Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print November 26, 2003, 10.1148/radiol.2301021047
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2301021047v1
230/1/281    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dave, U. R.
Right arrow Articles by deSouza, N. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dave, U. R.
Right arrow Articles by deSouza, N. M.
(Radiology 2004;230:281-286.)
© RSNA, 2004

Technical Developments

Esophageal Cancer Staging with Endoscopic MR Imaging: Pilot Study1

Umakant R. Dave, MD, Andreanna D. Williams, MD, Jason A. Wilson, MD, Zahir Amin, MD, David J. Gilderdale, MD, David J. Larkman, MD, Mark R. Thursz, MD, Simon D. Taylor-Robinson, MD and Nandita M. deSouza, MD

1 From the Robert Steiner MRI Unit (U.R.D., A.D.W., D.J.G., D.J.L., S.D.T.R., N.M.d.S.), Departments of Gastroenterology (U.R.D., S.D.T.R., N.M.d.S.) and Anesthesia (J.A.W.), Hammersmith Hospitals Trust, DuCane Rd, London W12 0HS, England; Division of Medicine, Imperial College, London, England (M.R.T.); and Department of Radiology, Chelsea & Westminster Hospital, London, England (Z.A.). Received September 15, 2002; revision requested November 6; final revision received May 15, 2003; accepted May 19. MEDLINK funding enabled the design and production of the MR-compatible endoscope and the coil. Marconi Medical Systems provided further financial support and technical assistance. Address correspondence to N.M.d.S. (e-mail: ndesouza@hhnt.org).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
The authors defined esophageal anatomy and evaluated esophageal cancer staging in a pilot group by comparing endoscopic magnetic resonance (MR) imaging results with pathologic and endoscopic ultrasonographic (US) results when available. A porcine esophagus, one volunteer, and 23 patients suspected of having esophageal cancer were imaged at 0.5 T. MR imaging was successful in 21 patients. Eight of these patients underwent esophagectomy (one after chemotherapy, which invalidated comparison with MR imaging; another did not undergo lymphadenectomy) and one underwent laparoscopy and nodal staging only; eight underwent US. When verified with pathologic staging, endoscopic MR imaging was accurate in six of seven patients (T stage) and five of six patients (N stage; nodal areas too obscured by artifact for comparison in one case). MR imaging and US results concurred in seven of eight (T stage) and five of eight (N stage) patients. No complications were observed. Endoscopic MR imaging is safe and probably comparable to endoscopic US, but with a tendency to overstage the disease.

© RSNA, 2003

Index terms: Esophagus, MR, 71.121411 • Esophagus, neoplasms, 71.32, 71.33 • Head and neck neoplasms, staging, 71.32, 71.33 • Magnetic resonance (MR), endoscopic, 71.121411 • Ultrasound (US), endoscopic, 71.12981


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and Methods
Results
Discussion
 REFERENCES
 
Distal esophageal and cardial cancer is the most rapidly rising and lethal type of gastrointestinal malignancy, with a 4%–10% increase in incidence per annum in men (1) and a 5-year survival rate of only 10% (2). Accurate tumor staging is critical in the selection of therapy and prediction of prognosis (3,4). Standard imaging techniques such as computed tomography (CT) lack adequate image resolution for staging the primary site (5). Ultrasound probes used in conjunction with endoscopy achieve higher frequency, and endoscopic ultrasonography (US) is currently the most accurate clinical tool available for staging esophageal cancer; reported accuracies are 61%–92% for T stage and 75%–85% for N stage (6,7). However, low soft-tissue contrast between the tumor and muscularis propria limits the application of this technique.

Magnetic resonance (MR) imaging provides superior soft-tissue contrast enhancement, but with conventional external whole-body or phased-array receiver coils, the image resolution of small areas of interest is generally low. Staging accuracies with MR imaging by using conventional external surface coils are comparable to those with CT imaging; reported values vary from 50% to 85% (8,9) without added advantage. Dedicated surface receiver coils can greatly improve the image resolution of a localized region of interest, and endocavitary coils for imaging of the prostate, uterine cervix, and anal sphincter are now routinely used (10–12). Receiver coils integrated with an MR-compatible gastroscope have demonstrated the feasibility of endoscopic MR imaging (13,14), and comparison with endoscopic US in a prospective pilot study has shown that findings are comparable (15). However, to our knowledge, there are no data available to compare results at endoscopic MR imaging with results at pathologic tumor staging. The purpose of this study was to define the anatomy of the esophagus on endoscopic MR images (ex vivo and in vivo) and subsequently evaluate esophageal cancer staging with endoscopic MR imaging in a pilot group of patients by comparing the findings with those from pathologic staging at surgery and from endoscopic US when available.


