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


     

DOI: 10.1148/radiol.2422051557
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
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 Chen, C.-Y.
Right arrow Articles by Liu, G.-C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, C.-Y.
Right arrow Articles by Liu, G.-C.
(Radiology 2007;242:472-482.)
© RSNA, 2007

Gastrointestinal Imaging

Gastric Cancer: Preoperative Local Staging with 3D Multi–Detector Row CT—Correlation with Surgical and Histopathologic Results1

Chiao-Yun Chen, MD, Jui-Sheng Hsu, MD, PhD, Deng-Chyang Wu, MD, PhD, Wan-Yi Kang, MD, Jan-Sing Hsieh, MD, Twei-Shiun Jaw, MD, MMS, Ming-Tsang Wu, MD, PhD and Gin-Chung Liu, MD

1 From the Departments of Medical Imaging (C.Y.C., J.S.H., T.S.J., G.C.L.), Radiology (J.S.H., T.S.J., G.C.L.), Medical Gastroenterology (D.C.W.), Pathology (W.Y.K.), Surgery (J.S.H.), and Public Health (M.T.W.), Kaohsiung Medical University, Chung-Ho Memorial Hospital, 100 Tz-You 1st Road, Kaohsiung, Taiwan. From the 2004 RSNA Annual Meeting. Received September 19, 2005; revision requested November 14; revision received December 22; accepted January 23, 2006; final version accepted May 15. Address correspondence to G.C.L. (e-mail: gcliu{at}kmu.edu.tw).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Purpose: To prospectively evaluate accuracy of multi–detector row computed tomographic (CT) images for preoperative staging of gastric cancer by using surgical and histopathologic results as reference standards.

Materials and Methods: This study had institutional review board approval; informed consent was obtained from all patients. Multi–detector row CT included acquisition of virtual gastroscopy images after air distention and contrast material–enhanced dynamic transverse and multiplanar reformation (MPR) images after water distention. Fifty-five consecutive patients with gastric cancer (38 men, 17 women; age range, 37–84 years; mean age, 63 years) underwent preoperative CT. All received 6 g of gas-producing crystals before unenhanced CT scanning for gastric distention and virtual gastroscopy. Patients drank 800–1000 mL of tap water to establish a background for dynamic contrast-enhanced CT scans. Images were obtained in late arterial, portal venous, and delayed phases with start delays of 40, 70, and 150 seconds, respectively. All patients underwent surgery. CT findings were compared with surgical and histopathologic results. Differences in accuracy of transverse and MPR images for T and N staging were assessed with the McNemar exact test. Statistical significance was inferred at P < .05.

Results: Detection rates of primary tumors with transverse images, MPRs, and combinations of MPR and virtual gastroscopy images were 91% (50 of 55), 96% (53 of 55), and 98% (54 of 55), respectively. Overall accuracy in assessment of tumor invasion of the gastric wall (T stage) was significantly better with MPR images (89% [49 of 55]) than with transverse images (73% [40 of 55]) (P < .01). Overall accuracy for lymph node (N) staging was 78% (43 of 55) with MPR images and 71% (39 of 55) with transverse images. This difference was not significant (P = .103).

Conclusion: Multi–detector row CT with combined water and air distention can improve the accuracy of preoperative staging of gastric cancer. MPRs yield significantly better overall accuracy than transverse images for tumor staging but not for lymph node staging.

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Despite its declining incidence, gastric cancer remains an important cause of cancer death in Japan (1) and elsewhere. To reduce mortality, it is essential to choose an optimal therapeutic approach, and this, in turn, depends on early detection and accurate preoperative staging. Indeed, prognosis is related to depth of invasion of the gastric wall and lymph node involvement (2). Survival is improved with curative resection and palliative chemotherapy (3). A small early gastric cancer confined to the submucosa (T1 stage) can be treated with nonsurgical endoscopic mucosal resection (4). Preoperative chemotherapy or radiation therapy is recommended for advanced gastric cancer. Accurate preoperative staging, therefore, can help increase cure rates and quality of life.

