A New Method for Detecting Urinary Bladder Cancer by Using FDG
PET/CT
Yen-Kung Chen, M.D., Ph.D.
1,4Chung-Hsin Yeh, M.D.
2,4Chen-Tau Su, M.D.
3Chih-Cheng Tsui, M.Sc.
1Ru-Hwa Cheng, M.Sc.
1Guang-Uei Hung, M.D.
5Chia-Hung Kao, M.D.
6,7*
Introduction
The bladder is the most frequent site of urinary tract cancer, which is the fourth most common cancer in men and the ninth most common cancer in women in the United States (1). More than 95% of primary bladder cancers are of epithelial origin, most commonly transitional-cell (urothelial) carcinoma (2). Fortunately, most bladder cancers are diagnosed early, before spreading beyond the bladder. Eight-five percent of bladder cancer cases are localized in the bladder. The incidence of bladder cancer tends to increase considerably with age; approximately two-thirds of cases occur in people aged 65 years and older (3).
Fluorine-18-2-fluoro-2-deoxy-d-glucose (18F-FDG) positron emission tomography (PET) is used to
detect a wide variety of tumor foci. However, high levels of 18F-FDG activity in urine can mask
abnormal activity in the urinary bladder. To overcome this limitation, some authors have advocated retrograde irrigation of the urinary bladder during PET data acquisition (4, 5) or delayed imaging after forced diuresis coupled with parenteral hydration (6), which are methods that enhance urinary flux and allow rapid evacuation of the urinary bladder. These methods are useful for detecting suspected lesions of the bladder; however, bladder catheterization and continuous irrigation are too invasive for routine use.
When 18F-FDG is injected intravenously in humans, non-metabolized 18F-FDG is eliminated by
glomerular filtration without complete reabsorption into the bladder. Heightened 18F-FDG
accumulation in the bladder and similarly increased 18F-FDG uptake in malignant lesions may be
misinterpreted. We hypothesized that using 18F-FDG PET/computed tomography (CT) with an
18F-FDG foci. The purpose of this study was to evaluate the diagnostic accuracy of 18F-FDG PET/CT
when applying an IULIT for detecting urinary bladder cancer.
Materials and Methods
Participants
This study followed the guidelines issued by the institutional review board (ethics committee) ethics committee on human studies of the hospital and the Helsinki Declaration. In this retrospective study conducted from February 2001 to January 2010, the inclusion criteria required that participants without a history of bladder cancer underwent FDG PET or PET/CT scans as part of their clinical care or screening program. In total, 28,973 consecutive patients (mean age, 52.4 15.6 years [standard deviation]; age range, 2 to 96 years) consisting of 14,631 men (mean age, 53 12 years; age range, 2 to 95 years) and 14,342 women (mean age, 51.6 23.3 years; age range, 5 to 96 years) participated. The sample comprised 6224 patients with a history of cancer (mean age, 56.5 15.6 years; age range, 2 to 96 years), who were undergoing 18F-FDG PET or PET/CT scans as part of their clinical care, and
22,543 participants without a history of cancer (mean age, 51.2 11.3 years; age range, 23 to 94 years), who were receiving 18F-FDG PET or PET/CT scans as part of a screening program. Among the
6224 patients with a history of cancer, 2 patients underwent 18F-FDG PET/CT and showed an
unexpected presence of secondary primary tumors of bladder cancer. Of the 22,543 participants without a history of cancer, 8 underwent cystoscopy and histology and were diagnosed with bladder cancer after 18F-FDG PET or PET/CT scanning. Patients were excluded if they were lost to follow-up
before 1 year; this constituted 512 of the 6224 patients with a history of cancer and 987 of the 22,543 participants without a history of cancer. In total, 1499 patients were excluded, whereas 28,767 patients met the study criteria.
Imaging Protocol
Initial 18F-FDG PET images were obtained using either a PET scanner (ECAT EXACT HR1, Siemens,
Knoxville, TN) or an integrated PET/CT scanner (Discovery LS, General Electric Medical Systems, Waukesha, WI). Patients were required to fast for at least 8 hours before the initial PET scanning, and blood-glucose measurements were obtained from all patients before 18F-FDG administration. A cut-off
value of less than 8 mmol/L indicated that the examination could be performed. As many sequential images as necessary were taken to include the entire head, thorax, abdomen, and pelvis in the initial PET or PET/CT scans. For the PET scanner, transmission images were obtained for 2 minutes per bed position to correct for photon attenuation, using a germanium-68 line source. For PET/CT scanning, the PET attenuation correction factors were calculated based on the CT images. CT was performed using a four-detector-row spiral CT scanner (LightSpeed, General Electric Medical Systems, Waukesha, WI). Images were acquired in 5 to 7 bed positions, using the following parameters: 140 kV, 40 mA, 0.8
seconds per CT rotation, a pitch of 6, a table speed of 22.5 mm/s, coverage of 722.5 to 1011.5 mm, and an acquisition time of 31.9 to 37 seconds. CT was performed before emission acquisition. CT data were resized from a 512 3 512 matrix to a 128 3 128 matrix to match the PET data and enable fusion of the images, and CT transmission maps were generated. The full width at half maximum of PET/CT was 4.8 mm. The emission scan occurred 50 to 70 minutes after intravenous administration of 370 MBq (10 mCi) of 18F-FDG. Patients were asked to void, and were then positioned on the scanner table. Emission
images were acquired at 5 minutes per bed position. Image datasets were iteratively reconstructed using the ordered subset expectation maximization method. Images were displayed for visual interpretation in three orthogonal projections and whole-body maximum intensity projection (MIP) images. Delayed images were obtained using PET or an integrated PET/CT scanner. Image acquisition and processing for delayed PET or PET/CT scans were performed in the same manner as the PET or PET/CT scans, except that the range of image acquisition was limited to the region of interest (ROI) identified in the initial scan. A workstation (Xeleris, Elegems, Haifa, Israel) was used for image display and analysis.
Image Interpretation
Images were interpreted by 2 experienced nuclear medicine physicians, by consensus. When a consensus could not be reached, the opinion of the senior nuclear medicine physician was accepted. They were aware of the patients’ clinical history, which was provided by the referring physician, but they were unaware of the results of other imaging examinations, if these had been performed. First, the initial attenuation corrected PET images were reviewed using MIP images of the transaxial, coronal, and sagittal planes. Visual analysis was used and foci with abnormal 18F-FDG uptake were recorded. In
uremia patients without urine or oliguria, the intensity of the 18F-FDG uptake in the bladder region was
compared with the intensity of the liver uptake, and graded subjectively on a 4-point scale: 11 indicated that the bladder 18F-FDG uptake was less than the liver uptake; 21 indicated that the bladder 18F-FDG
uptake was as intense as the liver uptake; 31 indicated a moderately intense bladder 18F-FDG uptake,
which was slightly higher than the liver uptake; and 41 indicated an intense bladder 18F-FDG uptake,
which was markedly higher than the liver uptake. In patients with a high level of urinary radioactivity, PET with an IULIT was performed, and the abnormal foci were graded as having less than, equal to, or higher activity than the urinary 18F-FDG activity. Hypermetabolic foci having higher or lower 18F-FDG
intensity than the urinary level of 18F-FDG accumulation in the bladder were considered to be
abnormal foci. CT images showing soft-tissue lesions, which were isometabolic relative to the urinary activity, required additional analyses using delayed PET/CT scanning following the administration of diuretics and oral hydration. The uptake within lesions was quantified by determining the mean activity within a circular ROI of a minimum of 3 3 3 pixels placed within the area of maximal activity. The standard uptake value (SUV) was calculated as follows:
Data and Statistical Analyses
Data were expressed as the mean the standard deviation or number (percentage) of patients in each group. Data analysis was based on data generated from the initial 18F-FDG PET and delayed PET/CT
interpretations. Cystoscopic findings and pathologic results served as the reference standard for all lesions. A clinical follow-up period of at least 1 year served as the reference standard for patients not undergoing cystoscopy or biopsy. The clinical follow-up period ranged from 1 to 9.9 years, with a mean of 4.7 years. Patterns of abnormal bladder foci in the initial PET images were analyzed based on the bladder distribution and reference standard. Abnormal bladder foci on the initial PET or PET/CT and delayed PET/ CT images were also categorized into grades. The accuracy of PET and PET/CT image interpretation was assessed using the reference standard. Considering negative results to be physiological and equivocal results to be physiological and benign, we classified the initial PET or PET/CT and delayed PET/CT findings as true positive, true negative, false positive, or false negative. Because only abnormal 18F-FDG foci of the bladder were analyzed, and patients without cystoscopic
evaluation did not undergo an adequate follow-up, sensitivity and specificity were not calculated; instead, accuracy was presented for the proportion of correctly classified foci on a per-patient basis. A
P value of 0.05 was considered to indicate statistical significance.