   Materials and Methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
In Vitro Porcine Studies
This study was approved by the local research ethics committee (Hammersmith Hospital). In three 45-kg large white pigs, sacrificed for independent experimental reasons, a section of normal midesophagus was resected and placed in saline at room temperature. These sections were imaged between 45 minutes and 41/2 hours after the animals’ deaths. Sterility issues precluded the use of the same coil for ex vivo animal studies and in vivo human studies, so a coil similar in geometry and signal characteristics to that used in the human endoscopic MR imaging studies (12-mm diameter saddle geometry cylinder) was used. The coil was inserted into the esophageal lumen, and T1- and T2-weighted transverse MR images were obtained from the midesophagus section with a 192 x 256 matrix, 4-mm section thickness, and 12-cm field of view. The tissue was then fixed and sectioned for histologic examination.

In Vivo Studies
The study complied with the guidelines set out in the 1975 Helsinki Declaration on Human Rights and was approved by the local research ethics committee. All patients and the volunteer gave written informed consent.

Volunteer Study
A 50-year-old healthy volunteer was intubated with a 12-mm-diameter saddle geometry surface receiver coil. T1-weighted spin-echo transverse (repetition time msec/echo time msec, 400/20) MR images were obtained with cardiac gating.

Patients
Twenty-three patients who consecutively agreed to participate in the study (12 men and 11 women; mean age, 65 years; age range, 31–79 years) were recruited; these patients were confirmed to have or were suspected of having esophageal cancer at diagnostic endoscopy and histopathologic examination. All patients had undergone CT scanning at the referring hospital as part of their clinical work-up. Final histopathologic results showed adenocarcinoma (n = 15), squamous cell carcinoma (n = 5), low-grade dysplasia without invasive cancer (n = 1), benign stricture (n = 1), and glycogenic acanthosis (n = 1). Patients who were unable to lie flat or who had contraindications to MR imaging, such as ferrous implants or claustrophobia, were excluded.

Endoscopic MR Imaging
A forward-viewing endoscope that was constructed from nonferrous material (16) and had a 5-m-long cable connection to keep the light source away from the magnetic field was used. The coil system comprised two components: the receive coil (12-mm-diameter saddle geometry), which was secured mechanically to and was detachable from the endoscope’s distal tip, and the pick-up coil, which was embedded within the distal tip (Fig 1). The coils were coupled by magnetic induction and required no direct electric connection. The coil system was designed and built for this study and is not currently available commercially. The external cable connection was made from the proximal (operator’s) end of the endoscope to the receive channel of the MR imager. Active detuning was achieved by switching the pick-up coil to couple resistance into the receiver coil during MR radiofrequency excitation. During MR receive operation, the coil output impedance was matched to 50 {Omega} by using the output cable as a transmission line transformer.



View larger version (133K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Photograph shows distal end of MR-compatible endoscope with integral coil system. The outer saddle geometry receive coil is detachable and is inductively coupled to the pick-up coil embedded within the endoscope tip.

 
Sedation was performed by using a combination of propafol (80–350 mg) and midazolam (2–8 mg) in 17 patients, a combination of pethidine (20–25 mg) and midazolam (3–8 mg) in five patients, and midazolam only (6 mg) in one patient. Patients in this study required additional sedation compared with our standard endoscopic practice, in which up to 5 mg of midazolam is used, because of the length of the examination. O2 was administered to all patients through nasal cannulae, and in no instance was airway assistance required.

