Japanese Journal of Clinical Oncology Advance Access originally published online on April 19, 2008
Japanese Journal of Clinical Oncology 2008 38(5):347-353; doi:10.1093/jjco/hyn032
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© The Author (2008). Published by Oxford University Press. All rights reserved
Preoperative Diagnosis of Lymph Node Metastases of Colorectal Cancer by FDG-PET/CT
1 Department of Colorectal and Pelvic Surgery, National Cancer Center Hospital East, Kashiwa, Chiba
2 Division of Nuclear Imaging Medicine, National Cancer Center Hospital East, Kashiwa, Chiba, Japan
3 Department of General Surgical Science (Surgery I), Gunma University Graduate School of Medicine, Maebashi, Gunma, Japan
For reprints and all correspondence: Masaaki Ito, Department of Colorectal and Pelvic Surgery, National Cancer Center Hospital East, 6-5-1 Kashiwa, Chiba 277-8577, Japan. E-mail: maito{at}east.ncc.go.jp
Received January 20, 2008; accepted March 23, 2008
 |
Abstract |
|---|
 TOP Abstract INTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONFundingReferences |
|---|
Purpose: The purpose of this study was to assess the diagnostic value of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) for lymph node (LN) metastasis of colorectal cancer.
Methods: FDG-PET/CT was used to preoperatively evaluate 88 patients with colorectal cancer. In this study, LN sites were divided into proximal and distant according to their distance from the primary tumor. The FDG-PET/CT images were evaluated by three criteria; nodal diameter, abnormal uptake and maximum standardized uptake value (SUV). We compared the diagnostic ability of these methods for LN metastasis at proximal and distant sites.
Results: The mean SUV of the malignant LNs was significantly higher than that of the benign LNs. The sensitivity, specificity and accuracy of diagnosis by abnormal uptake were 28.6, 92.9 and 75.0%, those by nodal diameter using cutoff value of 10 mm were 30.6, 95.3 and 74.4% and those by SUV using cutoff value of 1.5 were 53.1, 90.6 and 80.1%, respectively. The sensitivity, specificity and accuracy of diagnosis based on optimal SUV were 51.2, 85.1 and 69.3% in the proximal site and 62.5, 92.5 and 89.7%, respectively, in the distant site.
Conclusions: FDG-PET/CT is useful for preoperative diagnosis of distant LN metastases of colorectal cancers.
Key Words: FDG-PET/CT • colorectal cancer • lymph node metastasis
|
INTRODUCTION |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONFundingReferences |
|---|
Computed tomography (CT), magnetic resonance imaging (MRI) and endorectal ultrasonography (EUS) are commonly used for preoperative diagnosis of lymph node (LN) metastasis in colorectal cancer. However, many investigators have reported low sensitivity for LN diagnosis using these modalities (1,2) and that is the main reason for being unable to tailor the extent of LN dissection according to the results of these diagnoses.
18F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) has been reported to be an effective method of diagnosing malignant disease (3,4). FDG-PET has often been used to diagnose postoperative recurrence (5,6) and distant metastasis (7,8) of colorectal cancer. However, FDG-PET has also been reported to have low sensitivity for the detection of LN metastasis (9,10), and FDG-PET could not determine the anatomical location for small lesions, such as LNs.
In this respect, 18F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT), which combines the use of a full-ring-detector PET scanner and multidetector-row helical CT scanner, has been developed and introduced widely into clinical practice (11,12). To our knowledge, there has been little information reported on the value of FDG-PET/CT preoperative nodal diagnosis in colorectal cancer.
The purpose of this study was to assess the diagnostic value of FDG-PET/CT for LN metastases of colorectal cancer.
|
PATIENTS AND METHODS |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONFundingReferences |
|---|
Patients
The subjects of this study were 88 patients diagnosed with primary colorectal cancer between May 2004 and May 2005. They were consecutive patients who were scheduled for surgical treatment during that period. There were 52 males and 36 females, and their mean age was 60.6 years (range: 23–89 years). The location of the primary colorectal cancer was the colon in 37 patients, and the rectum in 51 patients. Lateral LN dissection was performed in 16 of 51 patients with lower rectal cancer. Of those 16 patients, using FDG PET/CT we were able to preoperatively detect malignant LNs in 10 patients, and to pathologically confirm involved nodes in 9 of 10 patients. We excluded evaluations concerning lateral LNs from this study, because we had not yet had enough patients with lateral LN dissection. All patients underwent whole-body (from the upper legs to the base of the skull) FDG-PET/CT preoperatively and surgical treatment in the National Cancer Center Hospital East, Japan.
