AJR AJR-based Continuing Ed for Technologists
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Roedl, J. B.
Right arrow Articles by Blake, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Roedl, J. B.
Right arrow Articles by Blake, M. A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Hotlight (NEW!)
Right arrow
What's Hotlight?
DOI:10.2214/AJR.07.3320
AJR 2008; 191:515-521
© American Roentgen Ray Society

Original Research

Visual PET/CT Scoring for Nonspecific 18F-FDG Uptake in the Differentiation of Early Malignant and Benign Esophageal Lesions

Johannes B. Roedl1,2, Rivka R. Colen1,2, Kevin King1, Alan J. Fischman2, Peter R. Mueller1 and Michael A. Blake1

1 Division of Abdominal and Interventional Radiology, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, 55 Fruit St., Boston, MA 02114.
2 Division of Nuclear Medicine, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA.

Received October 18, 2007; accepted after revision February 9, 2008.

 
Address correspondence to J. B. Roedl (johannes.roedl{at}gmail.com).


Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
OBJECTIVE. The purpose of our study was to evaluate a visual PET/CT scoring system for the differentiation of benign and early malignant esophageal uptake.

MATERIALS AND METHODS. Thirty-six consecutive patients with precancerous or early malignant esophageal lesions including Barrett's esophagus, Tis, T1, and T2 adenocarcinomas were eligible. Findings of these patients were compared with 66 patients who had reported increased esophageal 18F-FDG uptake due to benign esophageal disorders. Lesions were evaluated with scores using the following characteristics in PET/CT: FDG uptake intensity (low = 0, moderate = 1, high = 2), FDG uptake eccentricity (concentric = 0, eccentric = 1), FDG uptake focality (diffuse = 0, segmental = 1, focal = 2), esophageal thickness on the CT component (normal = 0, thickening = 1, mass = 2), and location (distal third of the esophagus = 0, middle third of the esophagus = 1, proximal third of the esophagus = 2).

RESULTS. Early malignant lesions had higher scores in FDG uptake intensity (p = 0.003; chi-square), eccentricity (p < 0.001), and focality (p < 0.001) compared with benign lesions. No significant difference was seen in esophageal thickness on CT (p = 0.168) and in location of the lesion (p = 0.291). Binary logistic regression analysis with a stepwise forward inclusion of all score components including the maximum standardized uptake value (SUV) of the lesions revealed that a total score combining eccentricity and focality scores has the highest accuracy of predicting early malignant disease. Using a threshold of equal or higher than 2 in the combined total focality–eccentricity score, the sensitivity was 83.3% and specificity was 68.2% for predicting early malignant disease.

CONCLUSION. Focality and eccentricity of FDG uptake prove to be valuable PET/CT characteristics for the differentiation of nonspecific FDG uptake in the esophagus.

Keywords: esophagus • 18F-FDG • nonspecific uptake • PET • PET/CT


Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Incidental 18F-FDG uptake in the gastrointestinal tract is a common phenomenon in PET/CT [1–3]. Incidental gastrointestinal FDG assimilation might be physiologic secondary to swallowed saliva; active smooth muscle; active mucosa, feces, microbial uptake; or lymphatic tissue uptake [4, 5]. Incidental uptake might also be inflammatory, which is commonly seen, especially in the sigmoid colon and in the distal esophagus [6].

The clinical dilemma is that early malignant esophageal lesions might present with an uptake pattern and CT findings similar to those of physiologic or benign sites [7, 8]. This creates the clinical challenge of deciding whether to recommend further evaluation with endoscopy in patients with increased FDG uptake on PET or nonspecific esophageal thickening on CT but without known underlying esophageal abnormality.

So far, early detection of malignant esophageal lesions and differentiation from benign uptake has been difficult with PET and CT alone [7, 8]. The advent of combined PET/CT, a technique that provides accurately fused metabolic and morphologic information in one examination, has the potential to reveal new criteria or characteristics for malignant lesions that were not available on PET or CT alone.

