| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Nuclear Medicine, National Cheng Kung University Medical Center,
138 Sheng-Li Rd., Tainan 704, Taiwan.
2 Department of Radiology, National Cheng Kung University Medical Center,
Tainan, Taiwan.
3 Department of Psychiatry, National Cheng Kung University Medical Center,
Tainan, Taiwan.
Received November 22, 2007;
accepted after revision February 23, 2008.
This study was partly supported by grants from the Taiwan National Science
Council (NSC 96-2321-B-006-003) and National Cheng Kung University Medical
Center, Tainan, Taiwan (grant 95-109).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Dual-phase FDG PET studies were conducted with imaging 1 and 2 hours after FDG injection, and pulmonary lesions with an initial mean SUV less than 2.5 were identified. Nodules with pathologic reports were included in the study. The differences in mean SUV, maximal SUV, and retention index between benign and malignant pulmonary lesions were analyzed. Receiver operating characteristic analysis was performed to evaluate the discriminating validity of the retention index.
RESULTS. A total of 31 lesions (15 benign, 16 malignant) were included in the study. Among the benign lesions, 12 were granulomatous inflammation, including 10 tuberculosis lesions and two cryptococcosis lesions, and three were focal fibrosis. A retention index greater than 0% was observed in 87% of the benign lesions; 60% of the benign lesions had a retention index greater than 10%. Among the malignant lesions, 75% had a retention index greater than 0%, and 62% had a retention index greater than 10%. We found no significant differences in mean SUV, maximal SUV, and retention index between benign and malignant lesions. The area under the receiver operating characteristic curve did not differ from 0.5.
CONCLUSION. Delayed FDG PET is not useful for differentiating benign and malignant pulmonary nodules with an initial mean SUV less than 2.5 in geographic regions with epidemic granulomatous disease such as tuberculosis or in patients at high risk of granulomatous inflammation.
Keywords: delayed imaging • FDG • inflammation • lung nodules • malignancy
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Some authors [8–10] have claimed that dual-time-point imaging, because it shows increased SUV over time in malignant lesions and decreased or stable SUV over time in benign lesions, has improved the accuracy of FDG PET. Moreover, dual-phase FDG PET has been reported [11] to have the potential to improve accuracy in the evaluation of lung nodules with only borderline levels of increased metabolic activity. Some studies [12, 13], however, have shown that dual-phase FDG PET is ineffective in the evaluation of nasopharyngeal carcinoma and focal abdominal lesions. Another study [14] showed persistently increasing accumulation of FDG in two patients with pulmonary inflammatory lesions. The purpose of our study was to investigate whether PET with acquisition 2 hours after FDG injection is useful in determining the nature of pulmonary lesions when the initial 1-hour mean SUV is less than 2.5.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
For FDG PET, all patients ingested nothing but water for approximately 6 hours. Serum glucose level was measured before injection of the radiotracer to ensure a level less than 150 mg/dL. Diazepam, known to affect the efficacy of dual-phase FDG PET, was not administered before injection of FDG [16]. While in the supine position, each patient was given 370 MBq (10 mCi) of FDG IV and stayed calmly in the supine position for 1 hour after administration. An integrated PET/CT scanner (Biograph, Siemens Medical Solutions) was used to acquire images from the head to the upper part of the thighs. Delayed images of the suspected pulmonary foci were obtained 2 hours after FDG injection.
The images were reconstructed with a standard ordered-subset expectation
maximization algorithm. The axial spatial resolution was 4 mm at the center of
the field of view. For drawing regions of interest, transverse, coronal,
sagittal, and maximum-intensity-projection PET and CT images acquired 1 hour
and 2 hours after FDG injection were displayed simultaneously on a monitor
with the same window and level. The CT images were set in the lung window
setting. The region of interest was drawn in the midportion of the pulmonary
lesion with highest radioactivity, and its margin was approximately 70% of the
maximal radioactivity. Size, mean SUV, and 1- and 2-hour maximal SUVs of the
abnormal pulmonary foci were measured and recorded. The retention index in the
pulmonary lesion was calculated as follows:
 |
CT images of all patients included in the study were reviewed by a radiologist with 16 years of experience after board certification. Interpretation of the CT images was based on imaging features [17, 18]; the reviewer was not aware of the medical history or of other imaging reports.
The significance of the difference in 1-hour mean SUV, 1-hour maximal SUV, 2-hour mean SUV, 2-hour maximal SUV, retention index, and size of the lesions between benign and malignant pulmonary lesions was statistically analyzed with the Mann-Whitney U test. Statistical significance was set at p < 0.05. The receiver operating char acteristic curve was used to evaluate the retention index suitable as a cutoff for malignancy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
We found that 13 (87%) of the 15 benign lesions had further increases in maximal SUV on delayed images (retention index > 0%). Nine (60%) of the 15 benign lesions had a retention index greater than 10% (Figs. 1A and 1B). In the cases of granulomatous inflammation, retention indexes greater than 0% and 10% were observed in 92% and 67%, respectively, of the lesions. Twelve (75%) of the 16 malignant lesions exhibited further increases in maximal SUV on delayed images. Ten (62%) of the 16 malignant lesions had a retention index greater than 10%. When a retention index greater than 0% was used as the cutoff for malignancy, the sensitivity, specificity, and accuracy were 75%, 13%, and 45%, respectively; at greater than 10%, these values were and 62%, 40%, and 52%. The area under the receiver operating characteristic curve was 0.53, not different from 0.50, which was the half probability (p = 0.80). There was no good cutoff because of the decline in both sensitivity and 1 minus specificity as retention index increased.
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although dual-time-point imaging has been reported to improve the accuracy of FDG PET in the evaluation of many kinds of tumors [8, 11, 19–22], results of some studies [12, 13] do not support the claim. Because all of the lesions that we studied had pathologic reports, our findings may provide some clues. Most of the benign lesions in this study were infectious granuloma. Retention indexes greater than 0% and 10% were observed in 92% and 67%, respectively, of these granulomas. Thus we determined that dual-phase FDG PET is inadequate for determining the nature of pulmonary lesions.
Matthies and colleagues [8], using a 10% increase in delayed SUV for pulmonary malignancy rather than a cutoff SUV of 2.5, claimed that the sensitivity increased from 80% to 100% and that the specificity decreased from 94% to 89%. Adopting the same threshold, Xiu and colleagues [11] reported that in evaluation of lung nodules with only borderline increased FDG uptake (initial SUV < 2.5), use of the dual-time FDG technique increased the accuracy of PET from 65.2% (sensitivity, 0.0%; specificity, 100.0%) to 84.8% (sensitivity, 81.3%; specificity, 86.7%). According to the results of those two studies, dual-phase FDG PET seems valuable.
Our results were in contrast to those of Matthies [8] and Xiu [11] and their colleagues. Using the criteria those investigators did, we found that sensitivity and specificity were only 62% and 40%. The reason for the relatively low rate of further increase in delayed SUV in the malignant lesions in our patients is not clear, but the high rate in benign lesions was related to granulomatous inflammation, primarily tuberculosis. A few reports [3, 6, 7] have indicated that granulomatous inflammation can lead to false-positive findings of malignancy on FDG PET. Moreover, further increases in FDG uptake on delayed imaging in cases of granulomatous inflammation have been observed in other studies [11, 15, 23]. In the study by Xiu and colleagues, all four benign pulmonary lesions with a retention index greater than 10% were granulomatous inflammation. We concluded that the efficacy of using dual-phase FDG PET for diagnosing pulmonary nodules with an initial mean SUV less than 2.5 is limited in geographic regions with a high prevalence of granulomatous diseases such as tuberculosis and in patients at high risk of granulomatous inflammation.
High glucose transporter 1 expression levels that lead to high FDG uptake can be found in tumors and inflammatory lesions [24]. In tumors, an increased ratio of hexokinase to glucose-6-phosphatase results in gradual accumulation of FDG and thus in a further increase in SUV on delayed imaging. Varying levels of glucose-6-phosphatase activity among tumor cell types may explain the wide range of FDG retention over time. In contrast, high levels of glucose-6-phosphatase, which causes rapid clearance of FDG and, subsequently, a low ratio of hexokinase to glucose-6-phosphatase, are expressed by mononuclear cells, which represent the major cell population in chronic inflammation and infection [10].
The differential ratio of hexokinase to phosphatase may contribute to the distinct FDG uptake over time observed between inflammatory and malignant lesions. Nevertheless, our results and those of other clinical observations have shown that granulomatous inflammation presents persistently increased FDG uptake. An animal study of dynamic FDG imaging showed a progressive and consistent increase in FDG accumulation in turpentine-induced granulomatous inflammation [25]. In addition, human and animal studies have shown substantial expression of hexokinase in granulomatous tissue [26, 27]. That finding may explain the further increase in FDG uptake in granulomatous inflammation on delayed FDG images. On the other hand, these findings also may indicate that delayed FDG imaging can reveal occult granulomatous inflammatory lesions that cannot be detected with initial FDG imaging.
The study had limitations. First, according to our inclusion criteria, only pulmonary lesions with a 1-hour mean SUV less than 2.5 were analyzed. Therefore, the change in SUV on delayed images in this study did not represent lesions with an initial mean SUV greater than 2.5. This limitation might have resulted in the lower rate of further increases in FDG uptake in our malignant pulmonary lesions than has been found in other series. Second, we acquired delayed images 2 hours after FDG injection, a method also adopted in other studies [8, 9, 11]. It may be possible to differentiate benign and malignant lesions after a longer delay. Third, to explore the possible causes of the controversial efficacy in previous dual-phase FDG PET studies, we included only lesions with pathologic findings in the study, and that requirement might have caused selection bias. Fourth, the numbers of patients and lesions in this study were small. A larger series with a substantially larger number of varied benign and malignant pulmonary lesions is needed to verify our findings.
According to our preliminary data, for pulmonary lesions with an initial mean SUV less than 2.5, further increases in FDG uptake are frequent in granulomatous inflammation. Therefore, delayed FDG PET is not useful for differentiating between benign and malignant pulmonary nodules with an initial mean SUV less than 2.5 in geographic areas with epidemic granulomatous diseases such as tuberculosis or in patients at high risk of granulomatous inflammation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