FDG-PET/CT
The FDG-PET/CT examinations were performed with a Discovery LS8 (GE Healthcare, Milwaukee, WI, USA). The axial field of view (FOV) is 15.2 cm, and the system produces 35 images with a slice thickness of 4.25 mm using 2D mode. The matrix size is 128 x 128 with a 50 cm transaxial FOV. All patients fasted for at least 6 h before the injection of FDG. The dose of FDG was 370 MBq, and the FDG-PET/CT examination was started 60 min after taking the dose of FDG. First, the CT scan was performed from the base of the skull to the pelvic floor with the following settings: 140 kV, 100–380 mA, 0.6 s per CT rotation, pitch 1.35:1, table rotation 22.5 mm/s, and slice thickness 5.0 mm. A whole-body PET scan was performed immediately after the CT scan was completed. The emission time was set at 4 min per bed position, and a total of six incremental bed positions were scanned. PET images were corrected for attenuation by using the CT data. The PET images were reconstructed by an iterative reconstruction with ordered-subset expectation maximization algorithm (two iterations, 14 subsets) and reformatted into transverse, coronal and sagittal views.
Classification of LNs
The intra-abdominal LNs were divided into two groups according their anatomical location. The first group consisted of LNs along the vascular arcades of marginal vessels, which were proximal to the primary tumor, and will be referred to as proximal group (PG). The second group consisted of LNs along the superior mesenteric artery, inferior mesenteric artery, ileocolic artery, right colic artery, middle colic artery, left colic artery, sigmoid arteries or superior rectal artery and in the para-aortic region and will be referred to as the distant group (DG).
Image Interpretation
The FDG-PET/CT images were interpreted on the manufacturer's review station (Xeleris; GE Healthcare, Milwaukee, WI, USA) by two observers who were experienced in FDG-PET/CT interpretation and blinded to the clinical information. When there was a discrepancy between the observers' interpretations, a consensus was reached by discussion. We evaluated visible LNs by three criteria; nodal diameter, abnormal uptake and SUV. We diagnosed LN metastasis by abnormal FDG uptake of the images, regardless of the shape or size of the node, and this method was defined as visual diagnosis. We also measured the maximum axial diameter (defined as size diagnosis) and maximum standardized uptake value (SUV) (defined as SUV diagnosis) of all LNs with greater FDG uptake than normal organs or surrounding tissue on the FDG-PET/CT images. The cutoff value of SUV was represented as the optimal cutoff value when accuracy was the highest. The widely accepted size criterion for LN metastasis by gastrointestinal tumors has been 
10 mm in diameter (13) and we used 10 mm as the cutoff value of nodal diameter.
When two or more LNs were diagnosed as malignant in the PG or DG, the LN with the highest SUV was selected as representative in each region and evaluated in this study. When LNs were not detected on the FDG-PET/CT images, the SUV was represented as the SUV of background in this study.
Analysis of the Results
We calculated the sensitivity, specificity, positive predictive value (PPV), negative predictive value and accuracy of FDG-PET/CT for diagnosing malignant LNs in each diagnosis based on nodal diameter, abnormal uptake and SUV. We assessed the diagnostic values in total of 176 LNs groups, which equally divided into 88 proximal LN groups and 88 distant LN groups for all patients.
Data were analysed by using statistical software (Dr SPSS II, version 11.01J, SPSS Japan Inc). Independent Student's t-test was used in this study. A corrected P < 0.05 was considered evidence of statistical significance.
|
RESULTS |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONFundingReferences |
|---|
Pathological Diagnosis for LN Metastasis
Surgical resection and regional LN dissection were performed in 88 patients. In 43 patients, malignant LNs were diagnosed by pathological examination in 49 groups of LNs. Of these 49 groups of LNs, 41 were PG and 8 were DG. Six patients had malignant LNs in both PG and DG.
FDG-PET/CT Findings
Table 1 shows rates of malignant LNs according to SUV and nodal diameter of representative LNs in 176 LN groups in 88 patients. LNs < 10 mm in diameter were found in 153 groups, whereas LNs 10 mm in diameter were in 23 groups. Rates of malignancy were 20.9% (32/153) of LNs < 10 mm in diameter and 73.9% (17/23) of LNs 10 mm in diameter. Of 153 LNs < 10 mm in diameter, rate of LN metastases were 16.8% (23/137) in 0–1.5 of SUV; 57.1% (4/7) in 1.5–2.5 of SUV; 42.9% (3/7) in 2.5–3.5 of SUV; and 100% (2/2) in more than 3.5 of SUV, respectively. Of 23 LNs 10 mm in diameter, there were no malignant LNs in 0–1.5 of SUV, and the rates were 60.0% (3/5) in 1.5–2.5 of SUV; 57.1% (4/7) in 2.5–3.5 of SUV; and 100% (10/10) in more than 3.5 of SUV, respectively.