In the present study, we evaluated whether early malignant lesions can be distinguished from benign lesions and which characteristics in PET/CT are useful to make this differentiation. A PET/CT score that encompasses findings from the PET component, the CT component, or from the combined PET/CT was evaluated for this clinically important differentiation.


Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Patients and Standard of Reference
Our institutional review board gave approval and informed consent was waived before this retrospective HIPAA-compliant study was performed. Study participants were consecutive patients with benign (n = 66), premalignant, or malignant (n = 36) esophageal lesions with a T stage lower than T3 and with increased FDG uptake (above liver FDG intensity) on PET/CT examinations. Six of the 36 early malignant lesions were incidentally found without a history or diagnosis of premalignant or malignant esophageal disease. All of the benign esophageal lesions (n = 66) were inci dentally found.

Final diagnosis of malignant esophageal disease with typing, grading, and T staging was determined within 2 weeks of the pretherapy PET/CT examination. The standard of reference was either pretherapy endoscopic sonography with histology or postsurgical pathology if patients were scheduled for primary surgery without neoadjuvant therapy. Benign esophageal uptake other than reflux esophagitis had to be confirmed by endoscopy. Reflux esophagitis was confirmed by endoscopy or, when endoscopy was not available, based on clinical diagnosis and a follow-up period of at least 6 months without signs of malignancy. Clinical diagnosis of reflux esophagitis was based on clinical symptoms and on an increased quantity and frequency of acidic (low pH) reflux at the gastroesophageal junction measured with a nasoesophageal tube over a period of 24 hours (24-hour pH-metry). For further analysis, the early malignant lesions were grouped on the basis of their therapeutic approach: Therapy group 1 consisted of metaplastic lesions, which are usually controlled and closely observed by regular endoscopic examinations. Therapy group 2 included dysplastic lesions and T1 adenocarcinomas, which are treated by primary surgery. Therapy group 3 encompassed T2 adenocarcinomas, which are treated by neoadjuvant chemoradiotherapy followed by surgery.

Patients with diabetes mellitus were not enrolled in the study. An electronic database of patient records was reviewed retrospectively to acquire the above-mentioned clinical information. Before assessment of the PET/CT score, these clinical data had been documented in an "esophageal tumor-metry" database run by statistical software (SPSS version 15.0, SPSS). The two readers who determined the PET/CT score were blinded to the patients' diagnoses and clinical information.

PET/CT
All patients were asked to fast for at least 6 hours before PET/CT. Subjects drank two cups of water (each 10 oz) as a negative contrast material 1 hour before imaging. A dose of 555 MBq of FDG was injected IV 60 minutes before imaging. Imaging was performed with an integrated PET/CT scanner (Biograph Sensation 16; Siemens Medical Solutions), which includes a 16-MDCT component and a high-resolution PET system with lutetium oxyorthosilicate crystals. A low-dose CT scan for attenuation correction was performed including an area from the neck to the midthigh. PET scan was performed in 7 acquisition beds with a scan-bed acquisition time ranging from 3 to 7 minutes depending on patient weight. PET images were reconstructed at a section thickness of 2.4 mm. After PET, contrast-enhanced CT was performed with 100 mL of 300 mg I/mL of contrast material (iopamidol [Isovue 300, Bracco]) along with 20 mL saline. The injection was automatic with a dual-head injector (Stellant, Medrad) at 2 mL/s with a 60-second delay. The tube voltage of the diagnostic CT was 140 kVp, the tube current–time product was 170 mAs, the table feed was 15 mm/s, and the pitch was 1.5. Diagnostic CT images were reconstructed with a 2-mm section thickness at 2-mm intervals. The PET data were fused with the contrast-enhanced CT data using fusion software (Reveal-MVS, Mirada Solutions).


Figure 1
View larger version (125K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1A —71-year-old man with diagnosis of T1 adenocarcinoma (arrows): focal and eccentric pattern (focality–eccentricity score, 3). "Hot iron" color map (A), axial (B), and sagittal (C) PET/CT images show 20-step color map.

 


Figure 2
View larger version (127K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1B —71-year-old man with diagnosis of T1 adenocarcinoma (arrows): focal and eccentric pattern (focality–eccentricity score, 3). "Hot iron" color map (A), axial (B), and sagittal (C) PET/CT images show 20-step color map.