The first six patients underwent endoscopic MR imaging with the internal receive coil only; in the subsequent 17 patients, an external surface coil was placed posteriorly on the patient’s chest at the level of the endoscopic coil, and images were obtained by using the combined internal-external array. Imaging was performed with a 0.5-T MR imager (Apollo; Marconi Medical Systems, Highland Heights, Ohio). T1-weighted spin-echo (545–1,410/20) and T2-weighted fast spin-echo (repetition time msec/echo time [effective] msec, 1,220–2,820/80) MR images were acquired with cardiac gating. Multiple (six to 12) sections in a plane transverse to the long axis of the coil, with a 4-mm section thickness and a 20.0–22.5-cm field of view, were obtained in one probe location only in order to keep the length of examination time acceptable. Mean image acquisition time was 37 minutes (range, 26–60 minutes), and mean time for the entire procedure was 45 minutes (range, 40–50 minutes). A note was made if the endoscope did not pass through the tumor, and in these cases, endoscopic MR images were obtained with the coil proximal to the lesion. In seven patients with large lesions, an external wrap-around array coil was used to obtain fast spin-echo MR images of the chest in three planes immediately after the endoscopic MR imaging examination to evaluate the mediastinum fully for the presence of lymphadenopathy.

Before patients were discharged, they were asked to quantify their discomfort on a semiquantitative scale (scale of 1–10) to compare the discomfort of endoscopic MR imaging with the discomfort of their previous standard endoscopy and were asked whether they would be willing to undergo a repeat endoscopic MR imaging if clinically indicated. Any complications related to endoscopic MR imaging were recorded immediately after the procedure (immediate complications) and 1 week later (late complications) by two authors in consensus (U.R.D. and A.D.W.).

Endoscopic US
Endoscopic US was performed on the basis of an individual patient’s clinical need and was attempted in 10 patients by using a radial scanning echoendoscope with a 13-mm diameter (GF-UM200 Evis; Olympus, Tokyo, Japan); this was performed by a single investigator (Z.A.). Endoscopic US was initially performed by using a 7.5-MHz probe, and then detailed higher resolution views of the esophageal wall and the surrounding structures were obtained by using a 12-MHz probe. The following medications were administered before endoscopic US: lignocaine throat spray (no sedation) in one patient, midazolam (3–5 mg) in six patients, and a combination of midazolam (4–8 mg) and pethidine (25–50 mg) in three patients.

Surgery
Nine patients underwent surgery. Eight underwent esophagectomy; one of these patients received neoadjuvant chemotherapy after endoscopic MR imaging and endoscopic US, and so imaging results were excluded from the comparison with histopathologic results. One patient with esophageal perforation underwent emergency esophagectomy without nodal staging. The remaining patient underwent only laparoscopy followed by radiation therapy, which allowed for pathologic nodal staging but not pathologic tumor staging. Staging was determined with both endoscopic MR imaging and endoscopic US in three of the nine patients who underwent surgery.

Image Analysis
On endoscopic MR images, lesions were assessed for depth of tumor infiltration (involving mucosa and submucosa only, T1; involving submucosa and muscularis propria, T2; breaching the full width of the muscularis propria but no extension beyond the esophageal wall, T3; and extension beyond esophageal wall to adjacent structures, T4) and the presence of lymph nodes. T1-weighted MR images were less susceptible to motion artifacts than were T2-weighted fast spin-echo MR images and were therefore used for staging the primary tumor. A single investigator (N.M.d.S.) evaluated the endocopic MR images and external-array MR images. Peritumoral lymph nodes were considered malignant if their maximum transverse diameter was greater than 5 mm or if they were hyperintense at T2-weighted MR imaging (17) or hypoechoic at endoscopic US and round and homogenous in structure (18).

Data Analysis
Accuracy rates of endoscopic MR imaging for the assessment of T stage and N stage were calculated by using the histopathologic examination of the operative specimen (when available) as a standard. Tumor staging was grouped in T1/T2 and T3/T4 stages because, for disease management and prognosis purposes, it is important to differentiate patients with T1 or T2 lesions from those with T3 or T4 lesions. In cases in which endoscopic US images were available, a comparison with endoscopic US was also performed.


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
Porcine ex Vivo Studies
Esophageal wall layers were identified as follows, and correlation was demonstrated between ex vivo MR imaging studies and histologic examinations (Fig 2).



View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Porcine esophagus in vitro. Transverse T1-weighted (355/20) (A) and T2-weighted (2,500/80) (B) MR images show four distinct wall layers that correspond to histopathologic examination of the transverse pathologic section in C. From inner to outer layers, these are high-signal-intensity (SI) mucus and epithelium (black arrowhead), low-SI muscularis mucosa (black arrow), high-SI submucosa (long white arrow), and low-SI muscularis propria (short white arrow). In addition, an outer layer with high SI (white arrowhead) is seen in A that corresponds to the fibrofatty adventitia in C (gray arrowhead).

 
Five layers were identified on the T1-weighted MR images (Fig 2, A), which corresponded with histologic results (Fig 2, C). The epithelium/mucus layer was innermost and demonstrated high SI. The mucularis mucosa showed low SI, the submucosa showed high SI, the striated muscularis propria showed low SI, and the outer high-SI layer corresponded with the fibrofatty adventitia. On the T2-weighted MR images (Fig 2, B), the fifth (outermost) high-SI layer, which corresponded with the fibrofatty adventitia, was difficult to visualize.

Human Studies
Three layers were seen in vivo in the normal esophagus: an inner mucosa/muscularis mucosa with intermediate SI, a submucosa with high SI, and a muscularis propria with low SI (Fig 3).



View larger version (171K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Transverse T1-weighted (840/20) MR image obtained through the midesophagus in a patient with normal esophageal wall anatomy. The innermost low-SI layer represents muscularis mucosa (long arrow). Around this layer is the thicker high-SI submucosa (short arrow), and the muscularis propria is the outermost low-SI band (arrowhead). E = endoscopic coil, L = lung, A = aorta, V = vertebral body. Note that the larger diameter of the esophagus compared with that of the coil results in collapsed "outpouching" of the esophageal wall on the left.

 
Endoscopic MR imaging was attempted in 23 patients suspected of having esophageal cancer, and intubation was successfully performed in 22 patients (intubation was unsuccessful in a patient with benign stricture who needed general anesthetic for routine endoscopy). In one patient with esophageal adenocarcinoma, a severe image artifact was obtained from an external source, so that data from 21 patients were analyzed. In nine patients, the endoscope would not pass through a stenotic lesion, so the coil was placed proximal to the lesion, which resulted in suboptimal image quality through the lesion. In four of seven patients undergoing external array imaging, lymph nodes were seen in the mediastinum. In all four patients with lymphadenopathy at external array imaging, lymph nodes were seen on endoscopic MR images.

Eight of 21 patients then underwent esophagectomy (in one case, after neoadjuvant chemotherapy; in another, a patient with esophageal perforation underwent emergency esophagectomy without nodal staging), two patients with no evidence of malignancy were treated conservatively, eight patients received chemotherapy and/or radiation therapy, and three patients underwent palliative therapy only. Three of the eight patients who underwent esophagectomy had T1 lesions, one had a T2 lesion, two had T3 lesions, and two had T4 lesions. Endoscopic MR imaging enabled correct staging of the primary tumor in five patients; one T1 lesion was overstaged as T3 (Table 1), and one node was also overstaged. In one patient, endoscopic MR imaging enabled correct identification of mediastinal nodes positive for cancer, which was confirmed at histopathologic examination. There was concordance between endoscopic MR imaging and histopathologic examination in six of seven patients (86%) for T stage (Fig 4) and in five of six (83%) for N stage.


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

  		TABLE 1. Concordance between Staging at Endoscopic MR Imaging and Pathologic Staging after Esophagectomy

 


View larger version (155K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a. (a) Transverse T1-weighted (840/20) MR image obtained in a patient with esophageal cancer and (b) corresponding histopathologic section. (Hematoxylin-eosin stain; original magnification, x40.) The tumor (T) is seen anteriorly in a, with a rim of intact muscularis propria (arrows) surrounding it. E = endoscopic coil, L = lung, A = aorta, V = vertebral body. In b, the tumor (T) has invaded the submucosa, while the muscularis propria (MP) is intact. There was good agreement between endoscopic MR imaging staging and pathologic staging of this tumor.