|
In 43 LNs that were detected, the mean SUV of the malignant LNs was 6.3 (range: 1.1–33.8), and the mean SUV of the benign LNs was 2.5 (range: 1.3–3.3). The mean maximum axial diameters of the malignant and benign LNs were 11.6 mm (range: 5.5–25.6 mm) and 9.7 mm (range: 5.0–15.0 mm), respectively. The difference between the SUVs was significant (P = 0.01), however, the difference in maximum axial diameter was not (P = 0.3) (Table 2).
|
Diagnostic Ability of FDG-PET/CT
The sensitivity, specificity and accuracy by visual diagnosis of FDG-PET/CT images were 28.6, 92.9 and 75.0%, respectively (Table 3). The receiver operating characteristic (ROC) curve analysis is shown in Fig. 1. The best combination between sensitivity and specificity, and thus the highest accuracy to distinguish benign from malignant LNs, was found at an SUV cutoff value of 1.5, and the optimal SUV cutoff value of 1.5 was chosen. At the cutoff value, the sensitivity, specificity and accuracy of SUV diagnosis were 53.1, 90.6 and 80.1%, respectively, and the sensitivity, specificity and accuracy of size diagnosis were 30.6, 95.3 and 74.4%, respectively. SUV diagnosis showed the best result in sensitivity among three methods of diagnosis, and demonstrated about 25% improvement of sensitivity compared to size or visual diagnosis.
|
|
Table 4 shows the results of the SUV diagnosis, which were assessed at each LN group. In the PG, the sensitivity, specificity and accuracy of SUV diagnosis were 51.2, 85.1 and 69.3%, respectively, and in the DG, 62.5, 92.5% and 89.7%, respectively. The results in the DG were superior to those in the PG. A representative image with LN metastasis in DG is shown in Fig. 2.
|
|
Next, we assessed influence of inflammation on diagnosis by FDG-PET/CT, because we had the impression that a false-positive accumulation of FDG was often seen at the proximal LN group. Figure 3 showed one of the false-positive cases. There was a huge tumor that invaded directly into the urinary bladder with abnormal white blood cell (WBC) counts. We could see the swollen LNs with abnormal uptake of FDG near primary tumor.
|
When FDG-PET/CT examinations were performed, the mean WBC count and C-reactive protein (CRP) values were 11 225 (range: 7700–14 600) and 3.6 (range: 0.09–8.20) in the four false-positive cases in the PG. In the true-positive cases, the WBC and CRP values were 6157 (range: 3600–10 800) and 1.1 (range: 0.06–4.3), respectively. The difference in WBC count in both false- and true-positive cases was statistically significant (P = 0.001) (Table 5). In patients with normal WBC counts, the diagnostic ability of SUV diagnosis was better than in those with abnormal WBC counts, in terms of specificity (85.1 versus 91.9) and PPV (75.0 versus 85.7) in the PG (Table 6). In the DG, only slightly improvements of specificity and PPV were found.
|
|
|
DISCUSSION |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONFundingReferences |
|---|
The results of this study indicated that FDG-PET/CT was a useful modality for preoperative diagnosis of LN metastases of colorectal cancer. Especially, the superiority was shown in the diagnosis of distant LNs, by using optimized cutoff value of SUV.
We compared the diagnostic abilities for LN metastasis among three methods based on different criteria; abnormal uptake, nodal diameter and SUV. The conventional evaluation of LN metastasis has mainly been based on nodal size. The widely accepted size criterion for LN metastasis by gastrointestinal tumors is 10 mm in diameter (13). Monig et al. (14) reported a significant difference in the average size of metastatic and non-metastatic LNs in colon cancer patients, but those up to 90% of the metastatic LNs were <10 mm in diameter, and Herrera-Ornelas et al. (15) reported similar results. The present study also described that malignant LNs were included at rates of 20.9% of LNs < 10 mm in diameter. Moreover, in such small LNs, we could catch many malignant LNs by using SUV diagnosis, and in LNs 10 mm in diameter, SUV of all malignant LNs was >1.5. These results indicated that size criterion was not reliable in correctly assessing the nodal status exactly, and SUV showed great benefit in accurately diagnosing LNs in colorectal cancer.