 


Figure 3
View larger version (157K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1C —71-year-old man with diagnosis of T1 adenocarcinoma (arrows): focal and eccentric pattern (focality–eccentricity score, 3). "Hot iron" color map (A), axial (B), and sagittal (C) PET/CT images show 20-step color map.

 


Figure 4
View larger version (121K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2A —62-year-old man with diagnosis of Barrett's esophagus (arrows): segmental and eccentric pattern (focality–eccentricity score, 2). "Hot iron" color map (A), axial (B), and sagittal (C) PET/CT images show 20-step color map.

 


Figure 5
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2B —62-year-old man with diagnosis of Barrett's esophagus (arrows): segmental and eccentric pattern (focality–eccentricity score, 2). "Hot iron" color map (A), axial (B), and sagittal (C) PET/CT images show 20-step color map.

 


Figure 6
View larger version (157K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2C —62-year-old man with diagnosis of Barrett's esophagus (arrows): segmental and eccentric pattern (focality–eccentricity score, 2). "Hot iron" color map (A), axial (B), and sagittal (C) PET/CT images show 20-step color map.

 
Image Interpretation
Two readers, each with 3 years of experience with PET/CT, who were unaware of the diagnosis and clinical data, reviewed the images. Assessment of the PET/CT score components was performed independently by both readers and interrater reliability was assessed. FDG focality, eccentricity, and location scores were evaluated qualitatively. Esophageal thickness and intensity scores were determined both qualitatively and quantitatively by both readers. Definitions of the PET/CT score components both for visual and quantitative analysis are given in Table 1. The maximum standardized uptake value (SUV) was assessed in a region of interest (ROI) (diameter, 0.3 cm) centered on the highest uptake of the lesion. For statistical analysis of the quantitative score components, the average values of the results of both readers were used. In case of discordant results between both readers in the qualitative score components, a third investigator was consulted to make the final decision.


View this table:
[in this window]
[in a new window]

  		TABLE 1: Components and Definitions of PET/CT Score for the Differentiation of Benign and Early Malignant Esophageal Lesions

 

Components of the PET/CT Score
Table 1 summarizes the PET/CT score with its components and definitions. Intensity score and esophageal thickness score can be assessed both in a qualitative (visual) and quantitative manner (Table 1). Because only benign and early malignant abnormalities were included, most lesions in this study were expected to have SUV results between 2 and 4. Therefore, an SUV threshold of 3, which lies between the endpoints of SUV 2 and SUV 4, was considered most appropriate for defining FDG intensity levels. For the qualitative (visual) analysis of the intensity score, the liver uptake (average liver uptake approximates an SUV of 2) [1, 9] and the brain uptake (average brain uptake approximates an SUV of 4) [10] were used as references. Uptake closer to liver than to brain (SUV ≤ 3) was defined as low, whereas uptake closer to brain than to liver (SUV > 3) was defined as moderate. Uptake equal to or higher than brain uptake was defined as intense (SUV ≥ 4) (Table 1).

Esophageal thickness on the CT component was classified in the visual score as normal, thickened, or masslike. This correlates on a quantitative scale with a maximal esophageal wall diameter of 15 mm or less, greater than 15 to 30 mm, and more than 30 mm [11]. Location was defined by the three parts of the esophagus including the proximal, middle, and distal thirds (Table 1). The examples of Figures 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B, 3C, 4A, 4B, and 4C are given to illustrate focality and eccentricity scores: An eccentric pattern was defined as asymmetric uptake with tendency to one side of the esophageal wall (Figs. 1A, 1B, 2A, and 2B), whereas concentric pattern was defined as uptake within or symmetric to the esophageal lumen or diffuse along the esophageal wall [12] (Figs. 3A, 3B, 4A, and 4B).


Figure 7
View larger version (116K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3A —64-year-old man with diagnosis of radiation esophagitis (arrows): segmental and concentric pattern (focality–eccentricity score: 1). "Hot iron" color map (A), axial (B), and sagittal (C) PET/CT images show 20-step color map.