Results
In 10 of the 28,767 patients, areas of abnormal tracer uptake (urothelial carcinoma) in the bladder were noted during PET or PET/CT examination (Table I). Two patients (Patient Nos. 2 and 12) showed negative results at the initial diagnosis, but proved to have urothelial carcinoma of the bladder after follow-up and histological interpretation. Based on the background of urinary accumulation of 18
F-FDG activity, patients were divided into groups of uremia patients and high-background urinary radioactivity patients. An IULIT for PET was applied to patients with a high level of urinary radioactivity, in which the patterns and grade of bladder foci in the high-activity bladder were divided into focal increased activity, central decreased activity, peripheral decreased activity, and isometabolic activity. Only the initial images of 2 of the 12 patients (Nos. 1 and 3) were recorded using PET scans. The delayed view of Patient No. 1 was recorded using the same PET, and the delayed view of Patient No. 3 was recorded using PET/CT scan. The bladder foci of Patient Nos. 1 and 2 showed relatively increased 18F-FDG activity in the urinary tract after increasing the upper limit of the image threshold.
For one patient, a PET scan was used to design the initial image status in the upper limit of the image threshold. A lesion with relatively focal intense activity (SUV 5 5.4) was located easily in the bladder (SUV 5 2.9). For another patient, who had a chronic cough and intermittent hematuria, a PET/CT scan was used to design the initial image status in the medium-upper threshold, which caused masking of the lesion by urinary radioactivity, resulting in a false negative identification of the bladder lesion (Figure 1A). One month later, the patient was diagnosed with urothelial carcinoma of the bladder. When images were reviewed retrospectively, focal-intense 18F-FDG activity in the anterior portion of the
bladder (Figure 1B and 1D) was detected using an IULIT, which corresponded to the soft-tissue lesion observed on the CT image (Figure 1C).
After using the IULIT, the bladder foci showed decreased 18F-FDG activity relative to the urinary
activity in the central (Patient Nos. 3 and 4) and peripheral (Patient Nos. 5 and 6) regions of the bladder (Figures 2 and 3). Hybrid PET/CT scans confirmed the presence of the soft-tissue lesions observed in the CT images. When using a PET image without a combined CT image, a false negative may be diagnosed on a lesion in the peripheral region of the bladder. When a suspicious bladder lesion having an 18F-FDG uptake lower than the urinary activity was observed, the delayed view after diuretic and
oral hydration were administered may have caused a reverse contrast of the lesion toward the urinary background ratio (such as in Patient No. 4). Patient No. 3 received only oral hydration without diuretic administration before the delayed images were captured.
Patients occasionally showed a soft-tissue lesion of the bladder, detected on CT images with an 18
F-FDG activity equal to the urinary activity (Patient Nos. 7 and 8 were asymptomatic). The delayed view after a diuretic with 20 mg of furosemide was injected intravenously and oral hydration with 800 mL of water was administered demonstrated the presence of a lesion with high 18F-FDG activity (Figure 4). A
change in position from supine to prone may exclude the possibility of a false 18F-FDG deposit in the
posterior portion of the bladder. Positive findings were obtained in 2 patients, with a more intense accumulation of FDG in the inferior-posterior portion of the bladder compared with the upper portion of the bladder. Cystoscope examinations yielded negative results.
Uremia patients with hematuria (Patient Nos. 9 and 11) and malignancy changes in urine cytology (Patient No. 10) underwent an 18F-FDG PET/CT examination. Urinary 18F-FDG activity was not
observed in the initial images of uremia patients (Patient Nos. 9, 10, and 11), but the 18F-FDG uptake in
the genitourinary tract may suggest abnormal findings (Figure 5). CT images that confirm these findings require further examination. Bladder cancer was not detected in the CT images of 2 of the 3 uremia patients. Patient No. 10 had oliguria and presented little urinary activity in the delayed image. Patient No. 12 was an asymptomatic man who underwent 18F-FDG PET/CT for cancer screening. Ten
months later, bladder urothelial carcinoma was diagnosed, and the patient received TUR-BT and chemotherapy. This was an 18F-FDG PET/CT false negative case. When applying an IULIT for 18
F-FDG PET or PET/CT and a delayed PET/CT combined with diuretics for detecting bladder cancer, the sensitivity and specificity were 91.17% (95% confidence interval (CI): 59.8% to 99.6%) and 99.99% (95% CI: 99.97% to 100%), respectively. The positive predictive value and negative predictive value were 84.6% (95% CI: 53.66% to 97.29%) and 99.99% (95% CI: 99.98% to 100%), respectively.
Discussion
eliminated in urine after 60 minutes, and 50% is eliminated after 135 minutes (7, 8). We performed PET/CT scans on 30 asymptomatic health checkup examinees (20 men and 10 women) 1 hour after the patients received an infusion of 18F-FDG and 500 mL of normal saline intravenously. The mean SUV
levels of the liver and bladder of these examinees were 2.01 0.27 (1.6 to 2.5) and 15.75 6.33 (8 to 33), respectively. The wide range and high mean of the bladder SUV level were affected by the 18
F-FDG injection dose, dehydration, hydration, and renal function of the patients. Some authors believe that a high level of radioactivity in the bladder may complicate lesion assessments. 18F-FDG PET/CT
has excellent sensitivity and specificity for metastatic bladder cancer detection, and provides additional diagnostic information that enhances clinical management compared with CT/MRI alone (9).
Patients with end-stage renal disease present oliguria or anuria, which causes low bladder 18F-FDG
activity. Lesions of the bladder with increased 18F-FDG uptake can be detected easily because of a
marked contrast with the peripheral region. Most uremic patients show rare excretory urinary radioactivity; therefore, when focal 18F-FDG activity in the bladder is observed, further examination is
necessary for the diagnosis of a highly possible bladder malignancy (10).
Early investigations of bladder neoplasm imaging using 18F-FDG PET were disappointing. Image
reconstruction with filtered back projection produced steak artifacts around the bladder. Iterative reconstruction using the ordered subset expectation maximization method provided a high image quality. Furthermore, for retrograde irrigation of the bladder to remove 18F-FDG radioactivity, saline
irrigate and a Foley catheter were used during 18F-FDG PET data acquisition (4). This technique has
the benefit of allowing the evaluation of tumors of the bladder; however, it is an invasive proce dure. Thereafter, PET combined with CT or hybrid PET/CT revealed the structure of the lesions, as well as anatomical information, providing additional diagnostic and prognostic accuracy (11). CT scans were used for correcting the attenuation of PET images. Patients were asked to void, and were positioned on the scanner table, complicating urinary bladder lesion assessments using CT images only. Another method involved recording delayed 18F-FDG PET/CT images after administering a diuretic coupled
with parenteral hydration (6, 12). This technique was noninvasive and could improve the detection of locally recurrent or residual bladder tumors; however, furosemide must be injected at least 2 hours after radiotracer injection, and many delayed images must be taken (6). In addition, early dynamic FDG PET images can demonstrate the presence of bladder lesions that are obscured by urine activity on routine images taken 1 hour after a radiotracer is injected (13).
The upper limit of the image threshold in the projection view was presented in counts per pixel, according to the intravenous dose of 18F-FDG, body weight, acquisition time, and reconstruction
method. In addition, the upper limit of the image threshold in volume view (transaxial, coronal, and sagittal) can be presented in counts per pixel or gigameters per milliliter units. In lesions with high radioactivity backgrounds, such as those in the bladder and brain, using a high upper limit of the image threshold may allow differentiation of high-activity lesions from the highly active background region.
If the upper limit of the image threshold is set too low, the difference between the lesion and the background may not be visualized, because the lesion intensity is above the threshold. In the present study, 18F-FDG PET with an IULIT allowed the visualization of the higher or lower contrasts of lesions
relative to urinary radioactivity. Delayed PET/CT images taken after administering a diuretic coupled with parenteral hydration were used to identify small existing isometabolisms or suspicious lesions of the bladder. Furthermore, hybrid PET/CT, which enabled detecting abnormal metabolism, could indicate lesion size, shape, and location. Delayed imaging and intravenous administration of furosemide were used to identify soft-tissue lesions with a high level of SUV. Papillary tumors represent the majority of urothelial carcinomas, which grow narrow, finger-like projections (14). In addition to papillary tumors, bladder cancer can develop in the form of a flat, red (erythematous) patch on the mucosal surface, requiring additional delayed PET/CT images to be recorded after administering a diuretic coupled with parenteral hydration and/or cystoscopy.
With the help of an IULIT, 18F-FDG PET enabled the visualization of focal FDG accumulation in the
dependent area of the bladder, possibly because of 18F-FDG deposition. Physiological changes usually
cause symmetrical focal 18F-FDG accumulation in the posterior-inferior area of the bladder. Requesting
patients to change position from supine to prone may resolve the pitfalls of image interpretation (15). Focal 18F-FDG retention in the extravesical area of the bladder may be due to the diverticulum of the
bladder (16). CT images or delayed PET/CT after patients received hydration and had a full bladder may aid in the diagnosis. When focal-intense 18F-FDG deposits occur in the dependent area of the
blad-der diverticulum, a change in the position of the patient from supine to prone is necessary. If focal 18
F-FDG is deposited in the dependent area, rather than in the bladder lesion, a position change causes a focal 18F-FDG move and persistent deposits in the dependent area (Figure 6).
This study had a number of limitations. Several comparisons of the various permutations and combinations were not considered, including a single delayed scan, a delayed scan after diuretic and oral hydration, only delayed scans, and dual-phase scans, all with diuretic administration. These com-parisons should be included in future studies. An additional limitation was the inherent possibility of missing potentially non-18F-FDG avid, small-sized, and flat-bladder lesions.
Using an IULIT for 18F-FDG PET clearly improved the visualization of the contrast between lesion
activity and urinary activity. Therefore, applying an IULIT may reduce false negatives and increase accuracy when detecting bladder cancer. Delayed images captured after administering diuretics and oral hydration to the patients reduced the number of equivocal (isometabolism) and false positive findings in patients with a dependent site of 18F-FDG deposition in the bladder. In initial 18F-FDG PET
or PET/CT images, an IULIT could be applied to screen lesions, regardless of whether they are located in the bladder. Delayed images captured after administering diuretics and oral hydration may improve the confidence of decisions made regarding cystoscopy. Therefore, we suggest using an initial image captured with an IULIT for evaluating bladder 18F-FDG uptake.