The presence of abnormal FDG uptake on FDG-PET or FDG-PET/CT images has been widely accepted as a criterion for differentiation between benign and malignant disease. FDG uptake has been evaluated mainly by visual inspection or calculation of the SUV. The value of the SUV has been reported to be affected by the size of the lesion and not to be very accurate in small lesions <15 mm (16), so each investigator has been applied his or her own visual criteria for diagnosis of LN metastasis (17–20). In the present series, although some malignant LNs did not have abnormal uptake, the mean SUV of malignant LNs was significantly higher than that of benign LNs. Consequently, SUV diagnosis yielded more accurate results than the visual or size diagnosis. Diagnosis using SUV could be the most valuable method of assessing FDG uptake objectively to diagnose LN metastasis of colorectal cancer.
In this study, diagnostic ability for metastasis at the distant sites showed better results than that at the proximal sites. FDG-PET/CT provided very low sensitivity at sites near the primary tumor, the same as that reported in previous studies of FDG-PET (9,10). The low sensitivity at the proximal sites in the present study may be explained by LNs near a primary tumor that cannot be distinguished from the tumor or physiological uptake on FDG-PET/CT images.
Inflammatory change by the primary colorectal cancer may have affected the results of our study. In the PG, the mean WBC count in the false-positive cases by SUV diagnosis was significantly higher than in the true-positive cases. The specificity and PPV of SUV diagnosis in the patients with a normal WBC count were better in the PG. No such difference was seen in the DG. This suggests that LNs near a primary tumor with inflammation often appear to have abnormal FDG uptake and are diagnosed as false positive. This is one of the limitations caused by the properties of FDG. FDG is a glucose analog that is taken up into cells by the large number of GLUT-1 transporters located on the cell membrane. Macrophages and neutrophils, whose numbers are increased in inflamed tissue, use glucose as an energy source and have an increased level of FDG uptake, the same as cancer cells. Several new tracers have been developed in recent years as agents to image tumor cells through different and specific aspects of malignant metabolism (21–23).
A sensitivity of 63% and an accuracy of 90% based on SUV-based diagnosis for distant LN are much better than those reported by CT, MRI or PET alone (1,2,9,10). SUV diagnosis in this study improves sensitivity by approximately 20–30% compared to sensitivity that has been reported by other modalities of images. Thus, FDG-PET/CT has a great ability to detect for small malignant lesions. However, this result is not still enough to tailor the range of LN dissection in the surgical treatments of colorectal cancer patients. More studies are needed using other imaging agents to achieve higher sensitivity.
In conclusion, FDG-PET/CT is a powerful tool for detection of distant LN metastases in colorectal cancer patients. SUV is a better criterion for detection of malignant LNs than abnormal FDG uptake or the nodal diameter. On the other hand, the inflammation near primary tumor yields false-positive accumulation of FDG to proximal nodes.
|
Funding |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONFundingReferences |
|---|
This study was supported in part by Third-term Comprehensive Control Research for Cancer grant from the Ministry of Health Labor, and Welfare of Japan.
Conflict of interest statement
None declared.
|
Acknowledgments |
|---|
We thank K. Inoue, H. Kitamura, K. Satoh and A. Hirayama for their technical support in FDG-PET/CT image acquisition and Y. Kojima for PET tracer production.
|
References |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONFundingReferences |
|---|
1 Bipat S, Glas AS, Slors FJ, Zwinderman AH, Bossuyt PM, Stoker J. Rectal cancer: local staging and assessment of lymph node involvement with endoluminal US, CT, and MR imaging–a meta-analysis. Radiology (2004) 232:773–83.
2 Heriot AG, Grundy A, Kumar D. Preoperative staging of rectal carcinoma. Br J Surg (1999) 86:17–28.[CrossRef][Web of Science][Medline]
3 Kato H, Miyazaki T, Nakajima M, Takita J, Kimura H, Faried A, et al. The incremental effect of positron emission tomography on diagnostic accuracy in the initial staging of esophageal carcinoma. Cancer (2005) 103:148–56.[CrossRef][Web of Science][Medline]
4 Mochiki E, Kuwano H, Katoh H, Asao T, Oriuchi N, Endo K. Evaluation of 18F-2-deoxy-2-fluoro-D-glucose positron emission tomography for gastric cancer. World J Surg (2004) 28:247–53.[CrossRef][Web of Science][Medline]
5 Strauss LG, Clorius JH, Schlag P, Lehner B, Kimmig B, Engenhart R, et al. Recurrence of colorectal tumors: PET evaluation. Radiology (1989) 170:329–32.