 

Figure 8
View larger version (117K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3B —64-year-old man with diagnosis of radiation esophagitis (arrows): segmental and concentric pattern (focality–eccentricity score: 1). "Hot iron" color map (A), axial (B), and sagittal (C) PET/CT images show 20-step color map.

 

Figure 9
View larger version (143K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3C —64-year-old man with diagnosis of radiation esophagitis (arrows): segmental and concentric pattern (focality–eccentricity score: 1). "Hot iron" color map (A), axial (B), and sagittal (C) PET/CT images show 20-step color map.

 

Figure 10
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4A —59-year-old woman with diagnosis of reflux esophagitis (arrows): diffuse and concentric pattern (focality–eccentricity score: 0). Axial PET/CT image shows "hot-iron" color map.

 

Figure 11
View larger version (82K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4B —59-year-old woman with diagnosis of reflux esophagitis (arrows): diffuse and concentric pattern (focality–eccentricity score: 0). Diffuse uptake is best seen on PET alone (gray-scale) as shown on axial (B) and sagittal (C) images.

 

Figure 12
View larger version (58K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4C —59-year-old woman with diagnosis of reflux esophagitis (arrows): diffuse and concentric pattern (focality–eccentricity score: 0). Diffuse uptake is best seen on PET alone (gray-scale) as shown on axial (B) and sagittal (C) images.

 

Focality of FDG uptake has been classified as focal (Figs. 1A, 1B, and 1C), segmental (Figs. 2A, 2B, 2C, 3A, 3B, and 3C), and diffuse (Figs. 4A, 4B, and 4C) in several previous studies on the gastrointestinal tract, especially the colon [1–3, 13]. Determining the focality (focal, segmental, or diffuse) of a lesion is a visual assessment that combines intensity and length of the lesion in one complementary parameter. Exact definitions of the subtypes of focality are given in Table 1. As depicted in Figures 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B, and 3C, we used both the standard "hot-iron" color map (Figs. 1A, 2A, and 3A) and the "20-step" color scale (Figs. 1B, 2B, and 3B) for image interpretation. In the 20-step color scale, each step represents a 5% difference in uptake. For example, if two tissues are four color steps apart, the difference in uptake is 20%. In our study the 20-step color scale was useful for identifying focal lesions and for the detection of subtle uptake. When comparing the axial hot-iron images with the 20-step color scale images in Figures 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B, and 3C, it becomes clear that the 20-step color scale accentuates the focal nature of the uptake. These "alternative" 20-step color scale images can be very helpful in highlighting abnormalities.

Statistical Analysis
Kappa values and linear regression analysis were performed to evaluate interrater reliability. Differences in demographic data, tumor characteristics, and score components were determined with chi-square, Fisher's exact, and Student's t tests. A binary logistic regression analysis with stepwise inclusion of all score components was performed to evaluate the best predictor of early malignant lesions. Furthermore, receiver operating characteristic (ROC) curves, area under the curve (AUC), sensitivity, and specificity were calculated. All statistical tests were two sided and performed at the 5% level of significance with SPSS for Windows, version 15.0 (SPSS).


Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Patient Characteristics and Patient Groups
Thirty-six patients with early malignant lesions and 66 patients with benign lesions with increased esophageal FDG uptake on PET/CT examinations were included in the study. There was no significant difference in patient characteristics between early malignant and benign lesions including sex (male 81% vs 65%; p = 0.160; chi-square test), age (67.9 ± 10.6 years vs 63.5 ± 14.2 years; t = –1.64; p = 0.104), weight (80.7 ± 15.8 kg vs 76.6 ± 14.8 kg; t = –1.3; p = 0.197), and height (171.4 ± 10.5 cm vs 168.5 ± 9.8 cm; t = –1.39; p = 0.167).