 


View larger version (180K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b. (a) Transverse T1-weighted (840/20) MR image obtained in a patient with esophageal cancer and (b) corresponding histopathologic section. (Hematoxylin-eosin stain; original magnification, x40.) The tumor (T) is seen anteriorly in a, with a rim of intact muscularis propria (arrows) surrounding it. E = endoscopic coil, L = lung, A = aorta, V = vertebral body. In b, the tumor (T) has invaded the submucosa, while the muscularis propria (MP) is intact. There was good agreement between endoscopic MR imaging staging and pathologic staging of this tumor.

 
Endoscopic US was attempted in 10 patients, and intubation was unsuccessful in one patient with adenocarcinoma. In one patient with benign stricture who successfully underwent endoscopic US with general anesthesia, intubation for endoscopic MR imaging failed, so comparative staging data from eight patients are available. Results of endoscopic MR imaging agreed with those of endoscopic US in seven of eight patients for T stage (Table 2). In one patient, endoscopic MR imaging caused the lesion to be overstaged in comparison with endoscopic US (from T2 to T3). In this patient, histopathologic results showed the lesion to be T1. In three of eight patients, endoscopic MR imaging failed to depict nodes seen at endoscopic US. Surgical histopathologic results were available only in one of these patients in whom no lymphadenopathy was identified. However, this patient had received neoadjuvant chemotherapy, which may have resulted in downstaging of the tumor. There was concordance between results of endoscopic MR imaging and those of endoscopic US in seven of eight patients (87%) for T stage and in five of eight (63%) for N stage.


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

  		TABLE 2. Concordance between Staging at Endoscopic MR Imaging and Endoscopic US

 
No immediate or late complications were observed in relation to the procedure. Nineteen patients described minimal discomfort (1 on a scale of 1–10) from the procedure, and the remaining two patients described moderate discomfort (3 on a scale of 1–10). All patients considered the discomfort of endoscopic MR imaging equal to or less than that of standard endoscopy (however, the level of sedation was much greater during endoscopic MR imaging), and all patients agreed to undergo further endoscopic MR imaging if clinically indicated.


   Discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
REFERENCES
 
This study demonstrates the potential of using endoscopic MR imaging for staging esophageal cancer: Endoscopic MR imaging findings compared well with histopathologic findings, with 86% concordance for T stage and 83% concordance for N stage with full staging correct in 63%. Also, there was 87% concordance between results of endoscopic MR imaging and those of endoscopic US for T1/T2 and T3/T4 stages. Since only three patients underwent both endoscopic US and surgery, it is not possible to comment on a comparison between endoscopic US findings and pathologic staging. Like endoscopic US, endoscopic MR imaging with currently available imaging sequences at 0.5 T tends to cause overstaging of the lesion. Improvements in imaging time and the increased use of T2-weighted sequences to improve the contrast between the layers of the esophageal wall should greatly improve the staging accuracy of this technique. Also, although in this pilot study images were obtained in one probe location only, with faster imaging techniques the option of acquiring data at multiple probe locations will be possible.

Clear delineation and identification of the mural layers are crucial for staging accuracy. Results of high-spatial-resolution MR imaging studies of esophageal cancer ex vivo at 4.7 T and subsequent correlation with histologic findings have revealed eight layers in the normal esophageal wall (19). On T2-weighted MR images, the mucosa was recognized as three layers: epithelium (low SI), lamina propria mucosae (high SI), and muscularis mucosae (low SI) (19). The submucosa had high SI and included loose connective tissue, capillary vessels, and lymphatics at histopathologic examination. The inner circular and outer longitudinal muscle layers were seen as discrete low-SI structures separated by a thin band of high SI that corresponded with loose connective tissue. The adventitia had high SI. This degree of soft-tissue contrast enhancement was not present on T1-weighted images. In our in vitro study at 0.5 T, five layers were identified on the T1-weighted MR images, which corresponded with histologic findings. However, the lower resolution of the in vivo images (lower matrix size and larger field of view), together with some motion artifact, made it difficult to separate the mucosa and muscularis mucosa on the T1-weighted MR images. Unfortunately, our in vivo T2-weighted images were prone to motion artifact, and we relied on T1-weighted images for in vivo staging with endoscopic MR imaging. Improvements in imaging time and pulse sequence parameters should enable more robust contrast enhancement on T2-weighted MR images in the future.

Kulling and colleagues (15) prospectively studied the role of endoscopic MR imaging in staging cancer of the esophagus (n = 9) or gastroesophageal junction (n = 6). Results of endoscopic MR imaging agreed with those of endoscopic US for T stage in 11 of 15 cases and for N stage in 12 of 15 cases, and the authors concluded that endoscopic MR imaging was comparable to endoscopic US. Endoscopic MR imaging was inadequate in four cases as a result of motion artifacts. Kulling and colleagues did not report surgical or pathologic confirmation of their imaging findings. Our study results compare endoscopic MR imaging prospectively with pathologic staging.

The importance of imaging in esophageal cancer staging has been demonstrated in a retrospective study of 203 patients in whom pretreatment endoscopic US was shown to enable prediction of survival on the basis of the initial T stage and N stage—in particular, on the basis of the presence of celiac axis lymph nodes. The presence of nodes was an important predictor of survival in multifactorial analysis (4). However, others have shown that while endoscopic US is effective for discrimination of stages T1 and T2 from stages T3 and T4 in gastroesophageal carcinomas, it has limited utility in local nodal staging (20). In this respect, endoscopic MR imaging offers potential advantage over endoscopic US: It may be coupled to an external coil or coils to improve the useful field of view (depicting nodes at greater distance from the esophagus), and it may be combined with a conventional external-coil examination to assess the T stage and N stage, as well as the metastatic status of cancer. The limited depth of penetration of endoscopic US makes it unsuitable as an imaging modality for the exclusion of distant metastasis, because the whole of the liver, the lungs, and more distant lymph node groups cannot be visualized. We conducted conventional external-coil MR imaging in seven patients after endoscopic MR imaging to assess the entire mediastinum.

Limitations of endoscopic US also result from its inability to traverse a stenotic cancer and lower performance at the cardia compared with the rest of the esophagus (20). These limitations also apply to endoscopic MR imaging. In addition, endoscopic US has demonstrated poor staging accuracy after neoadjuvant therapy (21). The accuracy of preoperative endoscopic US performed after chemotherapy or radiation therapy was only 37% for T stage, and we were unable to use it to distinguish radiation fibrosis and inflammation from residual tumor. MR imaging is less operator dependent and allows contrast enhancement. For these reasons endoscopic MR imaging may prove useful in the neoadjuvant setting, and this needs formal assessment.

Like endoscopic US, endoscopic MR imaging offers the advantages of optical visualization combined with cross-sectional imaging. For lesions accessible with the endoscope, the receiver coil can be placed immediately in contact with the region of interest to provide very high spatial resolution MR images of this area. In the thorax, image degradation due to coil motion arises mainly from the heart and great vessels and from respiration. The very high signal obtained at the surface of small internal receiver coils is "smeared" across the image as a result of motion. Cardiac gating of the images is therefore always necessary. Gross movement of the patient is an additional factor if imaging times are too long. In these examinations in a sedated patient, breath-hold techniques are not possible and single-section fast imaging techniques may be further used to reduce artifact by "freezing" motion. Gradient-echo techniques result in artifact from local field perturbations due to susceptibility mismatches between air, or the receiver coil itself, and the surrounding tissue. Single-shot fast spin-echo techniques are more immune to such artifacts, but the small fields of view and high spatial resolution required necessitate extremely high gradient performance. Such gradient strengths generally are not available at 0.5 T.

Limitations particular to the MR environment include long imaging time with repeated positioning of the receiver coil and repeated imaging if long segments of esophagus are involved. Also, it is crucial to follow special precautions when performing endoscopy in an MR environment, including the use of specially adapted MR-compatible endoscopes with the attendant problems of a long umbilical cord that is required to keep the light source out of the magnetic field. All other ferrous material also must be excluded from the magnetic field.

In conclusion, our preliminary results in a limited number of patients suggest that endoscopic MR imaging appears to be a safe technique that compares well with pathologic staging in esophageal cancer and can be combined with conventional external coil MR imaging to detect distant metastases. The rapid advances in MR technology will enhance the role of endoscopic MR imaging in the future.


   ACKNOWLEDGMENTS
 
Our thanks to Tomas Krausz, MD, FRCPath, for providing help with the histopathologic evaluation of the porcine specimens. Julian Teare, MD, FRCP, and David Rosin, MS, FRCS, FRCS(Ed), of St Mary’s Hospital, London; Devinder Bansi, MD, MRCP, Marta Carpani de Kaski, MD, Andrew Thillainayagam, MD, FRCP, and Nick Theodorou, MS, FRCS, of Charing Cross Hospital, London; Ajay Kakkar, BSc, PhD, FRCS, and Julian Walters, MA, MB, FRCP, of Hammersmith Hospital, London; and Alberto Isla, LMS, FRCS, of Ealing Hospital, all kindly provided patients for the study. David Harris, MD, FRCA, of the Department of Anesthesia, Hammersmith Hospital, London, provided initial advice on sedation during endoscopic MR imaging. Patrizia Cohen, BScHons, MBBCh, FFPath(SA), of Charing Cross Hospital, London, and Marjorie Walker, FRCPath, and Mary Thompson, MD, MRCP, FRCPath, of St Mary’s Hospital, London, reviewed the histopathologic results of the esophagectomy specimens.


   FOOTNOTES
 
Abbreviation: SI = signal intensity

Author contributions: Guarantors of integrity of entire study, U.R.D., S.D.T.R., N.M.d.S.; study concepts and design, U.R.D., S.D.T.R., M.R.T., N.M.d.S.; literature research, U.R.D., S.D.T.R., N.M.d.S.; clinical studies, all authors; experimental studies, N.M.d.S., A.D.W., D.J.G., D.J.L.; data acquisition, U.R.D., A.D.W., J.A.W., D.J.G., D.J.L., Z.A., S.D.T.R., N.M.d.S.; data analysis/interpretation, U.R.D., S.D.T.R., N.M.d.S.; manuscript preparation, definition of intellectual content, and revision/review, U.R.D., S.D.T.R., N.M.d.S.; manuscript editing, U.R.D., S.D.T.R., M.R.T., N.M.d.S.; manuscript final version approval, all authors


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 

  1. Blot WJ, Devesa SS, Kneller RW, Fraumeni JF, Jr. Rising incidence of adenocarcinoma of the esophagus and gastric cardia. JAMA 1991; 265:1287-1289.[Abstract/Free Full Text]
  2. Faivre J, Forman D, Esteve J, Gatta G. Survival of patients with oesophageal and gastric cancers in Europe. Eur J Cancer 1998; 34:2167-2175.[CrossRef][Medline]
  3. Allum WH. Staging of oesophagogastric cancer. Br J Surg 1998; 85:1-2.
  4. Pfau PR, Ginsberg GG, Lew RJ, Brensinger CM, Kochman ML. EUS predictors of long-term survival in esophageal carcinoma. Gastrointest Endosc 2001; 53:463-469.[CrossRef][Medline]
  5. Tio TL, Cohen P, and Coene PP. Endosonography and computed tomography of esophageal carcinoma: preoperative classification compared to the new (1987) TNM system. Gastroenterology 1989; 96:1478-1486.[Medline]
  6. Pham T, Roach E, Falk GL, Chu J, Ngu MC, Jones DB. Staging of oesophageal carcinoma by endoscopic ultrasound: preliminary experience. Aust N Z J Surg 1998; 68:209-212.[Medline]
  7. Vickers J. Role of endoscopic ultrasound in the preoperative assessment of patients with oesophageal cancer. Ann R Coll Surg Engl 1998; 80:233-239.[Medline]
  8. Angelelli G, Ianora AA, Scardapane A, Pedote P, Memeo M, Rotondo A. Role of computerized tomography in the staging of gastrointestinal neoplasms. Semin Surg Oncol 2001; 20:109-121.[CrossRef][Medline]
  9. Rosch T. Endoscopic staging of esophageal cancer: a review of the literature results. Gastrointest Endosc Clin N Am 1995; 5:537-547.[Medline]
  10. deSouza NM, Gilderdale DJ, Puni R, Coutts GA, Young IR. A solid re-usable endorectal receiver coil for magnetic resonance imaging of the prostate: design, use and comparison with an inflatable endorectal coil. J Magn Reson Imaging 1996; 6:801-804.[Medline]
  11. deSouza NM, Scoones DJ, Krausz T, Gilderdale DJ, Soutter WP. High-resolution MR imaging of stage I cervical neoplasia, using a dedicated endovaginal coil: MR features and correlation of imaging and pathologic findings. AJR Am J Roentgenol 1996; 166:553-559.[Abstract/Free Full Text]
  12. deSouza NM, Hall AS, Puni R, Gilderdale DJ, Young IR, Kmiot WA. High-resolution magnetic resonance of the anal sphincter using a dedicated endoanal coil: comparison of MR imaging with surgical findings. Dis Colon Rectum 1996; 39:926-934.[CrossRef][Medline]
  13. deSouza N, Gibbons A, Coutts G, Hall A, Puni R, Calam J, Young I. Magnetic resonance imaging during upper GI endoscopy: technical consideration and clinical feasibility. Minim Invasive Ther 1995; 4:277-281.
  14. Inui K, Nakazawa S, Yoshino J, et al. Endoscopic MRI: preliminary results of a new technique for visualization and staging of gastrointestinal tumors. Endoscopy 1995; 27:480-485.[Medline]
  15. Kulling D, Feldman DR, Kay CL, et al. Local staging of esophageal cancer using endoscopic magnetic resonance imaging: prospective comparison with endoscopic ultrasound. Endoscopy 1998; 30:745-749.[Medline]
  16. Gilderdale DJ, Williams AD, Dave U, deSouza NM. An inductively-coupled, detachable receiver coil system for use with magnetic resonance compatible endoscopes. J Magn Reson Imaging 2003; 18:131-135.[CrossRef][Medline]
  17. Mizowaki T, Nishimura Y, Shimada Y, et al. Optimal size criteria of malignant lymph nodes in the treatment planning of radiotherapy for esophageal cancer: evaluation by computed tomography and magnetic resonance imaging. Int J Radiat Oncol Biol Phys 1996; 36:1091-1098.[CrossRef][Medline]
  18. Chandawarkar RY, Kakegawa T, Fujita H, Yamana H, Toh Y, Fujitoh H. Endosonography for preoperative staging of specific nodal groups associated with esophageal cancer. World J Surg 1996; 20:700-702.[CrossRef][Medline]
  19. Yamada I, Murata Y, Izumi Y, et al. Staging of esophageal carcinoma in vitro with 4.7-T MR imaging. Radiology 1997; 204:521-526.[Abstract/Free Full Text]
  20. Kelly S, Harris KM, Berry E, et al. A systematic review of the staging performance of endoscopic ultrasound in gastro-oesophageal carcinoma. Gut 2001; 49:534-539.[Abstract/Free Full Text]
  21. Zuccaro G, Jr, Rice TW, Goldblum J, et al. Endoscopic ultrasound cannot determine suitability for esophagectomy after aggressive chemoradiotherapy for esophageal cancer. Am J Gastroenterol 1999; 94:906-912.[Medline]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
A. M. Riddell, D. C. Davies, W. H. Allum, A. C. Wotherspoon, C. Richardson, and G. Brown
High-Resolution MRI in Evaluation of the Surgical Anatomy of the Esophagus and Posterior Mediastinum
Am. J. Roentgenol., January 1, 2007; 188(1): W37 - W43.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. Radiol.Home page
A M Riddell, C Richardson, E Scurr, and G Brown
The development and optimization of high spatial resolution MRI for imaging the oesophagus using an external surface coil.
Br. J. Radiol., November 1, 2006; 79(947): 873 - 879.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Roentgenol.Home page
A. M. Riddell, J. Hillier, G. Brown, D. M. King, A. C. Wotherspoon, J. N. Thompson, D. Cunningham, and W. H. Allum
Potential of surface-coil MRI for staging of esophageal cancer.
Am. J. Roentgenol., November 1, 2006; 187(5): 1280 - 1287.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2301021047v1
230/1/281    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dave, U. R.
Right arrow Articles by deSouza, N. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dave, U. R.
Right arrow Articles by deSouza, N. M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE