respiration at end-tidal volume (quiet end-expiration), because the diaphragm spends the most time in this position during quiet respiration.
|
|
|
|
Accurate alignment of the PET and CT images requires that the patient remain still throughout the study. Comfort is important, and the patient is held securely on the scanning table with blankets and hook-and-loop (Velcro) straps. The patient is then positioned on the scanner with both arms up. An acrylic board with a set of handles on the scanner table above the patient's head is used to help the patient comfortably maintain the arms-up position and eliminate motion during PET and CT (Fig. 5). This device (Wing Board, model MT-WB01, Medtec) is the one used routinely for patients undergoing radiation oncology treatments. For all PET/CT examinations, the CT scan is obtained first in the craniocaudal direction. As soon as CT is complete, PET proceeds in the opposite direction beginning at the proximal part of the thighs. This technique minimizes misregistration in the pelvic organs due to normal peristalsis and bladder filling. For head and neck cancer patients, a second PET/CT acquisition is performed with the arms down to image the neck region (see later). In each case, keeping the arms out of the field of view is essential to maintain high quality of the CT images and to reduce truncation artifacts at the edge of the CT field of view.
Image quality of end-expiration chest CT— Over the past 2 years, the thoracic imaging division has interpreted more than 300 diagnostic chest CT scans obtained as part of PET/CT studies. Because these studies are obtained in quiet end-expiratory phase rather than full inspiration, the chest images from PET/diagnostic CT studies show lower lung volumes. More dependent atelectasis and ground-glass opacity typically are seen as a background on these images than on CT scans obtained at full inspiration (Fig. 6A, 6B). It is possible that small nodules are obscured by crowding of vascular structures, and comparison with previous chest CT scans can be difficult. Even after accounting for these factors, we have found that chest CT images obtained at end-tidal volume are adequate in the patient population referred for oncologic evaluation and undergoing concurrent PET. We have not found it necessary to perform follow-up full-inspiration CT after PET/CT.
In a study of 66 patients with known solid malignant tumors undergoing combined PET/CT, Juergens et al. [12] found that acquisition of an additional low-dose chest CT scan at full inspiration significantly improved sensitivity for detection of small pulmonary nodules. Therefore, although not currently a part of our PET/CT protocol, the indications and potential value of additional inspiratory chest CT warrant further investigation.
CT tube current modulation—Diagnostic CT exposes patients to a significantly higher radiation dose than do 68Ge transmission CT and low-dose CT (5 mAs) used for attenuation correction. Because the CT scan for PET/CT must be obtained in a single acquisition, optimization of CT parameters for the thoracic and abdominopelvic scans is difficult. At our institution, we typically use tube currents of 240 mA (effective tube current, 120 mAs) for chest CT and 340 mA (effective tube current, 170 mAs) for abdominopelvic CT. To maintain high image quality in the abdomen and pelvis while reducing dose to the chest and neck, automated CT tube current modulation is important. Techniques for automated CT tube current modulation vary among scanner manufacturers and have been summarized by Kalra et al. [13]. In our scanner, automated CT tube current modulation is implemented with the AutomA feature (GE Healthcare). With this feature, the tube current is adjusted in the z-direction according to the anteroposterior and lateral scout images obtained during the scan to maintain a specified noise factor [14]. In Figure 7, the tube current used to produce a diagnostic-quality CT scan from the skull base to the midthigh level is shown as a function of scan level.
Imaging Protocols
Details of the specific imaging protocols for all four options are in Tables 1, 2, 3 and 4. Before all PET studies, patients should have no caloric intake for at least 4 hours and preferably for more than 6 hours before the scan, although water intake is encouraged. Patients are asked to complete a questionnaire to provide general information, including clinical history, previous CT and PET studies, allergies, previous contrast reactions, and presence or absence of diabetes. Patients are given oral contrast material as follows: 1 mL of diatrizoate meglumine and diatrizoate sodium (37% organically bound iodine, Gastrografin, Bracco Diagnostics) per 30 mL of fruit-flavored beverage (Crystal Light, Kraft Foods), 900 mL in three divided doses over 90 minutes. Oral contrast administration is optional for patients being evaluated for head and neck or lung cancer.
|
|
|
|
Serum glucose is measured for all patients and should be 
200 mg/dL
before the PET study. Patients are given an IV injection of 18F-FDG
(5.2 MBq/kg; maximum dose, 740 MBq). The patient rests quietly for 1-2 hours
during the uptake phase. CT is performed with or without IV contrast
administration in the craniocaudal direction. When an IV contrast agent
(iopamidol 61%, 30% organically bound iodine, Isovue-300, Bracco Diagnostics)
is given, timing of the IV contrast bolus is optimized using automated
triggering with serial low-dose CT (SmartPrep, GE Healthcare). IV contrast
material (150 mL) is injected at 3 mL/s and is followed by a 30-mL saline
flush. The CT scan from the skull base to midthigh level takes approximately
16 seconds (Tables 1 and
2). Immediately after CT, PET
emission images are acquired starting with the most inferior table position
and moving in the caudo-cranial direction. This step minimizes misregistration
of the PET and CT images due to bowel transit and bladder filling.
Patients scheduled for options 1, 3, and 4 are allocated a 30-minute time slot on the PET/CT scanner. Longer time slots (45 minutes) are allocated for patients scheduled for option 2 to allow time for setup of the IV contrast injection. Forty-five-minute time slots also are allocated to patients undergoing whole-body (vertex to toes) studies or additional head and neck imaging (described later).
Image Interpretation
The combined PET and CT images are reviewed on an interactive workstation (Advantage Windows, GE Healthcare). These workstations allow parallel display of PET and CT images, multiplanar reconstruction, comparison with previous PET/CT images, and image fusion. PET images are reviewed by nuclear medicine specialists, and CT images are reviewed by the appropriate radiology staff. Thoracic, abdominal, and pelvic CT scans with and without contrast enhancement are interpreted by the abdominal imaging division, and chest CT scans without contrast enhancement are interpreted by the thoracic imaging division. These two divisions have dedicated PET/CT workstations for reviewing these studies.
This factor is important because the PET and CT interpretation areas are in different locations at our institution. Although PACS facilitates joint review of images, our current PACS workstations do not have the capabilities of image fusion and multiplanar reformatting. Therefore, it became essential in our practice to have PET/CT workstations at both locations for joint review of the images. With this configuration the combined study can be discussed over the telephone. Additional specialists in pediatric radiology, musculoskeletal imaging, and neuroradiology are consulted as needed.
The referring clinician is provided separate PET and CT reports. The body of each report ideally contains technique-specific findings and comparison with previous studies, and the conclusions are identical for both the CT and PET reports, giving a concise diagnostic impression formulated from both techniques. Communication and consensus among the physicians interpreting the PET and CT studies are crucial for maximizing the mutual information obtained with PET/CT and to provide consistent reporting to referring clinicians. The number and type of PET/CT studies according to diagnosis are shown in Table 5. Contrast-enhanced studies are most frequently requested for patients with esophageal cancer, lymphoma, colorectal cancer, and breast cancer. IV contrast material is used less frequently in examinations of patients with melanoma and lung cancer.
|
Additional Techniques
Whole-Body Imaging
When whole-body imaging (vertex to toes) is requested by referring
physicians, PET/CT acquisition extends from the top of the head to the
proximal part of the thighs according to the PET and CT or diagnostic CT
parameters used for conventional studies. The patient position then is
reversed on the scanner table for 2-minute emission images of the legs. CT is
performed with automated CT tube current modulation and the parameters shown
in Tables 1 and
2.
Pediatric Patients
Pediatric patients receive the same weight-adjusted dose of FDG as adult
patients. For PET/diagnostic CT of pediatric patients, we use the weight-based
color-coded system adopted by our pediatric radiology division for determining
CT scan parameters [15]. IV
and oral contrast doses are reduced accordingly. Other CT parameters (detector
configuration, table speed, rotation, and pitch) are the same as for adults,
except that the technique is reduced to 120 kVp. As with adults, automated CT
tube current modulation is used.
Head and Neck Tumors
Head and neck tumors are a special case, and we work directly with the
clinical subspecialists for tailoring these PET/CT studies. For example, an
immobilization device may be in place for radiation oncology treatment
planning. Two separate PET/diagnostic CT acquisitions are performed to
optimize the body and neck images. The body is imaged from midneck down in the
usual manner with the arms up. If IV contrast material is given for body CT,
the volume is reduced to 125 mL. After completion of body imaging, an
additional 8-minute scan (single bed position) of the neck is obtained with
the arms down and with a reduced (30-cm) field of view. We performed a study
[16] in which we found that
these dedicated images are particularly valuable for defining and evaluating
small lymph nodes in patients with head and neck cancer. When contrast
enhancement is used for these neck images, an additional 75 mL of contrast
agent is given IV at 2 mL/s with a 60-second delay between injection and
imaging. For these patients, the referring physicians can request a separate
head and neck CT report from the neuroradiology section.
Conclusions
New PET/CT scanners combine the diagnostic capabilities of MDCT and state-of-the-art PET. As of this writing, there is no consensus about or standardization of PET/CT protocols. The quality of CT scans obtained with PET studies and interpretation of these images are areas for discussion [17].
PET/CT is an emerging imaging technology, and standardization of imaging protocols is increasingly important for establishing the efficacy of PET/CT for specific clinical applications. Standardization will become a critical element in the design of multiple-institution clinical trials of PET. We and others have found that PET images can be obtained in parallel with high-quality contrast-enhanced CT scans for simultaneous diagnosis, anatomic correlation, and attenuation correction of the PET images. However, not all patients undergoing PET need contrast-enhanced CT.
We developed four PET/CT protocols that allow referring physicians to select from a range of reporting options. These protocols have effectively fulfilled the needs of our referring physicians for more than 2 years. Approximately 30% of our PET/CT studies include formal diagnostic CT interpretation. Our approach is to have the PET and CT studies interpreted jointly, combining the expertise of the appropriate radiology specialists. Communication and consensus among the interpreting imaging specialists are essential for provision of consistent and concordant reports to referring clinicians. This service is made possible through use of PACS and remote workstations.
It is likely that the number of PET studies done in combination with diagnostic CT will continue to increase. Further investigation is needed for clear definition of the specific indications for the various imaging protocols, and new protocols will continue to be developed. In particular, the role of IV contrast enhancement in the context of PET/CT warrants continued investigation. On one hand, more sophisticated multiphase contrast enhancement protocols may improve lesion identification and characterization of lesions on CT. On the other hand, the need for IV contrast enhancement may decrease when CT is combined with PET. For example, PET/diagnostic CT with contrast enhancement may be appropriate for initial staging of lymphoma, but follow-up imaging may be adequate with PET/diagnostic CT without IV contrast enhancement or PET/CT without a separate CT report.
In addition, there may be indications for which IV contrast administration is not necessary when CT scans are obtained as part of PET/CT. For example, we are investigating the possibility that IV contrast enhancement may not contribute significant additional information in PET/CT of patients with melanoma. If these indications can be established, the ability to perform these studies without IV contrast agents may save time, reduce cost, and eliminate contrast agent-associated risk to patients.
Ongoing collaboration among oncologists, radiologists, and nuclear medicine specialists will be essential for continued development of PET/CT protocols and for maximizing the benefit of combined PET/diagnostic CT studies.
Acknowledgments
We thank Jacqueline Ison, Mary Hawk, Robin Davis, Thomas Hawk, and Timothy Turkington for contributions to the development of the PET/CT protocols.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L. M. Ho, E. K. Paulson, M. J. Brady, T. Z. Wong, and S. T. Schindera Lipid-Poor Adenomas on Unenhanced CT: Does Histogram Analysis Increase Sensitivity Compared with a Mean Attenuation Threshold? Am. J. Roentgenol., July 1, 2008; 191(1): 234 - 238. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Gollub, R. Hong, D. M. Sarasohn, and T. Akhurst Limitations of CT During PET/CT J. Nucl. Med., October 1, 2007; 48(10): 1583 - 1591. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