PET/CT Score in Benign and Early Malignant Lesions
Evaluation of the PET/CT score components was successfully performed in all patients. Table 2 shows that there were significant differences between early malignant and benign esophageal lesions in the focality (p < 0.001), eccentricity (p < 0.001), and intensity (p = 0.003) scores (chi-square test). No significant difference was seen in esophageal thickness (p = 0.168), and location (p = 0.291) scores (chi-square test). The maximum SUV was significantly higher in early malignant than in benign esophageal lesions (p < 0.001, Mann-Whitney test). A binary logistic regression analysis with forward stepwise inclusion of all PET/CT score components and the maximum SUV of the lesions revealed that a combined focality–eccentricity score had the highest accuracy for distinguishing early malignant from benign esophageal disease: The sum of the focality and eccentricity score to a total focality–eccentricity score yielded higher accuracy than focality and eccentricity scores alone.


View this table:
[in this window]
[in a new window]

  		TABLE 2: PET/CT Characteristics in Benign and Malignant Esophageal Lesions

 

The focality–eccentricity score was also more accurate than the intensity score and the maximum SUV of the lesion. Using a focality–eccentricity score threshold of ≥ 2, the sensitivity was 83.3% and the specificity was 68.3% for differentiating benign from early malignant esophageal lesions. This means that lesions with a focal and eccentric (total score, 3), or segmental and eccentric (total score, 2), or focal and concentric pattern (total score, 2) are suspicious for malignant disease. ROC curves for focality–eccentricity score, intensity score, and maximum SUV are shown in Figure 5. As can be seen in Table 3, the focality–eccentricity score (AUC, 0.808) had a higher AUC, sensitivity, and specificity in differentiating early malignant from benign esophageal lesions than intensity score (AUC, 0.688; sensitivity, 80.6%; specificity, 51.5%) and maximum SUV (AUC, 0.717; sensitivity, 83.3%; specificity, 51.5%).


Figure 13
View larger version (10K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5 —Receiver operator characteristic (ROC) curves of focality–eccentricity score (solid line), intensity score (long dashes), and absolute standardized uptake value (short dashes) in differentiating benign from early malignant esophageal lesions. Detailed results are given in Table 3.

 

View this table:
[in this window]
[in a new window]

  		TABLE 3: Diagnostic Performance in the Differentiation of Benign from Early Malignant Esophageal Lesions

 

The concordance between both readers in determining the components of the PET/CT score was strong, with kappa values of 0.730 for the focality score, 0.705 for the eccentricity score, 0.732 for the qualitative intensity score, 0.834 for the quantitative intensity score, 0.742 for the qualitative esophageal thickness score, 0.820 for the quantitative esophageal thickness score, and 0.859 for the location score. The high intrareader reliability between the quantitative and qualitative assessment of the esophageal thickness score (kappa values for reader 1, 0.789; reader 2, 0.864) and the intensity score (reader 1, 0.865; reader 2, 0.791) showed that the visual score strongly correlates with the quantitative assessment of the score. A linear regression model showed excellent interrater reliability for maximum SUV measurements with an R [2] value of 0.988.

PET/CT Score in Different Therapy Groups of Early Malignant Lesions
As described in the "Materials and Methods" section, the early malignant lesions were subclassified for further analysis in three groups on the basis of the therapeutic approach (Table 4). Metaplastic lesions (therapy group 1, n = 6) are usually closely observed by regular endoscopic examinations, whereas dysplastic lesions and T1 adenocarcinomas (therapy group 2, n = 15) are treated with esophagectomy and T2 adenocarcinomas are treated with neoadjuvant chemoradiotherapy followed by surgery (therapy group 3, n = 15). Each of these therapy groups had significantly higher focality–eccentricity scores than benign lesions (group 1, p = 0.012; group 2, p < 0.001; group 3, p = 0.001; chi-square test). The intensity score, however, was not significantly different between group 1 and benign lesions (p = 0.415) and between group 2 and benign lesions (p = 0.091). Only group 3 showed a higher intensity score than benign lesions (p < 0.001) Sensitivities and specificities for distinguishing each therapy group from benign lesions when applying a threshold of ≥ 2 in the focality–eccentricity score and a threshold of ≥ 1 in the intensity score are shown in Table 4.


