| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Original Research |
1 Department of Radiology, Yonsei University College of Medicine, YongDong
Severance Hospital, 146-92 Dogok-Dong, Gangnam-Gu, Seoul 135-720, Republic of
Korea.
2 Department of Surgery, Yonsei University College of Medicine, Yong Dong
Severance Hospital, Seoul, Republic of Korea.
Received January 19, 2007;
accepted after revision June 10, 2007.
Address correspondence to J. S. Yu.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. The records of patients who had undergone radical gastrectomy for gastric cancer between August 2004 and August 2006 were retrospectively reviewed. Each patient was given 500-1,000 mL of water to drink, and 16-MDCT for preoperative staging was performed with IV contrast agents. Transverse and coronal reconstructed images had been prospectively interpreted as a part of daily practice. TNM stage, gross and histologic types, and location and size of the lesions were retrospectively analyzed for patients with primary lesions not visualized on preoperative MDCT.
RESULTS. In 44 patients with nonvisualized primary lesions, most (98%) of the lesions were stage pT1 confined to the mucosa (n = 27) or involving the submucosal layer (n = 16). Only one (2%) of the lesions had regional lymph node metastasis (pN1). Most (64%) of the lesions were located in the lower third of the stomach. Forty-one early gastric cancer lesions were flat, depressed, or excavated, and only two were elevated. Depending on the depth of invasion (mucosa or submucosa), there were no statistical differences (p > 0.05) in size, location, and gross or histologic type.
CONCLUSION. Almost all primary lesions of gastric cancer not visualized on preoperative MDCT with an optimized imaging protocol are early gastric cancer without regional lymph node metastasis. This negative finding may be helpful in planning minimally invasive management of gastric cancer.
Keywords: cancer staging • early gastric cancer • MDCT
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In daily clinical practice, however, in more than a few cases, lesions of gastric cancer not detected on preoperative MDCT have been histologically verified at endoscopic biopsy. We presume that in such cases, the presence of a nonvisualized primary lesion would implicate the presence of disease in an early stage that could be treated with a minimally invasive approach. The purpose of this study was to retrospectively determine the practical value of nonvisualized primary gastric lesions during staging of gastric cancer with a standard MDCT protocol in cases in which disease stage was eventually confirmed at radical gastrectomy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Preoperative CT Protocol
Two-phase (arterial and portal venous) contrast-enhanced dynamic CT was
performed with a 16-MDCT scanner (Somatom Sensation 16, Siemens Medical
Solutions). Patients fasted for at least 5 hours before the examinations and
then ingested 500-1,000 mL of tap water orally to distend the gastric wall
just before the CT examination. Because of the middle or distal gastric
locations of the primary lesions, most of the patients were placed in the
prone position on the scanning table. The exception was patients with known
primary lesions in the fundal portion of the stomach, who were placed in a
supine position. All patients received 150 mL of nonionic contrast material
(iopromide, Ultravist 300, Bayer HealthCare) IV by means of a power injector
(EnVisionCT, Medrad) at a rate of 3 mL/s. Scans were acquired in a
craniocaudal direction with the following parameters: detector collimation, 16
x 0.75 mm; table feed, 12 mm per rotation; section width, 5 mm;
reconstruction increment, 5 mm with 5-mm-thick sections; pitch, 1.2; tube
current, 120 kVp; 160 mAs.
For most patients, acquisition of arterial phase scans was initiated 15 seconds after enhancement of the thoracic aorta until 100 H was reached as measured with a bolus-tracking technique after injection of the contrast material. In a few patients, a test bolus injection of 20 mL of contrast material was administered and monitored by single-level image acquisition at the level of the hepatic hilum. The start time for the hepatic arterial phase was calculated by the addition of 15 seconds to the time of peak aortic enhancement during the early study period. Portal venous phase scans were acquired 70 seconds after the start of the contrast injection. Portal phase image sets were sent to a workstation (Wizard, Siemens Medical Solutions), and coronal reconstruction images with a section thickness of 2 mm without a gap were reformatted with the standard software that accompanied the scanner.
CT Interpretation
All of the arterial and portal phase transverse images and coronal
reformatted portal phase images of each patient were simultaneously displayed
on PACS monitors. These images had been routinely interpreted by two
observers, an attending radiologist and a radiology resident, at the official
interpretation. The resident usually reviewed the images first and then
reviewed and discussed them with the attending radiologist. The final
interpretation regarding TNM stage, size, contour, enhancement pattern, and
location of the primary lesion was recorded with agreement between these two
observers. In all, four attending radiologists, each with a minimum of 5 years
of experience in abdominal imaging, and nine junior and senior radiology
residents were involved in interpreting the images. The reviewers had
information about the presence of primary gastric cancer as verified by
endoscopic biopsy with or without knowledge of the gross contour or location
of the lesions. According to the guidelines agreed on by the abdominal imaging
section of our department, lack of visualization of a primary gastric lesion
implicated the absence of distinctive abnormal focal enhancement or localized
defect in mucosal enhancement regardless of the layering pattern of mural
enhancement without localized or diffuse mural thickening with abnormal
contrast enhancement. The presence of regional lymph nodes was recorded. Lymph
nodes were considered involved when the short-axis diameter was larger than 6
mm for perigastric lymph nodes and larger than 8 mm for extra-perigastric
lymph nodes [12,
13].
Gross and Microscopic Data Analysis
The gross and histopathologic features of gastric cancer were officially
reported by one pathologist with 30 years of experience in the field of
gastrointestinal disease. Mural invasion of gastric cancer into the gastric
wall and lymph node involvement were classified according to the TNM staging
of gastric cancer on the basis of the sixth edition of the International Union
Against Cancer TNM classification
[14]. For early gastric cancer
(EGC), depth of tumor cell invasion was subdivided into mucosa and submucosa,
and the lesions were further assigned to the gross classification of protruded
(type 1), elevated (type 2a), flat (type 2b), depressed (type 2c), or
excavated (type 3). Borrmann classification and depth of tumor invasion were
analyzed for advanced gastric cancer. The size of the lesion was reported in
square centimeters and calculated as follows: (long-axis diameter x 0.5)
x (short-axis diameter x 0.5) x 3.14. The location of the
lesion was subdivided into upper, middle, or lower third of the stomach.
According to the Japanese classification of gastric cancer
[15], the histologic types of
papillary, well differentiated, and moderately differentiated tubular
adenocarcinoma were classified as differentiated type. Poorly differentiated
adenocarcinoma, mucinous adenocarcinoma, and signet ring cell carcinoma were
classified as undifferentiated types.
Depending on the depth of tumor invasion and lymph node involvement of the gastric cancer, Student's t tests were used to compare the mean size of the lesions, and chi-square tests were used to compare the locations, gross types, and histologic types of the lesions. A value of p < 0.05 was considered a statistically significant difference.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Regardless of the depth of tumor invasion (p = 0.772), most (41 [95%] of 43) of the EGC lesions were of the flat, depressed, or excavated tumor type (Table 2). Only two lesions had an elevated appearance, and there were no protruded-type nonvisualized lesions on preoperative MDCT. Many of the EGC lesions confined to the mucosa (13 [48%] of 27) or involving the submucosa (six [38%] of 16) were histologically verified signet ring cell carcinomas. In addition, 13 (68%) of 19 signet ring cell carcinomas were confined to the mucosa. However, the ratio of differentiated types versus undifferentiated types showed no statistical difference (p = 0.780) according to depth of tumor cell invasion in patients with EGC (Table 2).
|
The presence of regional lymph nodes was described in 13 (30%) of 44 patients. Among these patients, only two had small (< 0.6 cm in short-axis diameter) benign-appearing nonenhancing perigastric lymph nodes. Seven patients had single or several (as many as six) perigastric lymph nodes measuring 0.6-0.8 cm in short-axis diameter, and another three patients had multiple small conglomerated perigastric lymph nodes (up to 1.6 cm in longest dimension) with contrast enhancement. The disease in these patients was staged N1. In another patient with 10 enlarged perigastric and extraperigastric lymph nodes as large as 2 cm, disease was staged N2 on preoperative MDCT despite the presence of nonvisualized primary gastric lesions. On review of the pathologic specimen, only one patient with EGC confined to the mucosa (2.5 x 2.0 cm) had five small metastatic lymph nodes (pN1) in the greater omentum. There was no lymph node involvement in the other 43 (98%) patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
With the use of less-invasive options, precise locoregional staging of tumors is essential for individual therapeutic planning with preoperative imaging studies [21]. Besides endoscopic mucosal resection of small superficial mucosal lesions determined on the basis of gross endoscopic findings and cellular differentiation, the indications for laparoscopy-assisted gastrectomy include T1N0M0, T1N1M0, and T2N0M0 disease [23]. When surgeons decide whether minimally invasive surgery should be performed on the basis of endoscopic or endoscopic sonography findings, a problem is that the diagnostic accuracy of endoscopic sonography in terms of depth of invasion is not optimal, and some patients with EGC miss the opportunity to be treated with minimally invasive surgery [24]. With an increase in diagnostic accuracy in the evaluation of EGC, more patients would be able to undergo minimally invasive surgery and thus have a better quality of life than they would with radical treatment.
After detection of primary lesions with endoscopy or double-contrast barium studies, preoperative staging is mainly performed with CT. Use of MDCT, which enables thin collimation and fast imaging of isotrophic 3D MPR images with enhanced resolution, has the potential to improve the accuracy of preoperative staging of gastric cancer [5-8]. To avoid false-positive and false-negative findings of mural or perigastric involvement, adequate water distention of the stomach for dynamic imaging with MPR allows the reviewer to choose the optimal imaging plane. This capability enables better delineation of intramural and extramural tumor involvement and recognition of small perigastric lymph nodes. Even with the optimized CT protocol, however, lack of visualization of primary lesions is not infrequent in daily practice. In a recent investigation, Shimizu et al. [6], who used a CT protocol similar to ours consisting of water-filled luminal distention and dynamic MPR imaging, found that only 14 (41%) of 34 cases of EGC were detected and that the detection rate of lesions confined to the mucosa was very low (17%). In that study, the detection rate of EGC involving the submucosal layer was 69%.
Detection of gastric cancer used to be influenced by morphologic features, thickening of the gastric wall, and the degree of the tumor. In this study, we found no protruded-type EGC lesions; most of the lesions had a flat, depressed, or excavated appearance without recognizable mural thickening. As mentioned previously [6, 25, 26], our findings suggest that CT may not be best suited for detection of nonprotruded EGC lesions, even with thin-slice MPR.
On dynamic imaging studies, EGC lesions may be detected as areas of prominent contrast enhancement or focal interruption of the enhancing gastric wall on arterial phase images [25-28]. However, the distinction between well-enhanced mucosal or nonprotruding EGC lesions and well-enhanced surrounding normal mucosal layers can be difficult to visualize on CT. For EGC lesions involving the submucosal layer, recognition of a two- or three-layered pattern of the gastric wall on the dynamic scans would be helpful for lesion depiction [25, 26, 29]. Basically, pT1-stage EGC cannot be differentiated from pT2-stage advanced gastric cancer on dynamic CT scans when the normal layered pattern is not recognized. The water-filling method to achieve adequate gastric distention has merit because it can show the low-attenuation density of the submucosal layer on early-phase images [26, 29]. However, problems with this layered pattern of a normal stomach are not constantly seen in daily clinical practice, even with the use of thin-slice MPR to minimize the influence of oblique scanning and partial volume averaging effects [9]. Because of difficulties correlating the exact location of an endoscopically recognized tumor with normal CT findings, it was not possible to verify the presence of the layered pattern at the site of the lesion in this study. However, almost all of the patients had pT1-stage EGC, and this finding can be considered to support a minimally invasive endoscopic or surgical approach.
The degree and pattern of contrast enhancement of gastric cancer are influenced by the histologic characterization [5, 25]. It has been suggested [25, 27] that undifferentiated infiltrative tumors are not well enhanced on early-phase images. Dependent on the degree of interstitial fibrosis, as in the scirrhous sub-types of advanced gastric cancer, such lesions can be well delineated on rather delayed-phase images. In this study, nearly one half of the nonvisualized lesions were histologically undifferentiated types of mostly signet ring cell carcinoma, and the degree of tumoral enhancement was not differentiated from adjacent normal mucosa. EGC with signet ring cell histologic features tends to be a depressed lesion confined to the mucosa [30]. It is not unusual to miss the primary lesion on preoperative CT, as occurred in this study. Early-stage lesions are difficult to detect on more delayed-phase images owing to the homogeneous enhancement of the lesion and the entire gastric wall [27, 28].
Regarding the size of the lesions from the gross pathologic specimens obtained from our patients, only 2D tumor extent was described. In general, the rate of detection of mass-forming lesions is influenced by the volume and size of the lesions in any organ or structure. In this study, however, most of the EGC lesions were not accompanied by mural thickening, and the 2D size difference alone could not have influenced the tumor detection rate. For the one case of missed advanced gastric cancer involving the muscularis propria, the small 2D extent (0.8 cm2) in addition to the ulceroinfiltrative gross type and undifferentiated histologic type might have influenced the lack of visualization of the primary lesion despite the considerable depth of tumor invasion.
Lesion location can influence lesion detection. Filling the entire stomach with water is impossible, and the portion filled with air depends on the position of the patient. In this study, most of the lesions were located in the middle and lower thirds of the stomach, and most of the patients who underwent CT were in the prone position for the examination. It is unlikely that one would miss primary lesions owing to inadequate positioning of the patients. Because we did not conduct a retrospective imaging-to-pathology analysis, we do not know the details about the more vulnerable locations of missed lesions in individual patients.
In this study, almost all of the patients had no pathologically confirmed regional lymph node metastasis regardless of the presence of enlarged lymph nodes around the stomach [12, 13]. The only case of EGC with regional lymph node metastasis (pN1) was confined to the mucosa, and there was no specific finding of lymph node involvement differentiated from the other false-positive findings of lymph node involvement. These results suggest that primary lesions not visualized on preoperative MDCT are likely statistically T1N0, even in cases of nonspecifically enlarged regional lymph nodes. This finding should be proven in a large enough series of patients.
A number of cases of EGC involving the submucosal layers were included in this study. In a comparison of these lesions with EGC lesions confined to the mucosa, there were no statistical differences in the gross and histologic types or in the size and location of the lesions. Because of the ability to depict the five layers of the gastric wall, endoscopic sonography may be the only method that can provide the information needed for differentiation of mucosa-confined lesions and detailed depth of invasion in the submucosal layer [1-4]. However, endoscopic sonography is operator-dependent, and the potential for understaging or overstaging is considerably high. In addition, the accuracy of endoscopic sonography as a single imaging technique for preoperative staging is still not sufficient for selecting patients for endoscopic mucosal resection [24]. For patients with controversial endoscopic sonography findings, lack of visualization of a primary lesion on preoperative MDCT would provide supplementary information suggesting EGC be subjected to more precise locoregional tumor staging. Even though we do not have any correlation data between CT and endoscopic sonography, depending on the endoscopic and CT findings, all but one of our patients underwent laparoscopy-assisted gastrectomy (D1 + ß). The exception was a patient with CT findings suggestive of involvement of the extraperigastric lymph nodes but findings at radical surgery with D2 lymph node dissection that confirmed stage pN0. Follow-up studies have shown no evidence of localized tumor recurrence or metastasis.
Our study had several limitations. First, the water-filling technique was not optimized for detection of EGC. Results of two investigations [31, 32] suggest that air distention of the stomach with virtual gastroscopy during MDCT may help detection of EGC. However, CT is not a screening imaging technique for initial lesion detection. The main purpose of a CT examination is preoperative staging of lesions already detected with endoscopy or double-contrast barium study. We know the water-filling technique is more favorable for recognition of the layered pattern of the normal gastric wall to evaluate the depth of intramural tumor invasion [6, 9]. We do not have any data on lack of visualization of primary lesions with the air distention technique; thus the results of our study cannot be adjusted for the patients examined with MDCT performed with air distention of the stomach. Second, we did not analyze the cases of all EGC and advanced gastric cancer patients regardless of the imaging features, so we do not know the overall incidence of nonvisualized lesions according to depth of tumor invasion or gross and histologic types. Moreover, we have no data on the relative value of arterial versus portal phase imaging for accuracy of tumor staging. These issues were beyond the scope of the design of the study.
In conclusion, the presence of nonvisualized primary lesions on preoperative MDCT performed with water-filling gastric distention and MPR strongly suggests the presence of EGC without lymph node metastasis. In an era of individual therapeutic approaches to gastric cancer according to depth of tumor invasion, this negative finding should be helpful in choosing minimally invasive surgical options for the management of early-stage lesions with reference to the endoscopic and histologic features in conjunction with the endoscopic sonography findings.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
