| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 All authors: Department of Radiology, University of California, San Francisco, Box 0628, M-372, 505 Parnassus Ave., San Francisco, CA 94143-0628.
Received January 29, 2008;
accepted after revision May 9, 2008.
Address correspondence to F. V. Coakley
(fergus.coakley{at}ucsf.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We retrospectively identified 39 patients (25 men and 14 women) with suspected isolated hepatic metastases from colorectal cancer who underwent FDG PET/CT. The CT protocol included acquisition of unenhanced and multiphase contrast-enhanced CT images through the liver. At two separate sittings, four readers (two radiologists and two nuclear medicine physicians) noted and characterized all hepatic lesions in consensus, first based on PET and unenhanced CT images and later based on PET and contrast-enhanced CT images. The nature of detected lesions was established by histopathologic or clinicoradiologic correlation.
RESULTS. A total of 178 hepatic lesions were identified, consisting of 137 metastases and 41 benign lesions. Using lesion-based analyses with Obuchowski's method for paired observations, 172 of 178 lesions (97%) were detected at PET/contrast-enhanced CT compared with only 135 of 178 (76%) at PET/unenhanced CT (p = 0.0004). Specifically, 114 of 137 (83%) hepatic metastases were detected on PET/contrast-enhanced CT compared with 92 of 137 (67%) on PET/unenhanced CT (p = 0.012). One hundred thirty-one of 178 lesions (73%) were accurately characterized at PET/contrast-enhanced CT compared with 101 of 178 (57%) at PET/unenhanced CT (p = 0.004).
CONCLUSION. IV iodinated contrast material administration improves the detection of hepatic metastases and the characterization of focal hepatic lesions at PET/CT.
Keywords: 18F-FDG • colorectal cancer • liver • metastatic liver disease • PET/CT
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Imaging Technique
All FDG PET/CT studies were performed on a Biograph 16 (Hi-Rez) PET/CT
scanner (Siemens Medical Solutions), which is a hybrid PET and 16-MDCT
scanner. Patients fasted for at least 6 hours before the PET/CT examination.
Six patients with diabetes were instructed to follow a low carbohydrate diet
the evening before and to withhold all insulin for 4 hours preceding the
PET/CT examination. In these patients, the blood glu cose level was checked by
the finger-stick method before injection of the radiopharmaceutical. IV
injection of 12.5 ± 2.5 mCi of FDG was followed by a 10-mL normal
saline flush. Patients rested for 60 ± 15 minutes and voided before
being positioned su pine on the scanner table. Unenhanced CT was performed
first, from the vertex of the skull through the mid thigh at 152 mAs, 120 kVp,
and 0.75-mm collimation. Images were reconstructed as contiguous 5-mm slices.
Multiphase contrast-enhanced CT was then performed. Early and late arterial
phase CT images were obtained 20 and 45 seconds, respectively, after starting
the injection of 150 mL of iohexol (Omnipaque 350, GE Healthcare). Contrast
material was injected at 3 mL/s using a power injector (Stellant D, Medrad).
Both sets of arterial phase images were obtained from the dome of the
diaphragm to the iliac crest at 180 mAs, 120 kVp, and 0.75-mm collimation.
Arterial phase images were reconstructed as contiguous 1-mm slices. Portal
venous phase images were acquired after a delay of 80 seconds from the vertex
of the skull through the mid thigh at 180 mAs, 120 kVp, and 1.5-mm
collimation. Portal venous phase images were reconstructed as contiguous 2-mm
slices. No oral contrast material was administered. PET was performed
immediately after CT, without repositioning the patient. PET images were
obtained at 7–10 stations per patient, with an acquisition time of 4
minutes per station, from the skull vertex through the mid thigh. The
unenhanced CT data were used for attenuation correction of PET emission
images, which were coregistered with all contrast-enhanced CT datasets. The
reconstructed CT, PET, and fused PET/CT images were displayed in axial,
sagittal, and coronal planes.
Image Interpretation
All PET and unenhanced CT images and then all PET and multiphase
contrast-enhanced CT images were reviewed at two different sessions separated
by at least 6 weeks. For both sessions, PET images included both
attenuation-corrected and non-attenuation-corrected images. Readings were
performed in consensus by a panel of two nuclear medicine physicians and two
abdominal imaging radiologists. Readers were aware that patients were
suspected of having liver metastases but were unaware of all other clinical,
radiologic, and histopathologic results. Studies were reviewed simultaneously
on both a PACS workstation (Impax, Agfa) and an image processing workstation
(Leonardo, Siemens Medical Solutions).
At each review session, the readers recorded the segmental location, size, CT density, maximum standardized uptake value (SUVmax), and diag nostic impression for all detected focal hepatic lesions on a standardized template. The diagnostic impression consisted of both the most likely specific diagnosis and the overall likelihood of malignancy (0–100%). In patients with more than 10 focal hepatic lesions, only the 10 largest lesions at CT were recorded. Although the specific diagnoses were made on the basis of the expert judgment of the consensus panel, certain general criteria can be described. For example, hepatic metastasis was considered present when a hypo-attenuating or heterogeneous mass measuring more than 10 mm in diameter was seen at CT or a focus of increased FDG uptake (SUVmax greater than 2.5) was seen in the liver at PET, whereas a hepatic cyst was considered present when an unenhancing mass of fluid density (–20 to 20 HU) greater than 10 mm in diameter was seen at CT and was not accompanied by increased FDG uptake. Hypodense lesions in the liver under 10–15 mm in diameter and not associated with increased FDG uptake were considered too small to accurately characterize, but the consensus panel was allowed to assign a rating as to the likelihood of malignancy, bearing in mind that such lesions rarely, if ever, progress to metastases [6].
Standard of Reference
The primary investigator, who was not one of the PET/CT readers, correlated
all detected lesions with findings at histopathology, serial imaging
evaluation, and details of cancer therapy. Imaging techniques contributing to
the final standard of reference based on lesion evolution on serial studies
included CT, MRI, sonography, and PET/CT (these additional PET/CT examinations
were used only for evaluation of the study PET/CT findings and were not
included in the analysis and were not available to the readers at the time of
imaging interpretation). The mean interval between the study PET/CT and
follow-up imaging or histopathology was 109 days (range, 0–326 days).
The mean interval between prior imaging and the study PET/CT was 188 days
(range, 9–450 days). In the absence of histopathologic proof, a lesion
was considered a metastasis if additional imaging showed unequivocal findings
of malignancy or if a lesion progressed on follow-up or if a lesion regressed
after chemotherapy. The treated sites of metastatic disease were still
considered metastases after chemotherapy but were considered as benign after
surgical ablation (unless imaging progression suggested residual or recurrent
disease at the ablation site). Hypodense lesions that were too small to
accurately characterize by CT and that were not associated with increased FDG
uptake were considered benign if serial imaging showed stability. Two patients
with seven hepatic lesions that were all considered unclassifiable were
excluded from the initial study group.
Statistical Analysis
For comparing the accuracy of PET/contrast-enhanced CT and PET/unenhanced
CT, we used Obuchowski's method, which extends the usual McNemar test for
paired proportions to the case in which the observations are sampled in
clusters [7]. Usually a McNemar
test would suffice for an analysis of lesions evaluated on two different
tests, but in this study many patients had multiple lesions, which violates
the independence assumption in the McNemar test. Obuchowski's method was
specifically designed to address this problem. The rating of suspiciousness in
the overall interpretation was used to dichotomize lesion characterization; a
threshold of 50% was set as the threshold value to distinguish benign versus
malignant lesions. Diagnostic performance was evaluated by lesion detection
and characterization. A lesion was considered accurately characterized if the
lesion was rated as benign (or malignant) by the consensus panel, and the
lesion was found to be benign (or malignant) by the standard of reference.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Diagnostic Performance of PET/CT
One hundred seventy-two of 178 lesions (97%) were detected at
PET/contrast-enhanced CT compared with 135 of 178 (76%) at PET/unenhanced CT
(p = 0.0004). Specifically, 114 of 137 (83%) hepatic metastases were
detected on PET/contrast-enhanced CT compared with 92 of 137 (67%) on
PET/unenhanced CT (p = 0.012). One hundred thirty-one of 178 lesions
(73%) were accurately characterized at PET/contrast-enhanced CT compared with
101 of 178 (57%) at PET/unenhanced CT (p = 0.004). A representative
case showing improved characterization with IV iodinated contrast material is
shown in Figure 1A,
1B,
1C.
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The term "nondiagnostic CT" is often used to refer to an unenhanced CT study performed as part of a PET/CT examination, but this term is clearly inappropriate—the CT data are not nondiagnostic, although the data may be less diagnostically useful than the data provided by a complete contrast-enhanced CT. We believe that both components of PET/CT should be performed to the highest possible standard and should routinely include the administration of iodinated IV contrast material and that such an approach is supported by our data. If this view of our study is accepted and validated in further, larger prospective studies, then the current common practice of performing PET/CT without contrast material will need to be reconsidered.
Similarly, the euphemistic term "non-diagnostic CT" as a synonym for unenhanced CT may need rethinking. These conclusions, although somewhat speculative, merit discussion because there are important impli cations for patient preparation, staffing, image interpretation, and possibly billing if the administration of IV iodinated contrast material becomes a routine part of PET/CT practice. We are mindful that such an adjustment is historically reminiscent of both body CT and MRI, which were initially performed with only limited and selective use of IV contrast material. We suspect a similar paradigm shift will develop for PET/CT.
We are aware that other studies investigating the role of IV iodinated contrast material administration in PET/CT have had mixed results and the concept remains controversial [2, 4, 5, 9–12]. Pfannenberg et al. [10] suggested contrast-enhanced PET/CT protocols are dispensable for plan ning 3D conformal radiation therapy and unconventional extended surgical resection in patients with lung cancer. Rodríguez-Vigil et al. [9] in a prospective study showed a good correlation between unenhanced low-dose PET/CT and contrast-enhanced full-dose PET/CT for lymph node and extranodal disease in lymphomas, suggesting that unenhanced low-dose PET/CT might suffice in some selected cases. Conversely, in a different study, Pfannenberg et al. [4] studied the additional value of contrast-enhanced multiphase CT in comparison with low-dose unenhanced CT in combined PET–PET/CT protocols for 100 patients with different malignant tumors in a retrospective study and concluded that diagnostic multiphase CT as part of the combined PET/CT protocol has the potential to provide considerable additional value in specific clinical conditions, with resultant change of management in a substantial proportion of patients. Likewise Cantwell [5] showed the superior value of contrast-enhanced CT over unenhanced CT in PET/CT protocols in evaluation of patients with metastatic liver lesions from colorectal cancer; a total of 112 hepatic lesions were evaluated in 33 patients with colorectal hepatic metastases, and overall detection with PET/contrast-enhanced CT was statistically superior to PET/unenhanced CT (90.9% vs 73.6%, p < 0.05).
Our study has a number of limitations. It was a single-institution retrospective study. The results of our study may have been influenced by our study population, with a majority, who were surgical candidates, presenting with colorectal primary and advanced metastatic hepatic diseases. As often occurs in studies in which tumor detection is assessed, a histologic diagnosis was not available for all detected hepatic lesions, and many lesions were characterized by clinicopathologic correlation. Image interpretation was performed by consensus, using a panel of four expert readers. Thus, our data did not allow us to examine the degree of interobserver variability in the evaluation of PET/CT findings. Although whole-body images were available for all patients, we chose to focus on hepatic lesions, and we did not examine the role of iodinated contrast material for extrahepatic findings. The CT technique in our study included acquisition of multiphase contrast-enhanced images, but we did not analyze which phase of contrast enhancement provided the incremental benefit with respect to lesion detection and characterization. Although the published literature does not strongly support the use of arterial phase imaging for the detection of traditionally hypovascular masses such as colorectal cancer metastases to the liver, it has been suggested that arterial phase imaging may improve the characterization of small hypoattenuating metastases because these lesions may have a hyperemic rim in the arterial phase that is not seen with small cysts [13, 14].
In conclusion, IV iodinated contrast material improves the detection of hepatic meta stases and the characterization of focal hepatic lesions at FDG PET/CT, and IV iodinated contrast material should be con sidered as part of the protocol for PET/CT when optimal detection of hepatic metastases is critical, such as in the preoperative evaluation of patients with suspected isolated hepatic metastases to the liver.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