View this table:
[in this window]
[in a new window]

  		TABLE 4: Focality–Eccentricity Score and Intensity Score in Subgroups of Early Malignant Lesions Compared with Benign Group

 

Focality–Eccentricity Score in the Differentiation of Incidental Uptake
Six of the 36 early malignant lesions were incidentally found without prior history or diagnosis of premalignant or malignant esophageal disease. Because all benign esophageal lesions (n = 66) were incidental, the rate of incidentally found malignant lesions was 8.3% (6/72) in this study population. These six malignancies included three dysplastic lesions, two T1 adenocarcinomas, and one T2 adenocarcinoma. When comparing those six incidentally found malignant lesions with the 66 incidentally found benign esophageal lesions, sensitivities for identifying the malignancies would have been 100% both for the focality–eccentricity score, the intensity score, and maximum SUV when applying the previously established thresholds: focality–eccentricityscore ≥ 2, intensity score ≥ 1, and maximum SUV > 2.98). However, the specificity was significantly higher for the focality–eccentricity score (68.2%) than for the intensity score (51.5%) and the maximum SUV (51.5%).


Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Nonspecific esophageal uptake, which is mainly secondary to reflux esophagitis, is a common finding when interpreting PET/CT or PET studies. This is not surprising considering that about 15% of the population suffers from reflux disease [14]. These benign lesions, however, might resemble early esophageal malignancies. The clinical challenge is that, on the one hand, the only chance for cure of esophageal carcinoma is its early detection, and, on the other hand, the rate of false positives among nonspecific esophageal uptake is too high to recommend further evaluation with endoscopy in all of those cases. Therefore, the differentiation between early malignant and benign lesions as described in the present study has important clinical implications. To make this differentiation, we applied a PET/CT score that encompasses various lesion characteristics on PET/CT.

We found no significant differences in esophageal thickness on CT and location of the lesion between the 36 early malignant and the 66 benign lesions. However, intensity, focality, and eccentricity scores were significantly increased in the early malignant group compared with benign lesions. Furthermore, it was shown that the focality and eccentricity scores combined in a total focality–eccentricity score had the highest sensitivity (83%) and specificity (68%) in predicting malignant disease. The focality–eccentricity score was more accurate than the intensity score and the maximum SUV.

Our study shows the advantages of PET/CT compared with PET alone: Assessment of eccentricity, for example, is only possible with combined PET/CT. Eccentric uptake means uptake relative to the esophageal lumen or esophageal wall, structures that are not visible on PET alone.

For a more detailed evaluation of the subgroups of early malignant disease, we grouped those stages together that had a similar prognosis and the same therapeutic approach. Barrett's esophagus is treated by endoscopic follow-up (group 1), both Tis tumors (dysplastic lesions) and T1 tumors are treated with surgery alone (group 2), and T2 tumors, with neoadjuvant chemoradiotherapy and surgery (group 3). All three groups had a significantly increased focality–eccentricity score compared with benign lesions. On the basis of the intensity of the lesion alone, only group 3 (T2 tumors), but not group 2 (Tis and T1) and group 1 (Barrett's esophagus) could be distinguished from benign esophageal lesions.

Several studies have specifically investigated incidental FDG uptake in the gastrointestinal tract. Kamel et al. [2] reported that 60.9% and Israel et al. [1] that 85.3% of patients with incidental gastrointestinal tract uptake were found to have premalignant or malignant lesions. Investigations of the colon alone found a malignant proportion of 61.9% [13] and 37.0% [3] among incidental colonic FDG uptakes. In the present investigation, we evaluated incidental uptake in the esophagus, and a rate of 8.3% (6/72) was found to be malignant. Our results are consistent with the only previous report, to our knowledge, specifically evaluating incidental uptake in the esophagus. Rampin et al. [12] described in a letter to the editor that 2.22% of unexpected esophageal FDG uptake harbored malignant disease on follow-up examinations. Both our study and the report by Rampin et al. showed that the relative proportion of malignancies among the incidental lesions was lower in the esophagus than in other parts of the gastrointestinal tract.

