Incidental Finding of Focal FDG Uptake in the Bowel During PET/ CT: CT Features and Correlation With Histopathologic Results

Transcription

1 Gastrointestinal Imaging Original Research Kei et al. Bowel Uptake at PET/CT Gastrointestinal Imaging Original Research Pin Lin Kei 1,2 Raghunandan Vikram 3 Henry W. D. Yeung 4 John R. Stroehlein 5 Homer A. Macapinlac 1 Kei PL, Vikram R, Yeung HWD, Stroehlein JR, Macapinlac HA Keywords: FDG, incidental finding, focal bowel uptake, PET/CT DOI: /AJR Received September 29, 2009; accepted without revision November 2, Department of Nuclear Medicine, The University of Texas M. D. Anderson Cancer Center, Houston, TX. 2 Department of Diagnostic Radiology, Singapore General Hospital, Outram Rd., Singapore Address correspondence to P. L. Kei (Kei.pin.lin@sgh.com.sg). 3 Department of Diagnostic Radiology, The University of Texas M. D. Anderson Cancer Center, Houston, TX. 4 Department of Nuclear Medicine and PET, Hong Kong Sanatorium and Hospital, Happy Valley, Hong Kong. 5 Department of Gastroenterology, Hepatology and Nutrition, The University of Texas M. D. Anderson Cancer Center, Houston, TX. WEB This is a Web exclusive article. AJR 2010; 194:W401 W X/10/1945 W401 American Roentgen Ray Society Incidental Finding of Focal FDG Uptake in the Bowel During PET/ CT: CT Features and Correlation With Histopathologic Results OBJECTIVE. The purpose of this study was to identify and characterize the clinically significant lesions associated with incidental detection of focal uptake of 18 F-FDG in the bowel at PET/CT. MATERIALS AND METHODS. Among 2,250 consecutively registered patients with various nongastrointestinal malignant diseases who underwent FDG PET/CT as part of their care, patients with the incidental finding of focal bowel uptake of FDG were included in the study. All patients underwent an endoscopic or surgical procedure for characterization of the lesions. The location, intensity of uptake, and appearance of the lesions on PET/CT images were recorded and compared with the endoscopic and surgical pathologic results. RESULTS. Twenty-one of 25 foci of intense uptake in the bowel were associated with endoscopic or surgical abnormalities (positive predictive value, 84%). Seven lesions were malignant (two primary, five secondary); 13 were premalignant (nine tubovillous adenoma, four tubular adenoma); and one lesion was benign (hyperplastic polyp). Eleven lesions detected with endoscopy were not FDG avid, and all 11 were smaller than 1 cm in diameter. There was no statistically significant difference in the maximum standardized uptake values of the benign and malignant lesions. CONCLUSION. The incidental finding of focal FDG uptake in the bowel justifies further investigation of these foci and should not be dismissed as physiologic uptake. Premalignant lesions, such as adenoma, are often found, and early treatment may prevent the development of. I ntense focal uptake of 18F-FDG in the bowel is found in approximately 1.3 3% of patients who undergo PET/CT [1 3]. The mechanism of increased FDG uptake in the gastrointestinal tract is unclear but may be the result of a physiologic, inflammatory, benign, or malignant process. Increased FDG uptake also can be related to physiologic activity in the bowel, such as that due to metabolically active smooth muscle, reactive lymphocytes in the terminal ileum and cecum, and swallowed secretions [2, 4, 5]. The most common sites of physiologic uptake are the distal esophagus, stomach, and colon [4, 6]. These foci of uptake due to benign conditions or physiologic activity can be misinterpreted as malignant lesions and vice versa. Although attempts have been made to use the pattern of FDG uptake to characterize lesions [6, 7], the appearance of such lesions on the CT scans obtained during PET/CT has not been emphasized, to our knowledge. In this study, we sought to determine the frequency, imaging appearance on combined PET/CT scans, and clinical significance of these incidentally found focal areas of FDG uptake. Materials and Methods Study Design and Patient Data We received approval from the institutional review board of the institution at which the patients were treated to perform this retrospective study, and the need for informed consent was waived. The study complied with HIPAA requirements. We retrospectively reviewed the electronic records of 2,250 consecutively registered patients older than 18 years who had undergone FDG PET/ CT for malignant disease during the 11-month period August 2006 through June We identified the cases of patients who had incidental focal uptake of FDG in the gastrointestinal tract by reviewing the reports. Any intense focal uptake found to be greater than that of background uptake in the liver was considered abnormal. Patients with a history of gastrointestinal malignant AJR:194, May 2010 W401

2 Kei et al. TABLE 1: Characteristics of Lesions Associated With Incidental Findings of Focal FDG Uptake in the Bowel Patient No. Primary Malignant Tumor disease and patients without follow-up information were excluded from the study. Patients with a history of any pathologic condition or malignant disease involving the sites of localized intense FDG uptake before PET/CT also were excluded from the study. Specimens from 22 of the 36 patients who fulfilled the initial criteria were subjected to histopathologic investigation after a follow-up procedure (endoscopy or surgery), and only these 22 patients (10 men, 12 women; mean age, 62 years; range, years) were included for the purposes of this study. The other 14 patients either did not return for follow-up or died before the follow-up investigation was performed. The average time between PET/CT True-Positive Diagnosis Final Diagnosis and the follow-up procedure was 1.3 months (range, 1 week 4 months). The primary malignant tumors in the 22 patients are tabulated in Table 1. FDG PET/CT The patients fasted for at least 6 hours before PET/CT. Blood glucose was measured 1 hour before injection of FDG. If the blood glucose level exceeded 150 mg/dl (8.3 mmol/l), the examination was deferred. Approximately 1 hour before imaging, patients received an injection of radioactively labeled FDG (mean, 15 mci [555 MBq]; range, 9 19 mci [ MBq]). All scans were obtained with an integrated PET/CT scanner (Discovery ST-8, GE Healthcare). False-Positive Diagnosis False-Negative Diagnosis Size Final (mm) SUV max diagnosis SUV max Final Diagnosis Size (mm) 1 Myxoid liposarcoma Colon Hyperplastic polyp 2 2 Lymphoma Colon Ovary Ovarian metastasis Ovary Ovarian metastasis Melanoma Melanoma metastasis Melanoma Melanoma metastasis Lymphoma Lymphoma metastasis Non small cell lung 9 Waldenstrom macroglobulinemia 10 Non small cell lung 11 Non small cell lung Tubulovillous adenoma with high grade dysplasia Tubulovillous adenoma with high grade dysplasia Tubular adenoma with high grade dysplasia Tubular adenoma with high grade dysplasia Physiologic activity Melanoma, lymphoma Tubulovillous adenoma Non small cell lung 14 Base of tongue, non small cell lung 15 Chronic lymphocytic leukemia, lymphoma 16 Non small cell lung Tubulovillous adenoma Tubular adenoma 4 2, 2, 3, 7 Tubulovillous adenoma Tubulovillous adenoma Tubulovillous adenoma Tubular adenoma 2; tubulovillous adenoma 17 Lymphoma Tubular adenoma Breast Tubular adenoma Tubular adenoma 3 19 Lymphoma Tubular adenoma Physiologic activity 8.0 7, 7; Tubular adenoma 3 20 Tonsil Tubular adenoma Tubular adenoma 3 21 Melanoma Hyperplastic polyp Sarcomatoid cancer of the thorax Note SUV max = maximum standardized uptake value. Physiologic activity 10.5, 10 Unenhanced CT scans from the base of the skull to the upper thigh were obtained for attenuation correction and to aid in diagnosis (300 ma; tube rotation time, 0.5 second; 120 kvp; table speed, 13.5 mm/rotation; beam collimation, mm). Axial CT images were reconstructed with a soft-tissue reconstruction kernel with a slice thickness of 3.3 mm and an interval of 3.3 mm to match the PET images. No oral or rectal contrast material was administered for any of the scans. PET scans were acquired in the 2D mode for 3 minutes per bed position, and the images were reconstructed with standard vendor-provided reconstruction algorithms incorporating ordered subset expectation maximization. Attenuation correction W402 AJR:194, May 2010

3 Bowel Uptake at PET/CT of the PET images was performed with attenuation data from the CT component of the examination. The manufacturer s software was used to correct emission data for scatter, random events, and dead-time losses. Imaging Data Analysis and Correlation With Endoscopy Two readers with experience in gastrointestinal imaging reviewed the existing CT and PET components of the study on workstations (Advantage Windows 4.2, GE Healthcare). Both readers analyzed all abnormal findings, and any discrepancies were reviewed and resolved by consensus. The readers analyzed the PET and fused images in both the axial and coronal planes without knowledge of the endoscopic or surgical pathologic results. Any focal bowel metabolic activity greater than that seen in the normal hepatic parenchyma was considered intense focal bowel uptake. We used the CT component of the study to assess any associated soft-tissue abnormality in the region of intense focal bowel uptake. The attenuation-corrected PET component of the study was used to measure the maximum standardized uptake value (SUV max ) over the appropriate region of interest. Contiguous diffuse FDG uptake in the gastrointestinal tract was not considered in this study because it is usually physiologic [7]. The gastrointestinal tract was divided into the stomach, small bowel, cecum, ascending colon, hepatic flexure of the colon, transverse colon, splenic flexure of the colon, descending colon, sigmoid colon, and rectum to allow direct comparison with the endoscopic findings. In our study, the histopathologic diagnoses were grouped as malignant lesions, which included primary and metastatic disease of the gastrointestinal tract; premalignant lesions, which included adenoma with varying degrees of dysplasia; and benign lesions, such as hyperplastic polyps. An area of increased focal bowel uptake without the finding of a corresponding mucosal or structural lesion at endoscopy was deemed physiologic uptake. Follow-up records were scrutinized to exclude false-negative results of endoscopy. A PET/CT result was considered true-positive when an FDG focus and an endoscopic or surgical abnormality were situated in the same location. True-positive results included malignant, premalignant, and benign lesions. A false-positive PET/ CT result was defined as focal FDG uptake without a matching endoscopic or surgical abnormality, and these findings were considered to represent physiologic uptake. When a mucosal lesion of any size was detected at endoscopy with no corresponding focal FDG uptake, the PET/CT result was interpreted as false-negative. Per-patient and per-lesion analyses were performed to evaluate the positive predictive value of FDG PET/CT in the detection of gastrointestinal tract abnormalities. We evaluated the utility of PET/CT in the detection of structural lesions smaller than 10 mm in diameter, which corresponds to the spatial resolution of the PET scanner. When adenoma was found, the histopathologic result in regard to the macrostructure (tubular or villous component) and the degree of dysplasia were recorded. The SUV max (mean ± SD) was correlated with the histopathologic findings. We recorded the CT findings in the region of FDG uptake taking into account misregistration due to any physiologic motion that might have occurred in the time between the CT and PET data acquisitions. We were specifically looking for associated wall thickening, polypoid lesions, feceslike material (soft-tissue with small areas of air lucency), and whether the lesion was in a gravity-dependent or nondependent position within the bowel. When the area of increased focal uptake corresponded to a collapsed segment of the bowel, this finding was also recorded. The final determination of incidental FDG uptake was based on correlative endoscopic or surgical findings. Colonoscopy was performed on 19 patients, and two additional patients underwent upper gastrointestinal endoscopy (i.e., gastroscopy). All lesions detected at endoscopy were excised or biopsied, and their locations were recorded. One patient underwent surgery for intestinal obstruction secondary to a small-bowel lesion that correlated with the abnormal FDG PET/ CT finding. Statistical Analysis We performed statistical data analysis using the nonparametric Wilcoxon s rank sum test to compare the SUVs of the different types of lesions. Owing to the small numbers in the analysis, to boost the statistical strength of the study we combined the data from the benign and physiologic uptake groups. We also performed Wilcoxon s rank sum test to compare the SUVs within the premalignant group of lesions to test whether there was a difference between premalignant lesions with high-grade dysplasia and premalignant lesions without high-grade dysplasia. Results Twenty-five foci of increased FDG uptake were found in the 22 patients who underwent follow-up procedures. Three of the 22 patients had two separate (at least 5 cm apart) foci of increased FDG uptake in the bowel, and PET/CT showed all of these foci were localized to the colon and rectum. PET/CT showed the 19 solitary foci were localized to the stomach (n = 2), small bowel (n = 1), and colon and rectum (n = 16). No esophageal lesions were detected in this study. In addition to the lesions detected with PET/CT, 11 other mucosal or structural lesions were detected TABLE 2: Locations of Focal FDG Uptake in the Gastrointestinal Tract and Standardized Uptake Values Site of FDG Uptake Stomach 2 Small bowel 1 True-Positive Malignant Premalignant Benign Physiologic (False-Positive) Non-FDG-Avid (False-Negative) Colon Cecum Ascending colon 1 1 Hepatic flexure 1 0 Transverse colon 1 3 Splenic flexure 1 0 Descending colon 1 Sigmoid colon Rectum Total SUV max Mean ± SD 15.8 ± ± ± 1.1 Range AJR:194, May 2010 W403

4 Kei et al. TABLE 3: CT Features of the Incidentally Detected Focus of FDG Uptake in Gastrointestinal Tract Feature Soft-tissue lesion larger than 1 cm, no internal areas of air lucency at endoscopy that were not FDG avid. All 11 lesions were smaller than 1 cm in diameter (range, 2 8 mm). The locations of the lesions and SUV max are tabulated in Table 2. Per-Patient Analysis Among the 22 patients who underwent follow-up endoscopy or surgery, 21 (95%) had mucosal or structural lesions detected at endoscopy or surgery that corresponded to the lesions seen on the PET/CT scans. Seven of these patients had malignant lesions, 13 had premalignant lesions, and one patient had an FDG-avid hyperplastic polyp, which is considered benign. One patient had two separate foci of intense uptake (5 cm apart) in the colon, but no structural lesions were detected at colonoscopy. This patient underwent follow-up for 12 months, and the focal area of intense FDG uptake had resolved on subsequent PET/CT scans; the abnormalities therefore were considered due to physiologic uptake. Per-Lesion Analysis A total of 32 mucosal lesions were found in 22 patients. These lesions measured cm. Of these 32 mucosal lesions, 21 were FDG avid. Four foci of intense bowel uptake detected with PET were not associated with FDG-Avid Malignant Lesion (n = 7) FDG-Avid Premalignant Lesion (n = 13) any mucosal abnormality. The positive predictive value of PET/CT in the detection of these lesions was 84% (21/25). Seven of the 21 true-positive lesions were malignant: two newly diagnosed (second primary) s of the colon, one lymphoma, two ovarian metastatic lesions, and two mucosal or submucosal metastatic lesions of melanoma. Thirteen lesions were considered premalignant: four tubulovillous adenomas with high-grade dysplasia, five tubulovillous adenomas with no dysplasia, and four tubular adenomas with no dysplasia. One lesion found in the sigmoid colon was a hyperplastic polyp and therefore was considered benign. Four focal areas of increased uptake of FDG in the bowel had no corresponding mucosal lesions at endoscopy and were followed for 12 months. All four of these foci of increased FDG uptake had resolved on subsequent PET/CT scans and were deemed physiologic uptake. In addition, 11 non-fdg-avid mucosal lesions were detected with endoscopy. All of these lesions were smaller than 10 mm; one was a tubulovillous adenoma with no dysplasia, nine were tubular adenomas with no dysplasia, and one was a benign hyperplastic polyp. The SUV max of the 25 foci of intense uptake in the gastrointestinal tract ranged from FDG-Avid Benign Lesion (n = 1) False-Positive or Physiologic Uptake (n = 4) Bowel wall thickening Feces-like matter Collapsed segment of bowel A Fig year-old woman with non small cell of lung. A, PET/CT scan shows unexpected focus of intense uptake (arrow) in rectum. Colonoscopy and histopathologic examination showed large tubulovillous adenoma. B, Unenhanced soft-tissue-window CT image corresponding to A shows broad-based polypoid lesion (arrow) arising from rectal wall in non-gravity-dependent position. B 5.0 to The mean SUV max was 15.8 ± 7.5 (range, ) for the seven malignant lesions, 20.7 ± 11.3 (range, ) for the 13 premalignant lesions, 12 for the solitary benign (hyperplastic polyp) lesion, and 10.2 ± 1.1 (range, ) for the four sites of physiologic activity. We did not observe a statistical significant difference in SUV max between the malignant and benign physiologic groups of lesions (p = 0.17). There also was no significant difference in SUV max between the premalignant lesions with high-grade dysplasia and the premalignant lesions with no highgrade dysplasia (p = 0.76). There was evidence that SUV max was higher in the premalignant group than in the benign physiologic group (p = 0.003), but this finding was likely caused by random statistical variability due to the small sample size and unlikely to be of clinical significance. In the CT component of PET/CT, we found associated soft-tissue abnormalities, such as bowel wall thickening, in five of the seven malignant lesions and eight of the 13 premalignant lesions (Table 3). Feces-like matter with internal areas of air lucency was found in three of the four sites of physiologic uptake (Table 3). Six foci of intense FDG uptake were found in nondependent positions of the bowel. At endoscopy all of these findings were confirmed to be true-positive lesions (one lymphoma, one ovarian metastatic lesion, one melanoma metastatic lesion, two tubular adenomas, and one tubulovillous adenoma) (Figs. 1 and 2). Discussion We found that 95% of the patients in our study group with incidental focal uptake of FDG had an abnormal finding at endoscopy or surgery. PET/CT scans depicted 84% (perlesion analysis) of the total number of mucosal lesions found. These results are comparable to those of other studies [1 3]. Several studies with large sample sizes have shown that FDG PET/CT is a useful tool for detecting premalignant and malignant lesions W404 AJR:194, May 2010

5 Bowel Uptake at PET/CT [3, 7]. Gutman et al. [3] reviewed a series of 1,716 PET/CT scans of patients with various malignant diseases other than colorectal cancer and found focal FDG uptake in the colons of 20 patients. The foci corresponded to three s and 10 adenomas found at colonoscopy. Per-patient analysis in that study showed that 75% of the increased focal FDG uptake was associated with a structural lesion detectable with colonoscopy. The positive predictive value for FDG PET/CT in localization of the abnormalities was 67%. In another study, Israel et al. [2] found 58 foci of intense FDG uptake in the gastrointestinal tract in a series of PET/CT scans of 4,390 patients. Among the 34 patients who underwent follow-up studies, 24 (71%) had gastrointestinal tract abnormalities, including malignant and premalignant processes. In yet another study, Kamel et al. [1] analyzed the FDG PET/CT scans of 3,281 patients and found that PET/CT had a positive predictive value of 87% in the detection of gastrointestinal tract abnormalities. Those authors concluded that the PET/CT results directly influenced subsequent treatment of 28% of the patients [1]. We did not find the semiquantitative measurement SUV max useful in predicting the presence of pathologic changes, even though some authors [8, 9] suggest this value is useful. Studies [2, 3] have since shown that this method is not discriminative enough to separate malignant, premalignant, and benign lesions. In our series, a wide range of SUVs were found with no clear statistical difference between in the SUVs of the malignant and benign physiologic groups of lesions. There also was no statistical difference between the SUVs of premalignant lesions with high-grade dysplasia and premalignant lesions without high-grade dysplasia. A Fig year-old woman with lymphoma. A, FDG PET/CT shows unexpected focus of intense uptake (arrow) in non-gravity-dependent aspect of body of stomach. B, Unenhanced soft-tissue-window CT image corresponding to A shows nodular soft tissue (arrow) that was later confirmed to be lymphomatous infiltration of stomach. All 11 lesions not found with PET were smaller than 10 mm in diameter. This result is not surprising, particularly when we consider that the highest resolution of most commercially available PET scanners is greater than 1 cm [10]. Yasuda et al. [11] found that the ability to detect colonic adenomas improved with increasing lesion size. Gollub et al. [12] found that use of a standalone PET scanner does not significantly increase sensitivity in the detection of mucosal lesions or polyps over use of conventional CT colonography, particularly for flat lesions and lesions smaller than 1 cm. In our study, all true-positive lesions measured at least 10 mm. Long-segment diffuse uptake in the gastrointestinal tract is generally associated with physiologic uptake [6, 7], and intense focal uptake in the bowel is more often associated with true lesions [7, 13]. Previous studies have been focused mainly on the FDG uptake pattern in an attempt to characterized lesions [6, 7], but the CT findings from PET/CT have not been emphasized, to our knowledge. The unique advantage of PET/ CT fusion imaging is the ability to correlate findings with two complementary imaging techniques. The CT component of PET/CT allows precise anatomic correlation and can even be used for diagnostic purposes. In 12 of the 21 patients with true-positive lesions, the CT component of the study showed a soft-tissue abnormality. Even in unprepared bowel, we found an associated soft-tissue lesion (larger than 1 cm) or bowel wall thickening in five of the seven FDG-avid malignant lesions, the one benign lesion, and eight of 13 premalignant lesions. In addition, if the lesion was found in the non-gravity-dependent aspect of the bowel at PET/CT, the likelihood of finding a mucosal lesion B at endoscopy was high (six of six lesions in the nondependent position of the bowel were true-positive findings). We believe this result occurs because food or fecal material tends to fall to the dependent portion of the bowel, whereas a structural lesion arising from the bowel mucosa in a nondependent location stays more or less in the same position [14]. This principle is used in CT colonography (prone and supine image acquisition), and the basic rule can be applied to PET/CT scan interpretation. This principle implies that whenever an FDG-avid lesion detectable on a PET/CT scan is found in a nondependent aspect of the bowel, the probability of finding a mucosal lesion at endoscopy is high. Potential applications of this principle would be PET/colonography [12, 15] and problem solving in the care of patients who decline or are unable to undergo an endoscopic examination. Even in the absence of a CT abnormality, or when the affected segment of bowel is collapsed, incidental focal bowel uptake should not be taken lightly. A substantial number of these patients may have malignant or premalignant lesions [1 3]. On the basis of our results, we recommend that foci of increased FDG uptake in the bowel be evaluated further, particularly if an associated area of soft-tissue attenuation is found during the CT component of PET/CT or when an FDGavid focus is found in the nondependent part of the bowel. Acknowledgments We thank Ping Liu and Kate Newberry for their assistance in the preparation of this manuscript. References 1. Kamel EM, Thumshirn M, Truninger K, et al. Significance of incidental 18 F-FDG accumulations in the gastrointestinal tract in PET/CT: correlation with endoscopic and histopathologic results. J Nucl Med 2004; 45: Israel O, Yefremov N, Bar-Shalom R, et al. PET/ CT detection of unexpected gastrointestinal foci of 18 F-FDG uptake: incidence, localization patterns, and clinical significance. J Nucl Med 2005; 46: Gutman F, Alberini JL, Wartski M, et al. Incidental colonic focal lesions detected by FDG PET/ CT. AJR 2005; 185: Kostakoglu L, Hardoff R, Mirtcheva R, Goldsmith SJ. PET-CT fusion imaging in differentiating physiologic from pathologic FDG uptake. RadioGraphics 2004; 24: AJR:194, May 2010 W405

6 Kei et al. 5. Cook GJ, Fogelman I, Maisey MN. Normal physi- Imaging 2003; 30: tection of cancer and precancerous lesions. AJR ological and benign pathological variants of 9. Chen YK, Kao CH, Liao AC, Shen YY, Su CT. 2007; 188: fluoro-2-deoxyglucose positron-emission to- Colorectal cancer screening in asymptomatic 13. Kresnik E, Mikosch P, Gallowitsch HJ, Hein- mography scanning: potential for error in inter- adults: the role of FDG PET scan. Anticancer Res isch M, Lind P. F-18 fluorodeoxyglucose posi- pretation. Semin Nucl Med 1996; 26: ; 23: tron emission tomography in the diagnosis of 6. Prabhakar HB, Sahani DV, Fischman AJ, Mueller 10. Mawlawi O, Podoloff DA, Kohlmyer S, et al. Per- inflammatory bowel disease. Clin Nucl Med PR, Blake MA. Bowel hot spots at PET-CT. RadioGraphics 2007; 27: Tatlidil R, Jadvar H, Bading JR, Conti PS. Incidental colonic fluorodeoxyglucose uptake: correlation with colonoscopic and histopathologic findings. Radiology 2002; 224: Zhuang H, Sam JW, Chacko TK, et al. Rapid normalization of osseous FDG uptake following traumatic or surgical fractures. Eur J Nucl Med Mol formance characteristics of a newly developed PET/CT scanner using NEMA standards in 2D and 3D modes. J Nucl Med 2004; 45: Yasuda S, Fujii H, Nakahara T, et al. 18 F-FDG PET detection of colonic adenomas. J Nucl Med 2001; 42: Gollub MJ, Akhurst T, Markowitz AJ, et al. Combined CT colonography and 18 F-FDG PET of colon polyps: potential technique for selective de- 2001; 26: Gluecker TM, Fletcher JG, Welch TJ, et al. Characterization of lesions missed on interpretation of CT colonography using a 2D search method. AJR 2004; 182: Kinner S, Antoch G, Bockisch A, Veit-Haibach P. Whole-body PET/CT-colonography: a possible new concept for colorectal cancer staging. Abdom Imaging 2007; 32: W406 AJR:194, May 2010