DOI: 10.1148/rg.276065742
RadioGraphics 2007;27:1635-1652
© RSNA, 2007
PET/CT of Esophageal Cancer: Its Role in Clinical Management1
John F. Bruzzi, FFRRCSI,
Reginald F. Munden, MD,
Mylene T. Truong, MD,
Edith M. Marom, MD,
Bradley S. Sabloff, MD,
Gregory W. Gladish, MD,
Revathy B. Iyer, MD,
Tin-Su Pan, PhD,
Homer A. Macapinlac, MD, and
Jeremy J. Erasmus, MD
1 From the Division of Diagnostic Imaging (J.F.B., R.F.M., M.T.T., E.M.M., B.S.S., G.W.G., R.B.I., H.A.M., J.J.E.) and Department of Imaging Physics (T.S.P.), M. D. Anderson Cancer Center, Houston, Tex. Received November 7, 2006; revision requested March 14, 2007, and received April 2; accepted April 13. H.A.M. receives grant support from General Electric, consults for General Electric and Radiology Corporation of America, and is a speaker for Siemens; all other authors have no financial relationships to disclose.
Address correspondence to J.F.B., Department of Radiology, University College Hospital, Galway, Ireland (e-mail: bruzzij{at}hotmail.com).
 |
Abstract
|
|---|
Positron emission tomography (PET)/computed tomography (CT) has important utility and limitations in the initial staging of esophageal cancer, evaluation of response to neoadjuvant therapy, and detection of recurrent malignancy. Esophageal cancer is often treated by using a combined modality approach (chemotherapy, radiation therapy, and esophagectomy); correct integration of PET/CT into the conventional work-up of esophageal cancer requires a multidisciplinary approach that combines the information from PET/CT with results of clinical assessment, diagnostic CT, endoscopic gastroduodenoscopy, and endoscopic ultrasonography. PET/CT has limited utility in T staging of esophageal cancer and relatively limited utility in detection of dissemination to locoregional lymph nodes. However, PET/CT allows detection of metastatic disease that may not be identifiable with other methods. PET/CT is not sufficiently reliable in the individual patient for determination of treatment response in the primary tumor. Interpretation of PET/CT results is optimized by understanding the diagnostic limitations and pitfalls that may be encountered, together with knowledge of the natural history of esophageal cancer and the staging and treatment options available.
© RSNA, 2007
|
LEARNING OBJECTIVES
|
|---|
After reading this article and taking the test, the reader will be able to:
- Describe the natural history of esophageal cancer.
- List the diagnostic limitations and pitfalls that may be encountered in PET/CT of esophageal cancer.
- Discuss the staging and treatment options available for esophageal cancer.
|
Introduction
|
|---|
The incidence of esophageal cancer in the United States is increasing, and an estimated 14,550 new cases were expected to be diagnosed in 2006 (1). In patients with early-stage malignancy at presentation, esophagectomy is the treatment of choice and is potentially curative. Unfortunately, most patients have locally advanced disease at presentation, and 20%–30% have distant metastases (2). In patients with locally advanced disease without distant metastases, esophagectomy is a potential treatment option after neoadjuvant chemotherapy and radiation therapy in those who do not develop distant metastases during therapy (3–12). Consequently, in all patients with potentially resectable disease, accurate staging at initial presentation and assessment of therapeutic response after neoadjuvant therapy are important in regard to optimal management.
In patients with esophageal carcinoma being considered for esophagectomy, conventional staging methods include upper endoscopic gastroduodenoscopy, endoscopic ultrasonography (US), and computed tomography (CT) of the thorax and abdomen. The routine use of integrated positron emission tomography (PET)/CT with 2-[fluorine 18]fluoro-2-deoxy-D-glucose (FDG) in evaluation of patients with esophageal carcinoma is increasing and has been reported to be useful in initial staging of esophageal carcinoma, assessment of therapeutic response after neoadjuvant therapy, and detection of recurrent malignancy (2,13–22). However, accurate interpretation of PET/CT results in patients with esophageal carcinoma requires knowledge of the technical aspects of PET/CT image acquisition and the interpretative pitfalls that may be encountered, as well as an understanding of how the disease manifests and disseminates, the staging criteria used, and the different management strategies available.
In this article, we review the clinical utility of PET/CT of esophageal carcinoma and describe the diagnostic findings and pitfalls that are important in multidisciplinary management of this malignancy.
|
Technique of PET/CT
|
|---|
PET/CT is performed on an integrated scanner that combines both multisection CT and PET capabilities in two sequential gantries, avoiding the need for patient motion between the CT and PET components of the study and thereby leading to accurate coregistration of the CT and PET data. Patients undergo fasting for at least 6 hours before the PET/CT study. PET images are acquired during shallow breathing in the two-dimensional mode for 3 minutes per bed position 60–90 minutes after intravenous administration of 555–740 MBq of FDG. PET images are reconstructed by using standard vendor-provided reconstruction algorithms that incorporate ordered subset expectation maximization. Attenuation correction of PET images is performed by using attenuation data from the CT component of the examination; emission data are corrected for scatter, random events, and dead-time losses by using the manufacturer’s software.