Our study has limitations. Endoscopic proof of diagnosis was not available in all patients with reflux esophagitis. Furthermore, the study is retrospective and from a single institution.

In conclusion, we found that the combined focality–eccentricity score as defined in this study may help in the interpretation of non-specific esophageal uptake on PET/CT examinations. This could assist in making the clinically challenging decision of whether to recommend further evaluation with endoscopy in cases of nonspecific esophageal FDG uptake. Endoscopy has the highest malignancy yield when FDG uptake is focal and eccentric, segmental and eccentric, or focal and concentric.


References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Israel O, Yefremov N, Bar-Shalom R, et al. PET/CT detection of unexpected gastrointestinal foci of 18F-FDG uptake: incidence, localization patterns, and clinical significance. J Nucl Med 2005; 46:758 -762[Abstract/Free Full Text]
  2. Kamel EM, Thumshirn M, Truninger K, et al. Significance of incidental 18F-FDG accumulations in the gastrointestinal tract in PET/CT: correlation with endoscopic and histopathologic results. J Nucl Med 2004; 45:1804 -1810[Abstract/Free Full Text]
  3. Tatlidil R, Jadvar H, Bading JR, Conti PS. Incidental colonic fluorodeoxyglucose uptake: correlation with colonoscopic and histopathologic findings. Radiology 2002;224 : 783-787[Abstract/Free Full Text]
  4. Shreve PD, Anzai Y, Wahl RL. Pitfalls in oncologic diagnosis with FDG PET imaging: physiologic and benign variants. RadioGraphics 1999;19 : 61-77[Abstract/Free Full Text]
  5. Cook GJ, Fogelman I, Maisey MN. Normal physiological and benign pathological variants of 18-fluoro-2-deoxyglucose positron-emission tomography scanning: potential for error in interpretation. Semin Nucl Med 1996; 26:308 -314[CrossRef][Medline]
  6. Prabhakar HB, Sahani DV, Fischman AJ, Mueller PR, Blake MA. Bowel hot spots at PET-CT. RadioGraphics 2007;27 : 145-159[Abstract/Free Full Text]
  7. Kato H, Miyazaki T, Nakajima M, et al. The incremental effect of positron emission tomography on diagnostic accuracy in the initial staging of esophageal carcinoma. Cancer 2005;103 : 148-156[CrossRef][Medline]
  8. Little SG, Rice TW, Bybel B, et al. Is FDG-PET indicated for superficial esophageal cancer? Eur J Cardiothorac Surg2007; 31:792 -797
  9. Blake MA, Slattery JM, Kalra MK, et al. Adrenal lesions: characterization with fused PET/CT image in patients with proved or suspected malignancy—initial experience. Radiology2006; 238:970 -977[Abstract/Free Full Text]
  10. von Schulthess GK. Clinical PET-CT and SPECT-CT in body oncology, 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2007: 261-286
  11. Lefor AT, Merino MM, Steinberg SM, et al. Computerized tomographic prediction of extraluminal spread and prognostic implications of lesion width in esophageal carcinoma. Cancer 1988;62 : 1287-1292[CrossRef][Medline]
  12. Rampin L, Nanni C, Fanti S, Rubello D. Value of PET-CT fusion imaging in avoiding potential pitfalls in the interpretation of 18F-FDG accumulation in the distal oesophagus. (letter) Eur J Nucl Med Mol Imaging 2005;32 : 990-992[CrossRef][Medline]
  13. Gutman F, Alberini JL, Wartski M, et al. Incidental colonic focal lesions detected by FDG PET/CT. AJR 2005;185 : 495-500[Abstract/Free Full Text]
  14. Locke GR 3rd, Talley NJ, Fett SL, Zinsmeister AR, Melton LJ 3rd. Prevalence and clinical spectrum of gastroesophageal reflux: a population-based study in Olmsted County, Minnesota. Gastroenterology 1997;112 : 1448-1456[CrossRef][Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Roedl, J. B.
Right arrow Articles by Blake, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Roedl, J. B.
Right arrow Articles by Blake, M. A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Hotlight (NEW!)
Right arrow
What's Hotlight?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS