2-(18F)-Fluoro-2-Deoxy-D-Glucose Positron Emission Tomography Detects Clinical Relevant Adenomas of the Colon: A Prospective Study
http://www.100md.com
《临床肿瘤学》
the Departments of Gastroenterology and Hepatology and Nuclear Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, the Netherlands
ABSTRACT
PURPOSE: 2-(18F)-fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET) is a noninvasive imaging technique used clinically to detect malignant tumors. FDG-PET has been established as a tool for diagnosis of recurrent or metastatic colorectal carcinoma. Several case series suggest that FDG-PET also detects larger adenomas. The goal of this study was to investigate whether FDG-PET is able to detect colonic adenomas.
PATIENTS AND METHODS: FDG-PET was performed in 100 consecutive patients in whom colonic adenomas were suspected on barium enema (n = 47) or sigmoidoscopy (n = 53). A positive scan was defined as focal large bowel FDG accumulation. FDG-PET was followed in all cases by colonoscopy, and removed adenomas were examined histopathologically.
RESULTS: Colonoscopy confirmed the presence of adenomas in 68 of 100 patients. In 35 patients, there was focal FDG accumulation at site of the adenoma. The sensitivity of FDG-PET increased with adenoma size (21%, adenomas 1 to 5 mm; 47%, 6 to 10 mm; and 72%, > 11 mm). The sensitivity of FDG-PET also increased with the grade of dysplasia (33%, low grade; 76%, high grade; and 89%, carcinomas). The overall specificity was 84%.
CONCLUSION: FDG-PET detects colonic adenomas and the diagnostic test characteristics improve with size and grade of dysplasia of the adenoma.
INTRODUCTION
Colorectal adenocarcinoma (CRC) is the second most common cause of cancer-related deaths in the Western world and the fourth most frequent malignancy.1,2 The most important indicator of outcome after surgical resection of CRC is the pathologic stage at presentation. This is reflected by a five-year survival of approximately 90% in stage I, 60% to 80% in stage II, 40% to 60% in stage III and 4% in stage IV.3 These data underscore the importance of diagnosing CRC in an early or premalignant stage.
Colon carcinogenesis is regarded as a multifaceted process influenced by a variety of hereditary and lifestyle factors and involves a histologic progression from adenomatous polyps to colonic cancer.4,5 The incidence of adenomas increases with age, and the risk for malignancy with size. For example, malignant transformation is detected in only 1% of adenomas measuring < 1 cm, compared with 10% in larger adenomas.6,7 Indeed, detection and removal of adenomas is thought to result in a decrease in the incidence and mortality from colorectal cancer.8 A broad range of screening modalities for the detection of adenomas are available, such as fecal occult blood testing, endoscopy, barium enemas, and virtual colonoscopy. Although colonoscopy is regarded as the gold standard for the detection of premalignant lesions such as adenomatous polyps, it is not optimal in terms of examination performance, safety and patient acceptance, and this calls for a different diagnostic approach.
Recently, 2-fluoro [18-]-2-deoxyglucose positron emission tomography (FDG-PET) has been introduced in clinical practice for imaging of various malignancies. FDG-PET is a noninvasive, functional imaging technique, which uses the radiopharmaceutical FDG.9 It is clinically used to detect a wide variety of tumors including lymphoma, melanoma, lung cancer, and colon cancer.10 Several studies indicate a potential role of whole-body FDG-PET for the detection of malignant lesions of the colon.11 Indeed, FDG-PET is probably superior to conventional abdominal computed tomography in the assessment of patients with recurrent and metastatic CRC, and is reimbursed under Medicare for the diagnosis, staging and restaging of CRC.12-14
The situation is less defined for premalignant colorectal disease. The results from two retrospective studies suggest that FDG-PET may detect colonic adenomas, but it is uncertain which factors affect the diagnostic ability of FDG-PET.15,16 Furthermore, focal uptake in the colon is frequently identified on whole-body FDG-PET in patients studied for malignancies other than CRC, and the implications of these findings are unclear. These considerations stimulated us to design a prospective cohort study to establish the diagnostic test properties of FDG-PET in the detection of colonic adenomas using colonoscopy and histology as the gold standard.
PATIENTS AND METHODS
Patient Population
Between July 2001 and May 2004, we enrolled 100 consecutive Dutch adult patients who were referred to the endoscopy unit of the Department of Gastroenterology and Hepatology (Radboud University Nijmegen Medical Centre, the Netherlands) because of the suspicion of colonic polyp(s) on sigmoidoscopy or barium enema. The group consisted of 58 males and 42 females, mean age being 62 (SD 11) years. Patients were excluded if they were known with active inflammatory bowel disease or diverticulitis, because those diseases could lead to false-positive FDG-PET. We also excluded patients with known or suspicion of cancer. Patients with diabetes mellitus were excluded if fasting glucose concentration exceeded 15 mmol/L, as high serum glucose levels are responsible for false-negative FDG-PET. The study was approved by the local ethical committee. Informed consent was obtained.
FDG-PET Imaging
A dedicated PET scanner (ECAT-EXACT, Siemens/CTI, Knoxville, TN) was used for data acquisition. The spatial resolution of the PET camera was 5 mm. Before FDG injection, patients fasted for at least 6 hours. Intake of sugar free-liquids was permitted. Immediately before the procedure, the patients were hydrated with 500 mL of water. One hour after intravenous injection of 200 to 220 MBq FDG (Mallinckrodt Medical, Petten, the Netherlands) and 12 mg furosemide, emission and transmission images of the abdomen were acquired (10 minutes per bed position). The images were corrected for attenuation and reconstructed using the ordered-subsets expectation maximization (OSEM) algorithm. The reconstructed images were displayed in coronal, transverse and sagittal planes.
Interpretation
For this study, FDG-PET was analyzed visually by one of the staff physicians from the Department of Nuclear Medicine (Radboud University Nijmegen Medical Centre) and then reviewed by one of us (W.J.G.O), unaware of the results of the original interpretation of the FDG-PET, other diagnostic tests, or the final diagnosis. Results were judged to be abnormal if focal accumulation of the tracer was detected in the large bowel. If focal uptake was demonstrated, the uptake within the lesion was quantified by calculating the maximum and mean standard uptake value (SUV) according to the formula:
Maximum or mean region of interest activity in mCi/mL were derived from the calibrated activity measured in a manually drawn region of interest over the focal abnormalities (converting measured counts to accumulated activity in mCi) divided by the size of this region of interest (converting voxel size to volumes). Diffuse bowel accumulation was interpreted as physiologic. Disagreements were resolved by consensus.
Colonoscopy
Flexible endoscopy procedures were performed after thorough oral lavage with Klean prep. The procedure was performed by a gastroenterology trainee directly supervised by an experienced gastroenterologist (F.M.N.). The endoscopist was unaware of the result from the FDG-PET. The coecum was reached in 97% of all procedures. All lesions appearing as possibly malignant were biopsied and, whenever possible, complete polypectomy was performed. If polypectomy was technically impossible, surgical resection was performed (n = 3). All tissue specimens were immediately fixed in 10% buffered formalin solution before examination using hematoxylin and eosin staining. Each polyp was classified by an experienced pathologist as adenoma or other type polyp. Dysplasia was defined according to Vienna criteria as either low-grade dysplasia, high-grade dysplasia or carcinoma.17 The grade of dysplasia for each separate adenoma was recorded, and we scored the highest grade of dysplasia (low grade-high grade-carcinoma) in the adenoma for the purpose of the study. The results of the colonoscopy were considered negative in case of normal mucosal findings or if polyps were hyperplastic on histopathologic examination. Colonoscopy results were considered positive when histopathology confirmed the presence of either adenomas or carcinomas. The colonoscopic findings were reported in terms of the site and size of adenomas. Adenoma size was determined by endoscopical examination and data were verified by comparison with the data from macroscopical examination by the endoscopist. In case of multiple polyps, we scored the largest adenoma.
Experimental Procedures
After diagnosis of the colonic adenoma by either diagnostic flexible sigmoidoscopy (n = 53) or by double-contrast barium enema (n = 47), patients were put on the waiting list for a therapeutic colonoscopy that included polypectomy. FDG-PET was performed in the time-window between the diagnostic and therapeutic procedure. This study used colonoscopy as the gold standard, and patients were classified on the basis of their colonoscopic findings.
Statistics
Statistical analysis was performed with SAS statistical software, version 8.0. Data were described by means of frequency tables and descriptive statistics. The sensitivity was defined as the proportion patients with colonic polyps who actually had a positive FDG-PET. The specificity was defined as the proportion of patients without colonic polyps with a negative FDG-PET. The positive predictive value (PPV; the probability that the patient had colonic polyps when restricted to those patients who had a positive FDG-PET) was calculated as follows:
The negative predictive value (NPV; the probability that the patient will not have colonic polyps when restricted to all patients who had a negative FDG-PET) was defined as
RESULTS
Detection of Presence of Adenomas
In 100 patients in whom a previous sigmoidoscopy (n = 53) or double contrast barium studies of the colon (n = 47) had shown lesions suggestive for colonic adenomas, colonoscopy detected adenomas in 68 patients, but failed to confirm the presence of an adenoma in the colon in 32 patients. In 18 of 32 patients the preceding examination had been a double contrast barium study, whereas sigmoidoscopy had been the diagnostic tool for 14 of 32 patients. Within the latter group, all adenomas detected were small and had already been completely removed by biopsy during the initial sigmoidoscopy. Histopathologic examination of removed lesions demonstrated hyperplastic polyps (n = 4), adenomatous polyps (n = 5), a lipoma (n = 1), an inflammation polyp (n = 1) and normal mucosa (n = 3). In all of these cases biopsy completely removed the lesion.
FDG-PET
All 100 patients underwent FDG-PET, and 40 patients had significant focal colonic FDG accumulation.
Diagnostic Characteristics of FDG-PET as a Test for Detecting Colonic Adenomas
In 35 patients with a positive FDG-PET, colonoscopy showed a significant mucosal lesion. These lesions included carcinomas (n = 8), and adenomas with high-grade dysplasia (n = 13) and low-grade dysplasia (n = 14). The median size of the adenomas was 15 mm (range, 2 to 50 mm) and most adenomas were at least 15 mm (n = 21). The adenomas were predominantly located in the sigmoid (n = 26). The calculated mean and SD of the maximum SUV of this total group was 5.0 (range, 1.3 to 14.4). In all cases, FDG accumulation corresponded with the adenoma site. In 33 patients, FDG-PET failed to demonstrate colonic hotspots, but colonoscopy did detect significant abnormalities. Twenty-eight patients had adenomas with low-grade dysplasia, whereas four patients had adenomas with high-grade dysplasia. In one patient, colonoscopy detected an 8 mm large tubular adenoma with histopathologic evidence for foci of invasive carcinoma that had been missed by PET. The median size of the PET-negative adenomas was 6 mm (range, 2 to 30 mm), and almost all adenomas were smaller than 11 mm (n = 23).
Table 1 summarizes the FDG-PET results. FDG-PET is positive in 80% of larger adenomas, but negative in the large majority of small adenomas (1 to 6 mm; P < .001). The sensitivity of FDG-PET increases with the diameter of the polyp and rises from 21% for small (1 to 5 mm) polyps to 72% for larger (> 11 mm) polyps. The specificity of FDG-PET was 84%. The sensitivity of FDG-PET was 33% for adenomas with low-grade dysplasia, 76% for adenomas with high-grade dysplasia, and 89% for carcinomas (Table 2). For the detection of polyps, regardless of size or dysplastic grade, FDG-PET has a PPV of 88% and a NPV of 45%, which suggests that FDG-PET has a better accuracy for detecting the presence polyps rather than excluding its absence. When compared with the subgroup of patients without polyps, the detection of large (> 11 mm) adenomas yields a PPV of 82% and a NPV of 75%. Likewise, for the detection of carcinoma, the PPV is 61% while the NPV is 96%. These results in our study suggest that FDG-PET was not likely to miss colonic carcinoma and that an FDG-PET with negative results in most instances excluded the presence of colonic carcinoma.
Specificity
In 27 patients, neither FDG-PET nor colonoscopy demonstrated any mucosal abnormality in the colon. There were five patients in whom PET demonstrated focal FDG accumulation in the large bowel, but in whom colonoscopy failed to reveal adenomas. In a single patient, colonoscopy was compatible with sigmoidal diverticulitis that corresponded with localized FDG uptake at the same site. Four patients had a normal colonoscopy. The localization of false-positive hot spots was sigmoid (n = 2) and ascending colon (n = 2). In this group, the calculated average and standard deviation (SD) of the maximum SUV were 3.5 and 1.8, respectively. The site of the adenoma had no influence on the FDG-PET outcome (P = .46).
DISCUSSION
Our results show that FDG-PET detects premalignant colonic adenomas. The grade of dysplasia and size of the adenoma were clearly important discriminatory factors that predicted detection, while other parameters such as location of the adenoma, sex, or age did not affect the results. We found that FDG-PET has sensitivity of 72% for the detection of larger (> 11 mm) and of 76% for adenomas with high-grade dysplasia. The specificity of FDG-PET was 84%, whereas for the detection of polyps (regardless of size or dysplastic grade) the PPV is 88% and NPV is 45%. Intensity of tracer uptake does not allow discrimination between adenomas (n = 27) and carcinomas (n = 8) as maximum SUVs (average, 5.9; SD, 3.7), respectively (average, 6.5; SD, 3.0) did not differ.18,19
Few retrospective studies suggest that detection of adenomas by FDG-PET is possible.15,16,19-21 In those studies, the adenoma size was the most important variable to predict outcome of FDG-PET, and the detection rate increased significantly with size of the adenoma. There are no studies on the effect of the grade of dysplasia on diagnostic performance of FDG-PET. In contrast to our results, one study suggested that localization matters.16 This study showed a higher detection rate in coecum, ascending and descending colon compared with the sigmoid and transverse colon.16 Significant shortcomings of these earlier studies is their cross-sectional nature, which makes them prone to positive selection. FDG-PET had been performed either for staging malignant disease other than CRC20,21 or had been done as part of a cancer-screening program,16 and the detection of adenomas was merely coincidental. More importantly, colonoscopy had not been performed in all cases.20,21
We detected that a positive double-contrast barium enema in 10 (21%) of 47 patients was followed by a negative colonoscopy. These data are in agreement with the false-positive rate of barium enemas in The National Polyp Study.22 In eight false-positive barium enemas, FDG-PET was true negative.
The PPV of FDG-PET was 88% which suggests that in 88% of the cases, colonic FDG accumulation accurately identifies an adenoma. One limitation of our finding is that the adenoma prevalence in our population (68%) was relatively high. It is possible that in a population with a lower prevalence of adenoma, the PPV of FDG-PET will decrease as the proportion of patients with false-positive tracer uptake may increase. False-positive results are an unwanted aspect of any diagnostic tool, and FDG-PET is no exception. In 5% of our cases FDG-PET generated false-positive results and we were able to identify several important contributing factors. Normal accumulation in the colonic wall mucosa may appear focal and can be misinterpreted as clinically significant (nonphysiological) FDG accumulation. In our study, two patients were classified as false positive, because colonoscopy failed to identify any mucosal abnormality despite focal FDG accumulation. It is possible that colonoscopy missed a polyp, as back-to-back colonoscopy studies demonstrated a missing rate of 15% to 24% for (mostly small) colonic adenomas.23,24 Further, we identified diverticulitis as a potential confounder. In these cases, FDG accumulation is most likely caused by uptake in inflammatory cells.
Our study was designed to investigate the diagnostic ability of FDG-PET in detection of colonic adenomas. We aimed to include patients with lesions indicative of colonic polyps on sigmoidoscopy or barium enema. On histopathologic examination nine polyps were carcinomas. FDG-PET detected all but a single 8-mm large polyp with histopathologic signs of invasive carcinoma. In addition, FDG-PET also missed three other clinical relevant large adenomas with high grade dysplasia. Given the overall NPV of 45% and the relatively high cost of FDG-PET, the procedure does not qualify as a screening modality for colonic adenomas because very small polyps and polyps with a low grade of dysplasia are missed. Conversely, because of the high PPV (88%) of FDG-PET for clinically significant colonic adenomas, we feel that incidental identification of focal FDG accumulation necessitates follow-up by total colonoscopy. The timing of colonoscopy is, of course, determined by the nature and severity of the underlying disease that led to the initial request for FDG-PET.
FDG-PET detects benign colonic adenomas, but the sensitivity of this test depends on the size and grade of dysplasia of the colonic adenoma, with the latter being the most important discriminative factor that predicts detection of the adenoma. Many dysplastic adenomas are missed by FDG-PET (especially when small and with low-grade dysplasia), which is one of the factors limiting its utility as a screenings modality for adenomas.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank M. van Oijen, MD, Department of Gastroenterology and Hepatology, University Medical Centre St Radboud, Nijmegen, the Netherlands, for his assistance with the statistical analysis.
NOTES
Both W.J.G. Oyen and J.P.H. Drenth contributed equally to this article.
J.P.H. Drenth is a recipient of a NWO-VIDI grant.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Greenlee RT, Hill-Harmon MB, Murray T, et al: Cancer statistics, 2001. CA Cancer J Clin 51:15-36, 2001
Troisi RJ, Freedman AN, Devesa SS: Incidence of colorectal carcinoma in the US—An update of trends by gender, race, age, subsite, and stage, 1975-1994. Cancer 85:1670-1676, 1999
Cohen AM, Tremiterra S, Candela F, et al: Prognosis of node-positive colon cancer. Cancer 67:1859-1861, 1991
Morson B: President's address: The polyp-cancer sequence in the large bowel. Proc R Soc Med 67:451-457, 1974
Calvert PM, Frucht H: The genetics of colorectal cancer. Ann Intern Med 137:603-612, 2002
Winawer SJ, Fletcher RH, Miller L, et al: Colorectal cancer screening: Clinical guidelines and rationale. Gastroenterology 112:594-642, 1997
Muto T, Bussey HJ, Morson BC: The evolution of cancer of the colon and rectum. Cancer 36:2251-2270, 1975
Winawer SJ, Zauber AG, Ho MN, et al: Prevention of colorectal-cancer by colonoscopic polypectomy. N Engl J Med 329:1977-1981, 1993
Bomanji JB, Costa DC, Ell PJ: Clinical role of positron emission tomography in oncology. Lancet Oncol 2:157-164, 2001
Gould MK, Kuschner WG, Rydzak CE, et al: Test performance of positron emission tomography and computed tomography for mediastinal staging in patients with non-small-cell lung cancer: A meta-analysis. Ann Intern Med 139:879-892, 2003
Delbeke D: Oncological applications of FDG PET imaging: Brain tumors, colorectal cancer lymphoma and melanoma. J Nucl Med 40:591-603, 1999
Flanagan FL, Dehdashti F, Ogunbiyi OA, et al: Utility of FDG-PET for investigating unexplained plasma CEA elevation in patients with colorectal cancer. Ann Surg 227:319-323, 1998
Vitola JV, Delbeke D, Meranze SG, et al: Positron emission tomography with F-18-fluorodeoxyglucose to evaluate the results of hepatic chemoembolization. Cancer 78:2216-2222, 1996
Ruers TJM, Langenhoff BS, Neeleman N, et al: Value of positron emission tomography with [F-18]fluorodeoxyglucose in patients with colorectal liver metastases: A prospective study. J Clin Oncol 20:388-395, 2002
Drenth JPH, Nagengast FM, Oyen WJG: Evaluation of (pre-) malignant colonic abnormalities: Endoscopic validation of FDG-PET findings. Eur J Nucl Med 28:1766-1769, 2001
Yasuda S, Fujii H, Nakahara T, et al: F-18-FDG PET detection of colonic adenomas. J Nucl Med 42:989-992, 2001
Schlemper RJ, Riddell PH, Kato Y, et al: The Vienna classification of gastrointestinal epithelial neoplasia. Gut 47:251-255, 2000
Kubota K: From tumor biology to clinical PET: A review of positron emission tomography (PET) in oncology. Ann Nucl Med 15:471-486, 2001
Chen YK, Kao CH, Liao AC, et al: Colorectal cancer screening in asymptomatic adults: The role of FDG PET scan. Anticancer Res 23:4357-4361, 2003
Zhuang HM, Hickeson M, Chacko TK, et al: Incidental detection of colon cancer by FDG positron emission tomography in patients examined for pulmonary nodules. Clin Nucl Med 27:628-632, 2002
Tatlidil R, Jadvar H, Bading JR, et al: Incidental colonic fluorodeoxyglucose uptake: Correlation with colonoscopic and histopathologic findings. Radiology 224:783-787, 2002
Winawer SJ, Stewart ET, Zauber AG, et al: A comparison of colonoscopy and double-contrast barium enema for surveillance after polypectomy. N Engl J Med 342:1766-1772, 2000
Rex DK, Cutler CS, Lemmel GT, et al: Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies. Gastroenterology 112:24-28, 1997
Hixson LJ, Fennerty MB, Sampliner RE, et al: Prospective-study of the frequency and size distribution of polyps missed by colonoscopy. J Natl Cancer Inst 82:1769-1772, 1990(Maritte C.A. van Kouwen, )
ABSTRACT
PURPOSE: 2-(18F)-fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET) is a noninvasive imaging technique used clinically to detect malignant tumors. FDG-PET has been established as a tool for diagnosis of recurrent or metastatic colorectal carcinoma. Several case series suggest that FDG-PET also detects larger adenomas. The goal of this study was to investigate whether FDG-PET is able to detect colonic adenomas.
PATIENTS AND METHODS: FDG-PET was performed in 100 consecutive patients in whom colonic adenomas were suspected on barium enema (n = 47) or sigmoidoscopy (n = 53). A positive scan was defined as focal large bowel FDG accumulation. FDG-PET was followed in all cases by colonoscopy, and removed adenomas were examined histopathologically.
RESULTS: Colonoscopy confirmed the presence of adenomas in 68 of 100 patients. In 35 patients, there was focal FDG accumulation at site of the adenoma. The sensitivity of FDG-PET increased with adenoma size (21%, adenomas 1 to 5 mm; 47%, 6 to 10 mm; and 72%, > 11 mm). The sensitivity of FDG-PET also increased with the grade of dysplasia (33%, low grade; 76%, high grade; and 89%, carcinomas). The overall specificity was 84%.
CONCLUSION: FDG-PET detects colonic adenomas and the diagnostic test characteristics improve with size and grade of dysplasia of the adenoma.
INTRODUCTION
Colorectal adenocarcinoma (CRC) is the second most common cause of cancer-related deaths in the Western world and the fourth most frequent malignancy.1,2 The most important indicator of outcome after surgical resection of CRC is the pathologic stage at presentation. This is reflected by a five-year survival of approximately 90% in stage I, 60% to 80% in stage II, 40% to 60% in stage III and 4% in stage IV.3 These data underscore the importance of diagnosing CRC in an early or premalignant stage.
Colon carcinogenesis is regarded as a multifaceted process influenced by a variety of hereditary and lifestyle factors and involves a histologic progression from adenomatous polyps to colonic cancer.4,5 The incidence of adenomas increases with age, and the risk for malignancy with size. For example, malignant transformation is detected in only 1% of adenomas measuring < 1 cm, compared with 10% in larger adenomas.6,7 Indeed, detection and removal of adenomas is thought to result in a decrease in the incidence and mortality from colorectal cancer.8 A broad range of screening modalities for the detection of adenomas are available, such as fecal occult blood testing, endoscopy, barium enemas, and virtual colonoscopy. Although colonoscopy is regarded as the gold standard for the detection of premalignant lesions such as adenomatous polyps, it is not optimal in terms of examination performance, safety and patient acceptance, and this calls for a different diagnostic approach.
Recently, 2-fluoro [18-]-2-deoxyglucose positron emission tomography (FDG-PET) has been introduced in clinical practice for imaging of various malignancies. FDG-PET is a noninvasive, functional imaging technique, which uses the radiopharmaceutical FDG.9 It is clinically used to detect a wide variety of tumors including lymphoma, melanoma, lung cancer, and colon cancer.10 Several studies indicate a potential role of whole-body FDG-PET for the detection of malignant lesions of the colon.11 Indeed, FDG-PET is probably superior to conventional abdominal computed tomography in the assessment of patients with recurrent and metastatic CRC, and is reimbursed under Medicare for the diagnosis, staging and restaging of CRC.12-14
The situation is less defined for premalignant colorectal disease. The results from two retrospective studies suggest that FDG-PET may detect colonic adenomas, but it is uncertain which factors affect the diagnostic ability of FDG-PET.15,16 Furthermore, focal uptake in the colon is frequently identified on whole-body FDG-PET in patients studied for malignancies other than CRC, and the implications of these findings are unclear. These considerations stimulated us to design a prospective cohort study to establish the diagnostic test properties of FDG-PET in the detection of colonic adenomas using colonoscopy and histology as the gold standard.
PATIENTS AND METHODS
Patient Population
Between July 2001 and May 2004, we enrolled 100 consecutive Dutch adult patients who were referred to the endoscopy unit of the Department of Gastroenterology and Hepatology (Radboud University Nijmegen Medical Centre, the Netherlands) because of the suspicion of colonic polyp(s) on sigmoidoscopy or barium enema. The group consisted of 58 males and 42 females, mean age being 62 (SD 11) years. Patients were excluded if they were known with active inflammatory bowel disease or diverticulitis, because those diseases could lead to false-positive FDG-PET. We also excluded patients with known or suspicion of cancer. Patients with diabetes mellitus were excluded if fasting glucose concentration exceeded 15 mmol/L, as high serum glucose levels are responsible for false-negative FDG-PET. The study was approved by the local ethical committee. Informed consent was obtained.
FDG-PET Imaging
A dedicated PET scanner (ECAT-EXACT, Siemens/CTI, Knoxville, TN) was used for data acquisition. The spatial resolution of the PET camera was 5 mm. Before FDG injection, patients fasted for at least 6 hours. Intake of sugar free-liquids was permitted. Immediately before the procedure, the patients were hydrated with 500 mL of water. One hour after intravenous injection of 200 to 220 MBq FDG (Mallinckrodt Medical, Petten, the Netherlands) and 12 mg furosemide, emission and transmission images of the abdomen were acquired (10 minutes per bed position). The images were corrected for attenuation and reconstructed using the ordered-subsets expectation maximization (OSEM) algorithm. The reconstructed images were displayed in coronal, transverse and sagittal planes.
Interpretation
For this study, FDG-PET was analyzed visually by one of the staff physicians from the Department of Nuclear Medicine (Radboud University Nijmegen Medical Centre) and then reviewed by one of us (W.J.G.O), unaware of the results of the original interpretation of the FDG-PET, other diagnostic tests, or the final diagnosis. Results were judged to be abnormal if focal accumulation of the tracer was detected in the large bowel. If focal uptake was demonstrated, the uptake within the lesion was quantified by calculating the maximum and mean standard uptake value (SUV) according to the formula:
Maximum or mean region of interest activity in mCi/mL were derived from the calibrated activity measured in a manually drawn region of interest over the focal abnormalities (converting measured counts to accumulated activity in mCi) divided by the size of this region of interest (converting voxel size to volumes). Diffuse bowel accumulation was interpreted as physiologic. Disagreements were resolved by consensus.
Colonoscopy
Flexible endoscopy procedures were performed after thorough oral lavage with Klean prep. The procedure was performed by a gastroenterology trainee directly supervised by an experienced gastroenterologist (F.M.N.). The endoscopist was unaware of the result from the FDG-PET. The coecum was reached in 97% of all procedures. All lesions appearing as possibly malignant were biopsied and, whenever possible, complete polypectomy was performed. If polypectomy was technically impossible, surgical resection was performed (n = 3). All tissue specimens were immediately fixed in 10% buffered formalin solution before examination using hematoxylin and eosin staining. Each polyp was classified by an experienced pathologist as adenoma or other type polyp. Dysplasia was defined according to Vienna criteria as either low-grade dysplasia, high-grade dysplasia or carcinoma.17 The grade of dysplasia for each separate adenoma was recorded, and we scored the highest grade of dysplasia (low grade-high grade-carcinoma) in the adenoma for the purpose of the study. The results of the colonoscopy were considered negative in case of normal mucosal findings or if polyps were hyperplastic on histopathologic examination. Colonoscopy results were considered positive when histopathology confirmed the presence of either adenomas or carcinomas. The colonoscopic findings were reported in terms of the site and size of adenomas. Adenoma size was determined by endoscopical examination and data were verified by comparison with the data from macroscopical examination by the endoscopist. In case of multiple polyps, we scored the largest adenoma.
Experimental Procedures
After diagnosis of the colonic adenoma by either diagnostic flexible sigmoidoscopy (n = 53) or by double-contrast barium enema (n = 47), patients were put on the waiting list for a therapeutic colonoscopy that included polypectomy. FDG-PET was performed in the time-window between the diagnostic and therapeutic procedure. This study used colonoscopy as the gold standard, and patients were classified on the basis of their colonoscopic findings.
Statistics
Statistical analysis was performed with SAS statistical software, version 8.0. Data were described by means of frequency tables and descriptive statistics. The sensitivity was defined as the proportion patients with colonic polyps who actually had a positive FDG-PET. The specificity was defined as the proportion of patients without colonic polyps with a negative FDG-PET. The positive predictive value (PPV; the probability that the patient had colonic polyps when restricted to those patients who had a positive FDG-PET) was calculated as follows:
The negative predictive value (NPV; the probability that the patient will not have colonic polyps when restricted to all patients who had a negative FDG-PET) was defined as
RESULTS
Detection of Presence of Adenomas
In 100 patients in whom a previous sigmoidoscopy (n = 53) or double contrast barium studies of the colon (n = 47) had shown lesions suggestive for colonic adenomas, colonoscopy detected adenomas in 68 patients, but failed to confirm the presence of an adenoma in the colon in 32 patients. In 18 of 32 patients the preceding examination had been a double contrast barium study, whereas sigmoidoscopy had been the diagnostic tool for 14 of 32 patients. Within the latter group, all adenomas detected were small and had already been completely removed by biopsy during the initial sigmoidoscopy. Histopathologic examination of removed lesions demonstrated hyperplastic polyps (n = 4), adenomatous polyps (n = 5), a lipoma (n = 1), an inflammation polyp (n = 1) and normal mucosa (n = 3). In all of these cases biopsy completely removed the lesion.
FDG-PET
All 100 patients underwent FDG-PET, and 40 patients had significant focal colonic FDG accumulation.
Diagnostic Characteristics of FDG-PET as a Test for Detecting Colonic Adenomas
In 35 patients with a positive FDG-PET, colonoscopy showed a significant mucosal lesion. These lesions included carcinomas (n = 8), and adenomas with high-grade dysplasia (n = 13) and low-grade dysplasia (n = 14). The median size of the adenomas was 15 mm (range, 2 to 50 mm) and most adenomas were at least 15 mm (n = 21). The adenomas were predominantly located in the sigmoid (n = 26). The calculated mean and SD of the maximum SUV of this total group was 5.0 (range, 1.3 to 14.4). In all cases, FDG accumulation corresponded with the adenoma site. In 33 patients, FDG-PET failed to demonstrate colonic hotspots, but colonoscopy did detect significant abnormalities. Twenty-eight patients had adenomas with low-grade dysplasia, whereas four patients had adenomas with high-grade dysplasia. In one patient, colonoscopy detected an 8 mm large tubular adenoma with histopathologic evidence for foci of invasive carcinoma that had been missed by PET. The median size of the PET-negative adenomas was 6 mm (range, 2 to 30 mm), and almost all adenomas were smaller than 11 mm (n = 23).
Table 1 summarizes the FDG-PET results. FDG-PET is positive in 80% of larger adenomas, but negative in the large majority of small adenomas (1 to 6 mm; P < .001). The sensitivity of FDG-PET increases with the diameter of the polyp and rises from 21% for small (1 to 5 mm) polyps to 72% for larger (> 11 mm) polyps. The specificity of FDG-PET was 84%. The sensitivity of FDG-PET was 33% for adenomas with low-grade dysplasia, 76% for adenomas with high-grade dysplasia, and 89% for carcinomas (Table 2). For the detection of polyps, regardless of size or dysplastic grade, FDG-PET has a PPV of 88% and a NPV of 45%, which suggests that FDG-PET has a better accuracy for detecting the presence polyps rather than excluding its absence. When compared with the subgroup of patients without polyps, the detection of large (> 11 mm) adenomas yields a PPV of 82% and a NPV of 75%. Likewise, for the detection of carcinoma, the PPV is 61% while the NPV is 96%. These results in our study suggest that FDG-PET was not likely to miss colonic carcinoma and that an FDG-PET with negative results in most instances excluded the presence of colonic carcinoma.
Specificity
In 27 patients, neither FDG-PET nor colonoscopy demonstrated any mucosal abnormality in the colon. There were five patients in whom PET demonstrated focal FDG accumulation in the large bowel, but in whom colonoscopy failed to reveal adenomas. In a single patient, colonoscopy was compatible with sigmoidal diverticulitis that corresponded with localized FDG uptake at the same site. Four patients had a normal colonoscopy. The localization of false-positive hot spots was sigmoid (n = 2) and ascending colon (n = 2). In this group, the calculated average and standard deviation (SD) of the maximum SUV were 3.5 and 1.8, respectively. The site of the adenoma had no influence on the FDG-PET outcome (P = .46).
DISCUSSION
Our results show that FDG-PET detects premalignant colonic adenomas. The grade of dysplasia and size of the adenoma were clearly important discriminatory factors that predicted detection, while other parameters such as location of the adenoma, sex, or age did not affect the results. We found that FDG-PET has sensitivity of 72% for the detection of larger (> 11 mm) and of 76% for adenomas with high-grade dysplasia. The specificity of FDG-PET was 84%, whereas for the detection of polyps (regardless of size or dysplastic grade) the PPV is 88% and NPV is 45%. Intensity of tracer uptake does not allow discrimination between adenomas (n = 27) and carcinomas (n = 8) as maximum SUVs (average, 5.9; SD, 3.7), respectively (average, 6.5; SD, 3.0) did not differ.18,19
Few retrospective studies suggest that detection of adenomas by FDG-PET is possible.15,16,19-21 In those studies, the adenoma size was the most important variable to predict outcome of FDG-PET, and the detection rate increased significantly with size of the adenoma. There are no studies on the effect of the grade of dysplasia on diagnostic performance of FDG-PET. In contrast to our results, one study suggested that localization matters.16 This study showed a higher detection rate in coecum, ascending and descending colon compared with the sigmoid and transverse colon.16 Significant shortcomings of these earlier studies is their cross-sectional nature, which makes them prone to positive selection. FDG-PET had been performed either for staging malignant disease other than CRC20,21 or had been done as part of a cancer-screening program,16 and the detection of adenomas was merely coincidental. More importantly, colonoscopy had not been performed in all cases.20,21
We detected that a positive double-contrast barium enema in 10 (21%) of 47 patients was followed by a negative colonoscopy. These data are in agreement with the false-positive rate of barium enemas in The National Polyp Study.22 In eight false-positive barium enemas, FDG-PET was true negative.
The PPV of FDG-PET was 88% which suggests that in 88% of the cases, colonic FDG accumulation accurately identifies an adenoma. One limitation of our finding is that the adenoma prevalence in our population (68%) was relatively high. It is possible that in a population with a lower prevalence of adenoma, the PPV of FDG-PET will decrease as the proportion of patients with false-positive tracer uptake may increase. False-positive results are an unwanted aspect of any diagnostic tool, and FDG-PET is no exception. In 5% of our cases FDG-PET generated false-positive results and we were able to identify several important contributing factors. Normal accumulation in the colonic wall mucosa may appear focal and can be misinterpreted as clinically significant (nonphysiological) FDG accumulation. In our study, two patients were classified as false positive, because colonoscopy failed to identify any mucosal abnormality despite focal FDG accumulation. It is possible that colonoscopy missed a polyp, as back-to-back colonoscopy studies demonstrated a missing rate of 15% to 24% for (mostly small) colonic adenomas.23,24 Further, we identified diverticulitis as a potential confounder. In these cases, FDG accumulation is most likely caused by uptake in inflammatory cells.
Our study was designed to investigate the diagnostic ability of FDG-PET in detection of colonic adenomas. We aimed to include patients with lesions indicative of colonic polyps on sigmoidoscopy or barium enema. On histopathologic examination nine polyps were carcinomas. FDG-PET detected all but a single 8-mm large polyp with histopathologic signs of invasive carcinoma. In addition, FDG-PET also missed three other clinical relevant large adenomas with high grade dysplasia. Given the overall NPV of 45% and the relatively high cost of FDG-PET, the procedure does not qualify as a screening modality for colonic adenomas because very small polyps and polyps with a low grade of dysplasia are missed. Conversely, because of the high PPV (88%) of FDG-PET for clinically significant colonic adenomas, we feel that incidental identification of focal FDG accumulation necessitates follow-up by total colonoscopy. The timing of colonoscopy is, of course, determined by the nature and severity of the underlying disease that led to the initial request for FDG-PET.
FDG-PET detects benign colonic adenomas, but the sensitivity of this test depends on the size and grade of dysplasia of the colonic adenoma, with the latter being the most important discriminative factor that predicts detection of the adenoma. Many dysplastic adenomas are missed by FDG-PET (especially when small and with low-grade dysplasia), which is one of the factors limiting its utility as a screenings modality for adenomas.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank M. van Oijen, MD, Department of Gastroenterology and Hepatology, University Medical Centre St Radboud, Nijmegen, the Netherlands, for his assistance with the statistical analysis.
NOTES
Both W.J.G. Oyen and J.P.H. Drenth contributed equally to this article.
J.P.H. Drenth is a recipient of a NWO-VIDI grant.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Greenlee RT, Hill-Harmon MB, Murray T, et al: Cancer statistics, 2001. CA Cancer J Clin 51:15-36, 2001
Troisi RJ, Freedman AN, Devesa SS: Incidence of colorectal carcinoma in the US—An update of trends by gender, race, age, subsite, and stage, 1975-1994. Cancer 85:1670-1676, 1999
Cohen AM, Tremiterra S, Candela F, et al: Prognosis of node-positive colon cancer. Cancer 67:1859-1861, 1991
Morson B: President's address: The polyp-cancer sequence in the large bowel. Proc R Soc Med 67:451-457, 1974
Calvert PM, Frucht H: The genetics of colorectal cancer. Ann Intern Med 137:603-612, 2002
Winawer SJ, Fletcher RH, Miller L, et al: Colorectal cancer screening: Clinical guidelines and rationale. Gastroenterology 112:594-642, 1997
Muto T, Bussey HJ, Morson BC: The evolution of cancer of the colon and rectum. Cancer 36:2251-2270, 1975
Winawer SJ, Zauber AG, Ho MN, et al: Prevention of colorectal-cancer by colonoscopic polypectomy. N Engl J Med 329:1977-1981, 1993
Bomanji JB, Costa DC, Ell PJ: Clinical role of positron emission tomography in oncology. Lancet Oncol 2:157-164, 2001
Gould MK, Kuschner WG, Rydzak CE, et al: Test performance of positron emission tomography and computed tomography for mediastinal staging in patients with non-small-cell lung cancer: A meta-analysis. Ann Intern Med 139:879-892, 2003
Delbeke D: Oncological applications of FDG PET imaging: Brain tumors, colorectal cancer lymphoma and melanoma. J Nucl Med 40:591-603, 1999
Flanagan FL, Dehdashti F, Ogunbiyi OA, et al: Utility of FDG-PET for investigating unexplained plasma CEA elevation in patients with colorectal cancer. Ann Surg 227:319-323, 1998
Vitola JV, Delbeke D, Meranze SG, et al: Positron emission tomography with F-18-fluorodeoxyglucose to evaluate the results of hepatic chemoembolization. Cancer 78:2216-2222, 1996
Ruers TJM, Langenhoff BS, Neeleman N, et al: Value of positron emission tomography with [F-18]fluorodeoxyglucose in patients with colorectal liver metastases: A prospective study. J Clin Oncol 20:388-395, 2002
Drenth JPH, Nagengast FM, Oyen WJG: Evaluation of (pre-) malignant colonic abnormalities: Endoscopic validation of FDG-PET findings. Eur J Nucl Med 28:1766-1769, 2001
Yasuda S, Fujii H, Nakahara T, et al: F-18-FDG PET detection of colonic adenomas. J Nucl Med 42:989-992, 2001
Schlemper RJ, Riddell PH, Kato Y, et al: The Vienna classification of gastrointestinal epithelial neoplasia. Gut 47:251-255, 2000
Kubota K: From tumor biology to clinical PET: A review of positron emission tomography (PET) in oncology. Ann Nucl Med 15:471-486, 2001
Chen YK, Kao CH, Liao AC, et al: Colorectal cancer screening in asymptomatic adults: The role of FDG PET scan. Anticancer Res 23:4357-4361, 2003
Zhuang HM, Hickeson M, Chacko TK, et al: Incidental detection of colon cancer by FDG positron emission tomography in patients examined for pulmonary nodules. Clin Nucl Med 27:628-632, 2002
Tatlidil R, Jadvar H, Bading JR, et al: Incidental colonic fluorodeoxyglucose uptake: Correlation with colonoscopic and histopathologic findings. Radiology 224:783-787, 2002
Winawer SJ, Stewart ET, Zauber AG, et al: A comparison of colonoscopy and double-contrast barium enema for surveillance after polypectomy. N Engl J Med 342:1766-1772, 2000
Rex DK, Cutler CS, Lemmel GT, et al: Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies. Gastroenterology 112:24-28, 1997
Hixson LJ, Fennerty MB, Sampliner RE, et al: Prospective-study of the frequency and size distribution of polyps missed by colonoscopy. J Natl Cancer Inst 82:1769-1772, 1990(Maritte C.A. van Kouwen, )