The CT component of the study comprises a multidetector CT examination from the base of the skull to the upper thighs (120 mA, 140 kVp, table speed = 13.5 mm per rotation). Use of both oral and intravenous contrast material is desirable to improve anatomic localization of abnormalities detected on the PET images. Multiplanar CT images are reconstructed with a section thickness of 3.75 mm. PET and fused PET/CT images are analyzed both qualitatively and semiquantitatively. The intensity of FDG uptake within specific lesions is calculated by using a volume of interest over the lesion, according to the following formula (19): SUVmax = mean measured activity in the volume of interest (millicuries per milliliter)/injected dose of FDG (millicuries) per gram of body weight, where SUV = standardized uptake value.
|
Initial Staging of Esophageal Cancer
|
|---|
Esophageal cancer is most commonly staged according to the American Joint Committee on Cancer (AJCC) staging guidelines, which incorporate the T, N, and M classification (23). The T descriptor refers to the depth of tumor penetration through the mucosal layers of the esophageal wall; the N descriptor specifies involvement of locoregional lymph nodes; and the M descriptor indicates the presence or absence of distant metastases (Table).
|
T Stage
|
|---|
The T descriptor is determined according to the extent of invasion by the primary tumor through the mucosal layers of the esophagus and into the adventitia and adjacent organs. A higher T classification is associated with a greater likelihood of nodal metastatic disease and poorer long-term survival (24,25). The location of the primary tumor in the esophagus (cervical, intrathoracic, gastroesophageal junction) does not affect prognosis (25,26) but does influence the management options available and is relevant for tumor staging, primarily by determining the location of locoregional lymph nodes (Table).
When the primary tumor is confined to the esophageal wall (T1–T2), primary resection is possible. Extension into the periesophageal adventitia signifies a T3 carcinoma, which is still potentially resectable but is usually treated with combined modality therapy. Invasion of tumor into adjacent organs such as the aorta, diaphragmatic crus, liver, or pancreas indicates T4 disease.
Endoscopic US is the modality of choice for assessing the depth of penetration of the primary tumor through the esophageal wall at initial staging (27–29). PET/CT has limited utility in T staging, although advanced tumor invasion into adjacent organs (T4) can sometimes be detected at initial staging. Signs of invasion into adjacent organs at CT and PET/CT include blurring of the periesophageal fat, loss of the normal fat planes between the esophagus and adjacent structures, and a large concave interface with adjacent structures (Fig 1).
View larger version (93K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. T4 adenocarcinoma of the midesophagus with invasion into the thoracic aorta. (a) Coronal fused PET/CT image shows a primary FDG-avid tumor in the midesophagus (arrowhead). (b) Axial contrast-enhanced CT image obtained at the level of the midesophagus shows marked esophageal wall thickening (arrowhead), which corresponds to the primary tumor. In addition, there is loss of the normal fat plane between the esophagus and thoracic aorta with tumor extension into the periesophageal fat. Note the broad interface with the aorta (arrow), a finding suggestive of locoregional invasion. (c, d) Axial PET (c) and fused PET/CT (d) images obtained at the same level as b show marked FDG uptake in the primary tumor (arrowhead) and in the component of adventitial invasion (arrow). Aortic invasion by the tumor was confirmed at attempted esophagectomy.
|
|
View larger version (43K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. T4 adenocarcinoma of the midesophagus with invasion into the thoracic aorta. (a) Coronal fused PET/CT image shows a primary FDG-avid tumor in the midesophagus (arrowhead). (b) Axial contrast-enhanced CT image obtained at the level of the midesophagus shows marked esophageal wall thickening (arrowhead), which corresponds to the primary tumor. In addition, there is loss of the normal fat plane between the esophagus and thoracic aorta with tumor extension into the periesophageal fat. Note the broad interface with the aorta (arrow), a finding suggestive of locoregional invasion. (c, d) Axial PET (c) and fused PET/CT (d) images obtained at the same level as b show marked FDG uptake in the primary tumor (arrowhead) and in the component of adventitial invasion (arrow). Aortic invasion by the tumor was confirmed at attempted esophagectomy.
|
|
View larger version (35K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1c. T4 adenocarcinoma of the midesophagus with invasion into the thoracic aorta. (a) Coronal fused PET/CT image shows a primary FDG-avid tumor in the midesophagus (arrowhead). (b) Axial contrast-enhanced CT image obtained at the level of the midesophagus shows marked esophageal wall thickening (arrowhead), which corresponds to the primary tumor. In addition, there is loss of the normal fat plane between the esophagus and thoracic aorta with tumor extension into the periesophageal fat. Note the broad interface with the aorta (arrow), a finding suggestive of locoregional invasion. (c, d) Axial PET (c) and fused PET/CT (d) images obtained at the same level as b show marked FDG uptake in the primary tumor (arrowhead) and in the component of adventitial invasion (arrow). Aortic invasion by the tumor was confirmed at attempted esophagectomy.
|
|
View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1d. T4 adenocarcinoma of the midesophagus with invasion into the thoracic aorta. (a) Coronal fused PET/CT image shows a primary FDG-avid tumor in the midesophagus (arrowhead). (b) Axial contrast-enhanced CT image obtained at the level of the midesophagus shows marked esophageal wall thickening (arrowhead), which corresponds to the primary tumor. In addition, there is loss of the normal fat plane between the esophagus and thoracic aorta with tumor extension into the periesophageal fat. Note the broad interface with the aorta (arrow), a finding suggestive of locoregional invasion. (c, d) Axial PET (c) and fused PET/CT (d) images obtained at the same level as b show marked FDG uptake in the primary tumor (arrowhead) and in the component of adventitial invasion (arrow). Aortic invasion by the tumor was confirmed at attempted esophagectomy.
|
|
Although T4 disease precludes primary resection, esophagectomy remains a treatment option in patients in whom the extent of local invasion responds to neoadjuvant therapy to the point where it is considered to be T3 or less. 
However, after neoadjuvant therapy, loss of the normal tissue planes due to radiation-induced fibrosis and necrosis makes repeat evaluation of the T classification difficult with all imaging modalities (20,30), and local organ invasion may not be detected until the time of attempted esophagectomy (Fig 1).
|
Interpretative Pitfalls in Determination of T Classification with PET/CT
|
|---|
Although most esophageal carcinomas appear FDG avid at PET/CT, the reduced spatial and contrast resolutions of PET/CT limit visualization of the anatomic extent of the primary mass and preclude evaluation of the depth of local tumor invasion in most cases. Early-stage carcinomas, in particular, may not be detectable at all with either CT or PET/CT (31).
|
N Stage
|
|---|
Staging of nodal metastases in esophageal cancer is either N0 (no malignant lymph nodes) or N1 (lymphatic metastases to locoregional lymph nodes); that is, there is no N2 or N3 descriptor (23). Locoregional lymph nodes are those that represent the sites of primary lymphatic drainage from the esophagus and are normally resected with the primary tumor at the time of esophagectomy. According to the AJCC classification, the precise definition of a locoregional lymph node depends to some extent on the location of the primary tumor in the esophagus (cervical, intrathoracic, or gastroesophageal) (23).
For tumors in the cervical esophagus, locoregional lymph nodes include scalene, internal jugular, and supraclavicular lymph nodes. For tumors located in the thoracic esophagus, they include periesophageal and subcarinal lymph nodes. For tumors arising from the gastroesophageal junction, locoregional lymph nodes comprise lower periesophageal and pulmonary ligament lymph nodes, diaphragmatic lymph nodes (lying on the dome of the diaphragm or in the retrocrural regions), pericardial lymph nodes (located immediately adjacent to the gastroesophageal junction), left gastric lymph nodes, and celiac lymph nodes.
Lymphadenopathy outside these designated areas is regarded as M1a disease (nonregional lymph node metastases). The designation of nonregional lymph nodes as M1a is used because patients with metastases to nonregional lymph nodes have a much worse outcome than patients with involvement of locoregional lymph nodes only, but have a better long-term prognosis than patients with distant organ metastases (M1b). An important point is that lymph node metastases in nonregional or distant lymph nodes without involvement of intervening locoregional lymph nodes have been reported to occur in 25% of cases (32,33).
The nomenclature used for the designation of nonregional lymph nodes depends on the anatomic location of the primary tumor. For example, cervical lymphadenopathy is designated an M1a descriptor when the primary tumor is located in the upper thoracic esophagus; for tumors located more distally, an M1b descriptor is assigned. Similarly, celiac lymphadenopathy is classified as M1a disease when the primary tumor is located in the distal thoracic esophagus but as M1b disease when the tumor is located more proximally. For tumors of the midesophagus, both cervical and supraclavicular lymphadenopathy is considered M1b disease. (Owing to the limited lymphatic drainage from the midesophagus, metastases to nonregional lymph nodes are associated with a very poor prognosis equivalent to that of distant organ metastases, and the M1a descriptor is not used.)
For tumors occurring in the lower thoracic esophagus, the lymphatic drainage follows the blood supply of the left gastric artery to left gastric lymph nodes and lymph nodes in the gastrohepatic ligament (common hepatic and splenic lymph nodes) (34). In these cases, it is important at CT and PET/CT to differentiate between enlarged or FDG-avid lymph nodes in the gastrohepatic ligament (which are still resectable) and celiac lymph nodes (which are defined as nonregional lymph nodes or M1a disease and which normally indicate unresectability) (16,23,35,36). Owing to the close anatomic relationship of the liver, stomach, gastrohepatic ligament, and celiac axis in the upper abdomen, lymph nodes located in the caudal portion of the gastrohepatic ligament may appear to be close to the celiac axis on axial images (Fig 2); however, in the coronal plane they are located slightly more cephalad and more anteriorly than the origin of the celiac artery and can normally be identified adjacent to branches of the celiac axis (left gastric artery, hepatic artery, splenic artery) rather than adjacent to the celiac artery itself.
View larger version (69K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. N1 metastatic disease manifesting as FDG-avid left gastric lymphadenopathy. (a) Coronal maximum intensity projection (MIP) PET image shows an FDG-avid primary tumor at the gastroesophageal junction (short arrow) and an FDG-avid lymph node in the upper abdomen (long arrow). (b) Axial contrast-enhanced CT image shows the lymphadenopathy (arrow) between the celiac artery inferiorly (not shown) and the left gastric artery laterally (arrowhead). Differentiation between celiac and left gastric lymphadenopathy is difficult on axial images. (c) Coronal oblique MIP CT image shows the close anatomic relationship between the lymphadenopathy (arrow), celiac artery (C), left gastric artery (arrowhead), and stomach (S). (d) Sagittal oblique MIP CT image shows the lymphadenopathy (arrows), which is adjacent to the left gastric artery (arrowhead) and superior to the celiac artery (C). Therefore, the lymphadenopathy is localized to the left gastric region (a finding indicative of N1 disease) rather than the celiac region (M1a disease).
|
|
View larger version (44K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. N1 metastatic disease manifesting as FDG-avid left gastric lymphadenopathy. (a) Coronal maximum intensity projection (MIP) PET image shows an FDG-avid primary tumor at the gastroesophageal junction (short arrow) and an FDG-avid lymph node in the upper abdomen (long arrow). (b) Axial contrast-enhanced CT image shows the lymphadenopathy (arrow) between the celiac artery inferiorly (not shown) and the left gastric artery laterally (arrowhead). Differentiation between celiac and left gastric lymphadenopathy is difficult on axial images. (c) Coronal oblique MIP CT image shows the close anatomic relationship between the lymphadenopathy (arrow), celiac artery (C), left gastric artery (arrowhead), and stomach (S). (d) Sagittal oblique MIP CT image shows the lymphadenopathy (arrows), which is adjacent to the left gastric artery (arrowhead) and superior to the celiac artery (C). Therefore, the lymphadenopathy is localized to the left gastric region (a finding indicative of N1 disease) rather than the celiac region (M1a disease).
|
|
View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2c. N1 metastatic disease manifesting as FDG-avid left gastric lymphadenopathy. (a) Coronal maximum intensity projection (MIP) PET image shows an FDG-avid primary tumor at the gastroesophageal junction (short arrow) and an FDG-avid lymph node in the upper abdomen (long arrow). (b) Axial contrast-enhanced CT image shows the lymphadenopathy (arrow) between the celiac artery inferiorly (not shown) and the left gastric artery laterally (arrowhead). Differentiation between celiac and left gastric lymphadenopathy is difficult on axial images. (c) Coronal oblique MIP CT image shows the close anatomic relationship between the lymphadenopathy (arrow), celiac artery (C), left gastric artery (arrowhead), and stomach (S). (d) Sagittal oblique MIP CT image shows the lymphadenopathy (arrows), which is adjacent to the left gastric artery (arrowhead) and superior to the celiac artery (C). Therefore, the lymphadenopathy is localized to the left gastric region (a finding indicative of N1 disease) rather than the celiac region (M1a disease).
|
|
View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2d. N1 metastatic disease manifesting as FDG-avid left gastric lymphadenopathy. (a) Coronal maximum intensity projection (MIP) PET image shows an FDG-avid primary tumor at the gastroesophageal junction (short arrow) and an FDG-avid lymph node in the upper abdomen (long arrow). (b) Axial contrast-enhanced CT image shows the lymphadenopathy (arrow) between the celiac artery inferiorly (not shown) and the left gastric artery laterally (arrowhead). Differentiation between celiac and left gastric lymphadenopathy is difficult on axial images. (c) Coronal oblique MIP CT image shows the close anatomic relationship between the lymphadenopathy (arrow), celiac artery (C), left gastric artery (arrowhead), and stomach (S). (d) Sagittal oblique MIP CT image shows the lymphadenopathy (arrows), which is adjacent to the left gastric artery (arrowhead) and superior to the celiac artery (C). Therefore, the lymphadenopathy is localized to the left gastric region (a finding indicative of N1 disease) rather than the celiac region (M1a disease).
|
|
In our clinical experience and in agreement with other investigators, we have observed that difficulties in making this differentiation can occur with both PET/CT and endoscopic US (28,37). In cases of uncertainty, it may be useful to obtain additional reformatted CT images in the oblique coronal and sagittal planes to analyze the exact anatomic relationships of the suspected lymphadenopathy (Fig 2). In clinical practice, nodal status should be determined by using a multidisciplinary approach that includes information from both endoscopic US and PET/CT.
|
Interpretative Pitfalls in Determination of N Classification with PET/CT
|
|---|
PET/CT has relatively limited utility for detection of metastatic dissemination to locoregional lymph nodes (37–40). FDG uptake within periesophageal lymph nodes that are anatomically close to the primary tumor is difficult to differentiate from uptake within the esophagus itself owing to the limited spatial resolution of PET (Fig 3) (39). Furthermore, microscopic metastatic disease within lymph nodes may not demonstrate sufficient FDG uptake for detection with PET (37). In addition, FDG uptake within lymph nodes can occur in benign disease such as granulomatous infection (particularly in regions of endemic histoplasmosis or tuberculosis) or sarcoidosis(41–44) (Fig 4). In many cases, a confident interpretation of benign disease is not possible.
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. N1 metastatic disease manifesting as periesophageal lymphadenopathy. (a) Axial contrast-enhanced CT image obtained at the level of the subcarinal region shows marked esophageal wall thickening (arrowhead), which corresponds to a midesophageal adenocarcinoma. In addition, there is an enlarged local periesophageal lymph node (arrow), a finding suggestive of N1 disease. (b) Axial fused PET/CT image obtained at the same level shows intense FDG uptake in the primary tumor (arrowhead), but this activity cannot be resolved from uptake in the adjacent periesophageal lymph node. Therefore, the nodal status of this tumor remained indeterminate at PET/CT.
|
|
View larger version (103K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. N1 metastatic disease manifesting as periesophageal lymphadenopathy. (a) Axial contrast-enhanced CT image obtained at the level of the subcarinal region shows marked esophageal wall thickening (arrowhead), which corresponds to a midesophageal adenocarcinoma. In addition, there is an enlarged local periesophageal lymph node (arrow), a finding suggestive of N1 disease. (b) Axial fused PET/CT image obtained at the same level shows intense FDG uptake in the primary tumor (arrowhead), but this activity cannot be resolved from uptake in the adjacent periesophageal lymph node. Therefore, the nodal status of this tumor remained indeterminate at PET/CT.
|
|
View larger version (80K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. False-positive PET/CT results due to sarcoidosis. (a) Coronal MIP PET image shows an FDG-avid primary tumor in the distal esophagus (arrowhead) and several distant foci of FDG uptake in the axillae and right inguinal region (arrows). (b) Axial CT image shows several small lymph nodes in the mediastinum and axilla (arrows). (c) On an axial fused PET/CT image, the small lymph nodes appear as foci of increased FDG uptake, findings suggestive of distant lymph node metastases. Core biopsy of one of the inguinal lymph nodes demonstrated noncaseating granulomas, a finding consistent with sarcoidosis. The patient subsequently underwent Ivor Lewis esophagectomy.
|
|
View larger version (44K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. False-positive PET/CT results due to sarcoidosis. (a) Coronal MIP PET image shows an FDG-avid primary tumor in the distal esophagus (arrowhead) and several distant foci of FDG uptake in the axillae and right inguinal region (arrows). (b) Axial CT image shows several small lymph nodes in the mediastinum and axilla (arrows). (c) On an axial fused PET/CT image, the small lymph nodes appear as foci of increased FDG uptake, findings suggestive of distant lymph node metastases. Core biopsy of one of the inguinal lymph nodes demonstrated noncaseating granulomas, a finding consistent with sarcoidosis. The patient subsequently underwent Ivor Lewis esophagectomy.
|
|
View larger version (98K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c. False-positive PET/CT results due to sarcoidosis. (a) Coronal MIP PET image shows an FDG-avid primary tumor in the distal esophagus (arrowhead) and several distant foci of FDG uptake in the axillae and right inguinal region (arrows). (b) Axial CT image shows several small lymph nodes in the mediastinum and axilla (arrows). (c) On an axial fused PET/CT image, the small lymph nodes appear as foci of increased FDG uptake, findings suggestive of distant lymph node metastases. Core biopsy of one of the inguinal lymph nodes demonstrated noncaseating granulomas, a finding consistent with sarcoidosis. The patient subsequently underwent Ivor Lewis esophagectomy.
|
|
In a recent meta-analysis of 12 studies that examined the diagnostic accuracy of PET/CT in preoperative staging of esophageal cancer, PET/CT had a sensitivity of 51% and specificity of 84% for detection of nodal metastases (38). Endoscopic US is superior to PET/CT in detection of locoregional nodal metastases and is the primary modality used in this regard (45); in one recent prospective study, the combination of endoscopic US and CT had a reported sensitivity of 83% for local nodal metastases (vs 22% for PET, P = .0026). Endoscopic US has the added advantage over PET/CT of allowing biopsy of suspicious lymph nodes at the time of detection.
|
M Stage
|
|---|
Metastatic disease is present in 20%–30% of patients with esophageal cancer at initial evaluation (2). According to the AJCC classification, metastatic disease is subdivided into M1a and M1b, where M1a refers to metastases to nonregional lymph nodes (as discussed earlier) and M1b indicates distant organ metastases (23). Both epidemiologic and pathologic studies have shown that patients with M1a disease have a slightly better prognosis than patients with M1b disease (26,46). Patients with M1a disease who have a response to neoadjuvant therapy may still be candidates for esophagectomy and are potentially curable, but M1b signifies incurable disease for which esophagectomy is not indicated.
PET/CT allows detection of metastatic disease that may not be identifiable with other methods (Fig 5). PET/CT has been shown to improve preoperative staging and prevent inappropriate surgery when used in combination with conventional work-up (13,15,16). Even in patients who are not suitable for esophagectomy, the detection of unsuspected metastases can help guide palliative management.
View larger version (51K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Distant metastatic disease that was occult at conventional imaging. (a) Axial contrast-enhanced CT image shows a large soft-tissue mass at the gastroesophageal junction (arrowhead), a finding consistent with a locally advanced esophageal adenocarcinoma. There was no evidence of distant metastases at conventional imaging. (b, c) Axial (b) and coronal (c) fused PET/CT images show intense FDG uptake by the primary tumor (arrowhead) and an unexpected additional focus of FDG uptake in the liver (arrow). The additional focus of uptake was confirmed to be a hepatic metastasis with magnetic resonance (MR) imaging.
|
|
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Distant metastatic disease that was occult at conventional imaging. (a) Axial contrast-enhanced CT image shows a large soft-tissue mass at the gastroesophageal junction (arrowhead), a finding consistent with a locally advanced esophageal adenocarcinoma. There was no evidence of distant metastases at conventional imaging. (b, c) Axial (b) and coronal (c) fused PET/CT images show intense FDG uptake by the primary tumor (arrowhead) and an unexpected additional focus of FDG uptake in the liver (arrow). The additional focus of uptake was confirmed to be a hepatic metastasis with magnetic resonance (MR) imaging.
|
|
View larger version (71K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c. Distant metastatic disease that was occult at conventional imaging. (a) Axial contrast-enhanced CT image shows a large soft-tissue mass at the gastroesophageal junction (arrowhead), a finding consistent with a locally advanced esophageal adenocarcinoma. There was no evidence of distant metastases at conventional imaging. (b, c) Axial (b) and coronal (c) fused PET/CT images show intense FDG uptake by the primary tumor (arrowhead) and an unexpected additional focus of FDG uptake in the liver (arrow). The additional focus of uptake was confirmed to be a hepatic metastasis with magnetic resonance (MR) imaging.
|
|
|
Interpretative Pitfalls in Determination of M Classification with PET/CT
|
|---|
The commonest sites of visceral metastases (M1b) include the lungs, liver, bones, and adrenal glands (47,48). However, metastases from esophageal cancer can occur in unusual and unexpected locations and can be radiologically occult when traditional imaging methods such as CT are used for detection (12,49) (Fig 6). Uncommon sites of organ metastases include the brain, skeletal muscle, subcutaneous tissues, thyroid gland, and pancreas (49) (Figs 7, 8). In some cases, these metastases can simulate more benign entities such as injection granulomas, thyroid adenomas, or pancreatic pseudocysts (49). Accordingly, when interpreting initial staging PET/CT scans in patients with esophageal cancer, it is important to know the most common sites of metastatic disease, but also to be aware of the possibility of the more uncommon sites of distant metastases.
View larger version (69K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Distant metastatic disease that was occult at conventional imaging after neoadjuvant chemotherapy and radiation therapy. (a) Coronal MIP PET image shows diffuse FDG uptake throughout the esophagus (arrowhead), a finding consistent with posttreatment esophagitis. In addition, there are two unexpected foci of FDG uptake (arrows), findings suggestive of new distant metastases. (b) Axial CT image obtained at the level of the left axilla shows no anatomic abnormality. (c) Axial fused PET/CT image shows a hypermetabolic lesion in the left supraspinatus muscle (arrow). Subsequent biopsy demonstrated an intramuscular metastasis. The second focus of increased FDG uptake corresponded to a metastasis in the right gluteus medius muscle.
|
|
View larger version (41K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Distant metastatic disease that was occult at conventional imaging after neoadjuvant chemotherapy and radiation therapy. (a) Coronal MIP PET image shows diffuse FDG uptake throughout the esophagus (arrowhead), a finding consistent with posttreatment esophagitis. In addition, there are two unexpected foci of FDG uptake (arrows), findings suggestive of new distant metastases. (b) Axial CT image obtained at the level of the left axilla shows no anatomic abnormality. (c) Axial fused PET/CT image shows a hypermetabolic lesion in the left supraspinatus muscle (arrow). Subsequent biopsy demonstrated an intramuscular metastasis. The second focus of increased FDG uptake corresponded to a metastasis in the right gluteus medius muscle.
|
|
View larger version (47K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6c. Distant metastatic disease that was occult at conventional imaging after neoadjuvant chemotherapy and radiation therapy. (a) Coronal MIP PET image shows diffuse FDG uptake throughout the esophagus (arrowhead), a finding consistent with posttreatment esophagitis. In addition, there are two unexpected foci of FDG uptake (arrows), findings suggestive of new distant metastases. (b) Axial CT image obtained at the level of the left axilla shows no anatomic abnormality. (c) Axial fused PET/CT image shows a hypermetabolic lesion in the left supraspinatus muscle (arrow). Subsequent biopsy demonstrated an intramuscular metastasis. The second focus of increased FDG uptake corresponded to a metastasis in the right gluteus medius muscle.
|
|
View larger version (40K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. Unsuspected brain metastasis at initial staging in a patient with esophageal adenocarcinoma. (a) Axial unenhanced CT image shows a partially calcified lesion in the left lobe of the cerebellum (arrow). (b, c) On axial PET (b) and fused PET/CT (c) images, the lesion (arrow) appears as a subtle photopenic area. It is difficult to identify due to the increased FDG uptake in the surrounding cerebral tissue. (d) Axial T1-weighted MR image obtained after administration of gadolinium contrast material shows the metastasis (arrow).
|
|
View larger version (37K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. Unsuspected brain metastasis at initial staging in a patient with esophageal adenocarcinoma. (a) Axial unenhanced CT image shows a partially calcified lesion in the left lobe of the cerebellum (arrow). (b, c) On axial PET (b) and fused PET/CT (c) images, the lesion (arrow) appears as a subtle photopenic area. It is difficult to identify due to the increased FDG uptake in the surrounding cerebral tissue. (d) Axial T1-weighted MR image obtained after administration of gadolinium contrast material shows the metastasis (arrow).
|
|
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7c. Unsuspected brain metastasis at initial staging in a patient with esophageal adenocarcinoma. (a) Axial unenhanced CT image shows a partially calcified lesion in the left lobe of the cerebellum (arrow). (b, c) On axial PET (b) and fused PET/CT (c) images, the lesion (arrow) appears as a subtle photopenic area. It is difficult to identify due to the increased FDG uptake in the surrounding cerebral tissue. (d) Axial T1-weighted MR image obtained after administration of gadolinium contrast material shows the metastasis (arrow).
|
|
View larger version (52K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7d. Unsuspected brain metastasis at initial staging in a patient with esophageal adenocarcinoma. (a) Axial unenhanced CT image shows a partially calcified lesion in the left lobe of the cerebellum (arrow). (b, c) On axial PET (b) and fused PET/CT (c) images, the lesion (arrow) appears as a subtle photopenic area. It is difficult to identify due to the increased FDG uptake in the surrounding cerebral tissue. (d) Axial T1-weighted MR image obtained after administration of gadolinium contrast material shows the metastasis (arrow).
|
|
View larger version (35K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. Unsuspected intramuscular metastases at initial staging in a patient with potentially resectable esophageal adenocarcinoma. (a) Axial unenhanced CT image obtained at the level of the aortic arch shows no anatomic abnormality. (b) Axial fused PET/CT image obtained at the same level shows several foci of abnormal FDG uptake in skeletal muscles (arrows). Subsequent evaluation showed that these foci represented intramuscular metastases.
|
|
View larger version (72K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. Unsuspected intramuscular metastases at initial staging in a patient with potentially resectable esophageal adenocarcinoma. (a) Axial unenhanced CT image obtained at the level of the aortic arch shows no anatomic abnormality. (b) Axial fused PET/CT image obtained at the same level shows several foci of abnormal FDG uptake in skeletal muscles (arrows). Subsequent evaluation showed that these foci represented intramuscular metastases.
|
|
In addition, PET/CT occasionally allows detection of unsuspected synchronous neoplasms. Synchronous neoplastic disease can be present in 1.5%–5.5% of patients with esophageal cancer at initial presentation, most commonly in the stomach, head and neck, and colon (49–51). In patients with esophageal cancer, any unusual foci of FDG uptake at PET/CT should be regarded with suspicion and should prompt further evaluation to exclude neoplastic disease (Fig 9).
View larger version (82K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. Unexpected synchronous neoplasm in a patient with newly diagnosed adenocarcinoma of the distal esophagus. (a) Coronal MIP PET image shows intense focal FDG uptake by the primary mass (arrow) as well as by a lesion in the pelvis (arrowhead). (b) Axial fused PET/CT image shows the hypermetabolic mass in the sigmoid colon (arrowhead). (c) Axial contrast-enhanced CT image obtained at the same level shows the large colonic mass (arrowhead). A primary adenocarcinoma of the sigmoid colon was found at subsequent sigmoidoscopy and was resected.
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. Unexpected synchronous neoplasm in a patient with newly diagnosed adenocarcinoma of the distal esophagus. (a) Coronal MIP PET image shows intense focal FDG uptake by the primary mass (arrow) as well as by a lesion in the pelvis (arrowhead). (b) Axial fused PET/CT image shows the hypermetabolic mass in the sigmoid colon (arrowhead). (c) Axial contrast-enhanced CT image obtained at the same level shows the large colonic mass (arrowhead). A primary adenocarcinoma of the sigmoid colon was found at subsequent sigmoidoscopy and was resected.
|
|
View larger version (49K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9c. Unexpected synchronous neoplasm in a patient with newly diagnosed adenocarcinoma of the distal esophagus. (a) Coronal MIP PET image shows intense focal FDG uptake by the primary mass (arrow) as well as by a lesion in the pelvis (arrowhead). (b) Axial fused PET/CT image shows the hypermetabolic mass in the sigmoid colon (arrowhead). (c) Axial contrast-enhanced CT image obtained at the same level shows the large colonic mass (arrowhead). A primary adenocarcinoma of the sigmoid colon was found at subsequent sigmoidoscopy and was resected.
|
|
|
Assessment of Treatment Response
|
|---|
Patients with locally advanced esophageal cancer who are suitable for esophagectomy may have a prolonged long-term survival if they are first treated with chemotherapy and radiation therapy (3–5,10,52,53). Two recent meta-analyses confirmed an improved 3-year survival in patients who received neoadjuvant chemotherapy and radiation therapy over patients who underwent surgery alone, in addition to a higher chance of complete resection and a lower chance of local recurrence after completion of therapy (54,55). Patients who are treated with such multimodality therapy will benefit most from esophagectomy if they respond to the neoadjuvant therapy (3–8); for those patients who do not respond but nevertheless undergo esophagectomy, the long-term survival is poorer than if they had proceeded to surgery de novo (5,8,56). Consequently, the assessment of therapeutic response is important for deciding on the most appropriate form of definitive treatment.
Assessment of therapeutic response has traditionally been performed with CT and with endoscopic gastroduodenoscopy and endoscopic US (19,20). However, the extensive tumor necrosis and fibrosis that follow chemotherapy and radiation therapy have made it difficult to evaluate the extent of residual malignancy by using these methods; a recent meta-analysis by Westerterp et al (20) reported maximum joint sensitivity and specificity (Q point) values of 54% for diagnostic CT and 86% for endoscopic US. Recently, there has been increasing interest in use of PET/CT for assessing treatment response (19–21,57,58). A pathologic response within thetumor has been reported to correspond to decreases in SUVmax within the primary tumor of 35%–60% between initial staging PET scans and reevaluation imaging (21,57–61).
However, to date there have been mixed results in use of PET for assessment of treatment response. Some investigators have reported that a single posttreatment PET/CT scan is sufficient for assessing therapeutic response, by showing that persistent FDG uptake within the primary tumor site (with SUVmax 
4) correlates with residual viable macroscopic tumor and poor survival after esophagectomy (18,19,62). However, more recent studies have shown that PET/CT assessment of therapeutic response is less optimal than previously reported (12,63), due in part to technical differences between PET and PET/CT, but also as a result of the many false-positive findings that may occur secondary to posttreatment esophagitis and ulceration (discussed further in the following section).
In most studies of treatment response in esophageal cancer (19,30,57,58,61,62), PET/CT has been performed after completion of chemotherapy and radiation therapy; however, it may be of greater clinical utility to perform PET/CT earlier in the course of therapy, so that neoadjuvant therapy can be "tailored" to achieve the maximum benefit. Weber et al (21) have shown that PET/CT performed after only two cycles of induction chemotherapy allows predicti
Top
Abstract
LEARNING OBJECTIVES