6 Valk PE, Abella-Columna E, Haseman MK, Pounds TR, Tesar RD, Myers RW, et al. Whole-body PET imaging with [18F]fluorodeoxyglucose in management of recurrent colorectal cancer. Arch Surg (1999) 134:503–11. (Discussion 11–3).
7 Erturk SM, Ichikawa T, Fujii H, Yasuda S, Ros PR. PET imaging for evaluation of metastatic colorectal cancer of the liver. Eur J Radiol (2006) 58:229–35.[CrossRef][Web of Science][Medline]
8 Wiering B, Krabbe PF, Jager GJ, Oyen WJ, Ruers TJ. The impact of fluor-18-deoxyglucose-positron emission tomography in the management of colorectal liver metastases. Cancer (2005) 104:2658–70.[CrossRef][Web of Science][Medline]
9 Abdel-Nabi H, Doerr RJ, Lamonica DM, Cronin VR, Galantowicz PJ, Carbone GM, et al. Staging of primary colorectal carcinomas with fluorine-18 fluorodeoxyglucose whole-body PET: correlation with histopathologic and CT findings. Radiology (1998) 206:755–60.
10 Kantorova I, Lipska L, Belohlavek O, Visokai V, Trubac M, Schneiderova M. Routine (18)F-FDG PET preoperative staging of colorectal cancer: comparison with conventional staging and its impact on treatment decision making. J Nucl Med (2003) 44:1784–8.
11 Bar-Shalom R, Yefremov N, Guralnik L, Gaitini D, Frenkel A, Kuten A, et al. Clinical performance of PET/CT in evaluation of cancer: additional value for diagnostic imaging and patient management. J Nucl Med (2003) 44:1200–9.
12 Ell PJ. The contribution of PET/CT to improved patient management. Br J Radiol (2006) 79:32–6.
13 Monig SP, Zirbes TK, Schroder W, Baldus SE, Lindemann DG, Dienes HP, et al. Staging of gastric cancer: correlation of lymph node size and metastatic infiltration. AJR Am J Roentgenol (1999) 173:365–7.
14 Monig SP, Schroder W, Baldus SE, Holscher AH. Preoperative lymph-node staging in gastrointestinal cancer–correlation between size and tumor stage. Onkologie (2002) 25:342–4.[CrossRef][Web of Science][Medline]
15 Herrera-Ornelas L, Justiniano J, Castillo N, Petrelli NJ, Stulc JP, Mittelman A. Metastases in small lymph nodes from colon cancer. Arch Surg (1987) 122:1253–6.
16 Hoffman EJ, Huang SC, Phelps ME. Quantitation in positron emission computed tomography: 1. Effect of object size. J Comput Assist Tomogr (1979) 3:299–308.[Web of Science][Medline]
17 de Jong IJ, Pruim J, Elsinga PH, Vaalburg W, Mensink HJ. Preoperative staging of pelvic lymph nodes in prostate cancer by 11C-choline PET. J Nucl Med (2003) 44:331–5.
18 Sironi S, Buda A, Picchio M, Perego P, Moreni R, Pellegrino A, et al. Lymph node metastasis in patients with clinical early-stage cervical cancer: detection with integrated FDG PET/CT. Radiology (2006) 238:272–9.[CrossRef][Web of Science][Medline]
19 Williams AD, Cousins C, Soutter WP, Mubashar M, Peters AM, Dina R, et al. Detection of pelvic lymph node metastases in gynecologic malignancy: a comparison of CT, MR imaging, and positron emission tomography. AJR Am J Roentgenol (2001) 177:343–8.
20 Yun M, Lim JS, Noh SH, Hyung WJ, Cheong JH, Bong JK, et al. Lymph node staging of gastric cancer using (18)F-FDG PET: a comparison study with CT. J Nucl Med (2005) 46:1582–8.
21 Been LB, Suurmeijer AJ, Cobben DC, Jager PL, Hoekstra HJ, Elsinga PH. [18F]FLT-PET in oncology: current status and opportunities. Eur J Nucl Med Mol Imaging (2004) 31:1659–72.[CrossRef][Web of Science][Medline]
22 Francis DL, Freeman A, Visvikis D, Costa DC, Luthra SK, Novelli M, et al. In vivo imaging of cellular proliferation in colorectal cancer using positron emission tomography. Gut (2003) 52:1602–6.
23 Francis DL, Visvikis D, Costa DC, Arulampalam TH, Townsend C, Luthra SK, et al. Potential impact of [18F]3'-deoxy-3'-fluorothymidine versus [18F]fluoro-2-deoxy-D-glucose in positron emission tomography for colorectal cancer. Eur J Nucl Med Mol Imaging (2003) 30:988–94.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||