Preoperative staging has often included endoscopic ultrasonography (US) and computed tomography (CT), but hydrodynamic helical CT might replace preoperative endoscopic US in T and N staging (5). Multi–detector row CT with thin collimation offers near-isotropic imaging of the stomach and allows high-quality multiplanar reformations (MPRs) and endoluminal three-dimensional virtual gastroscopy of gastric images. With air distention of the stomach, virtual gastroscopy images allow evaluation of gastric endoluminal disease. With adequate distention of the stomach by using water as negative contrast, dynamic contrast material–enhanced CT images offer superior differentiation of tumor tissue from normal mucosa. Also, MPRs have advantages for assessing both intra- and extraluminal processes of the gastric wall and for evaluating more distant regions, such as paraaortic lymph nodes and abdominal organs. Thus, the purpose of our study was to prospectively evaluate the accuracy of multi–detector row CT images for preoperative staging of gastric cancer by using surgical and histopathologic results as reference standards.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Patients
Our study had institutional review board approval. Informed consent was obtained from all patients. Between June 2003 and November 2004, 69 consecutive patients with gastric cancer were prospectively and preoperatively examined with multi–detector row CT. All enrollees had local gastric disease (T1–T4, N1–N3) without distant metastasis that was diagnosed with multi–detector row CT; 14 enrollees did not undergo surgery and were excluded because CT showed advanced disease–multiple liver metastases, extensive peritoneal carcinomatosis, or paraaortic lymph nodes and/or pancreatic involvement. The remaining 55 patients included 38 men and 17 women (age range, 37–84 years; mean age, 63 years) (Fig 1). All CT examinations followed a protocol designed for patients with gastric cancer. Before CT, all patients underwent endoscopic biopsy that provided histologic confirmation of gastric cancer. Tumors were found in the fundus (n = 4), body (n = 14), antrum (n = 28), or pyloric region (n = 6) of the stomach. Three patients had linitis plastica that involved the body of the stomach to the antrum. All patients underwent laparotomy and, within 7 days, resection of the primary gastric cancer with perigastric nodal resection. Histologically, all tumors were adenocarcinomas of the gastric mucosa.


Figure 1
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Flowchart of study enrollment.

 
CT Protocol
Multi–detector row CT was performed by using a 16–detector row CT scanner (LightSpeed H16; GE Healthcare, Milwaukee, Wis) in patients who had fasted more than 8 hours. Each patient received 6 g of gas-producing crystals with 10 mL of water orally shortly before unenhanced CT to enable gastric distention and virtual gastroscopy. In cases of insufficient air distention of the stomach, an additional 3 g of gas-producing crystals were given. Unenhanced upper abdominal CT scans from the diaphragmatic domes to 2 cm below the lower margin of the air-distended gastric body were acquired with a collimation of 1.25 mm, a table speed of 27.5 mm/sec, 250–300 mAs, and 120 kVp. In five patients in whom much residual fluid covered the tumor, changing to the right or left decubitus position and additional scanning were performed. All procedures were performed by one author (G.C.L., who has specialized in gastrointestinal CT for more than 15 years), who was not involved in the evaluation of virtual gastroscopy data.

Immediately after unenhanced CT, each patient drank 800–1000 mL of tap water while in the CT gantry to provide negative gastric contrast for contrast-enhanced dynamic CT. This allows good visualization of the enhancing gastric wall (6) and avoids overshooting artifacts due to intraluminal air obscuring the normal gastric wall pattern (7). One hundred milliliters of a nonionic iodinated contrast agent (Ultravist; Schering, Berlin, Germany) was administered via the antecubital vein at 3 mL/sec by using a 20-gauge needle and an automatic injector. CT was performed in the late arterial phase (start delay, 40 seconds), in the portal venous phase (70 seconds), and in the delayed phase (150 seconds). The late arterial and delayed phases were used to evaluate T staging; the portal venous phase was used to evaluate N staging (8). The entire procedure was performed within 20 minutes.

To reduce radiation dose from dynamic scanning, the investigator (G.C.L.) tried to locate the lesion on unenhanced images and to define the minimum possible scanning range in late arterial and delayed phases. If the tumor site could not be definitely located in unenhanced images, the CT examination included the whole stomach. Portal venous phase CT scanning included the whole abdomen and pelvis, from the diaphragmatic domes to the anal verge. We reconstructed raw data sets at 1.25-mm section thicknesses and 0.9-mm reconstruction intervals to create virtual gastroscopy images (from the unenhanced images obtained with air distention) and MPRs. We used a 5-mm section thickness for transverse CT images.

Image Analysis
Two abdominal radiologists (C.Y.C. and J.S.H., with 8 and 6 years of experience, respectively, in abdominal CT and 2 years of experience in three-dimensional endoluminal approaches and interpretation) performed the image analysis. They were blinded to endoscopic results, lesion size, and macroscopic features and independently evaluated CT images at a workstation (AW 4.1; GE Healthcare) before surgery by using a navigator for virtual gastroscopy images and soft-tissue window settings for transverse CT (5 mm) and MPR images (1.25 mm or greater). Transverse images were initially assessed in random order. MPR and virtual gastroscopy images were then evaluated separately at a different time (interval, ≥2 weeks), also in random order. With MPR images, the depth of tumor invasion was detected with a projection vertical to the tumor to avoid partial volume effects. All lymph nodes were simultaneously evaluated with transverse, coronal, and sagittal images at the workstation. Differences in assessment were resolved by consensus.

Like conventional gastroscopy, virtual gastroscopy depicted abnormal endoluminal lesions (9–12). We classified the focal morphologic abnormalities depicted by using virtual gastroscopy into three early gastric cancer groups (13) and four Borrmann categories for advanced gastric cancer (Fig 2) (14,15). Triple-phase contrast-enhanced transverse images and MPRs were evaluated separately. Normal gastric walls frequently showed multilayered (two- or three-layered) patterns, with a markedly enhanced inner mucosal layer and an intermediate hypoattenuating submucosal layer. The outer, slightly high-attenuating, muscular-serosal layer is very thin (16,17). We assessed and compared the following: the presence, sites, and morphologic features of tumors; the depth of tumor invasion; the degree of tumor enhancement (for T staging); and regional lymph node involvement (for N staging).


Figure 2A
View larger version (45K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: (a) Japanese classification of early gastric cancer and (b) Borrmann classification of advanced gastric cancer.

 

Figure 2B
View larger version (45K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: (a) Japanese classification of early gastric cancer and (b) Borrmann classification of advanced gastric cancer.

 
CT criteria for mural invasion of cancer into the gastric wall are shown in Table 1 (1,5,16–18). Regional lymph nodes were considered to represent local metastases if they were solitary or separate nodes 8 mm or greater in long-axis diameter with marked enhancement, which was defined as attenuation greater than 85 HU in the postcontrast portal venous phase (19). Attenuation measurement was performed by placing a region of interest in the central portion of the node with the greatest size on transverse sections. Regional lymph nodes were also considered to represent local metastases if they were cluster nodes or solitary nodes with associated reticular strands, independent of size or enhancement pattern. Imaging-pathologic correlations of lymph nodes were performed by using group-by-group analysis. Our CT staging method for nodal metastasis (Table 1) was based on the nomenclature of the American Joint Committee on Cancer and the International Union Against Cancer (20).


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

  		Table 1. CT Criteria for T and N Staging of Gastric Cancer

 
Tumor enhancement patterns were analyzed in all phases. Good, intermediate, and poor tumor enhancement, respectively, indicate higher attenuation, isoattenuation, and lower attenuation compared with the normal gastric wall on contrast-enhanced images. The extragastric compartment of the abdomen and pelvis were evaluated carefully on portal venous phase contrast-enhanced MPR images. Metastases were recorded as present or absent.

Surgery
All 55 patients underwent partial or complete gastrectomy. The surgical procedure was performed on the basis of clinical and surgical findings defined according to the Japanese Research Society for Gastric Cancer (14). Of the 55 patients, 37 underwent D1 lymphadenectomy, and 18 underwent D2 lymphadenectomy. D2 lymphadenectomy was defined, according to the Japanese Research Society for Gastric Cancer, as a resection of all N1 and N2 nodes with en bloc resection of the stomach. If any of the second-tier stations were not resected, the procedure was downgraded to represent a D1 lymphadenectomy. Individual lymph node groups were dissected and placed in separate specimen pots. The gastrectomy specimen and associated lymph node groups were transferred to the pathology department.

Pathologic TNM Staging
Histopathologic findings were reported by a specialist with 6 years of experience in gastrointestinal pathology (W.Y.K.). Location, gross appearance, tumor size, and depth of invasion were determined. All metastatic nodes were labeled regarding location and size. The standard for liver, peritoneal, and retroperitoneal metastases was results of histopathologic examination, intraoperative hepatic US, or surface exploration. T and N staging was based on the pathologic TNM (pTNM) staging system developed by the American Joint Committee on Cancer and the International Union Against Cancer (Table 2) (20).


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

  		Table 2. Pathologic TNM Staging System Developed by American Joint Committee on Cancer and International Union Against Cancer in 2002

 
Statistical Analysis
Detection rates with transverse images, MPR images, and the combination of MPR and virtual gastroscopy images were calculated. Test performance characteristics (sensitivity, specificity, accuracy) were calculated with transverse and MPR images for T and N staging. The differences in diagnostic accuracy between transverse and MPR images for T and N staging were assessed by using the McNemar exact test. Statistical significance was inferred for a two-tailed P value less than .05.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Detection and Morphology of Primary Tumors
Detection rates of primary tumors with transverse images, MPR images, and the combination of MPR and virtual gastroscopy images were 91% (50 of 55 neoplasms), 96% (53 of 55 neoplasms), and 98% (54 of 55 neoplasms), respectively. One primary tumor (type IIc) could not be detected with MPRs but was detected with virtual gastroscopy images. One primary tumor (type IIb early gastric cancer) could not be detected with either MPR or virtual gastroscopy images. Five of nine early gastric cancers were not detected with transverse images. Virtual gastroscopy images showed seven early gastric cancers (one type I, four type II, and two type III cancers) and 46 advanced gastric cancers (five type I, 14 type II, 22 type III, three type IV, and two type I + II cancers according to the Borrmann classification). All neoplasms were verified with pathologic correlation.

Enhancement Patterns of Primary Tumors on Dynamic Contrast-enhanced MPRs
On dynamic contrast-enhanced MPRs, well-enhanced tumors (51 of 55) showed variable degrees of focal thickening of the gastric wall with gradually increasing enhancement of the tumor parts from the inner mucosal layer (late arterial phase) to the outer tumoral margin (delayed phase). Two of nine early gastric cancers showed no focal gastric wall thickening and had an intermediate enhancement pattern. In the delayed phase, seven of nine early gastric cancers showed low-attenuation strips in the outer margin corresponding to the submucosal layer (Fig 3). Transmural tumor enhancement with smooth outer gastric surfaces was found in 15 of 17 T2 gastric tumors (Fig 4). Two tumors showed poor enhancement: One was a mucinous tumor, the other a signet-ring cell tumor.


Figure 3
View larger version (143K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Early gastric cancer (type III) in 66-year-old woman. A–C, Transverse dynamic contrast-enhanced CT images show enhancing tumor from arterial phase to delayed phase. There is a low-attenuation strip (arrow) representing the submucosal layer in outer tumor border; this finding suggests pathologic stage T1. Also shown is a well-enhanced lymph node (arrowhead) in infrapyloric region. D, Coronal MPR shows well-enhanced mucosal tumor (arrow) with visible outer submucosal layer and clear fat plane around gastric wall. E, Photomicrograph shows submucosal invasion of early gastric cancer (pT1). (Hematoxylin-eosin stain; original magnification, x10.) F, Virtual gastroscopy image shows excavated lesion (arrow) at gastric antrum. G, Conventional gastroscopy revealed a similar excavated ulcerated lesion (arrow).

 

Figure 4
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Advanced gastric cancer (Borrmann type III) in 50-year-old woman. A, Virtual gastroscopy image shows typical ulcerated carcinoma in stomach body. B, Surgical specimen shows a similar lesion. C–E, Transverse dynamic contrast-enhanced CT images obtained in, C, arterial phase, D, portal venous phase, and, E, delayed phase show transmural, gradually enhancing tumor (arrow) with smooth outer border of gastric wall; these findings suggest pathologic stage T2. F, Photomicrograph shows subserosal invasion of gastric cancer (pT2). (Hematoxylin-eosin stain; original magnification, x5.)

 
Depth of Invasion of Gastric Wall
At histopathologic examination, nine of 55 neoplasms (16%) were staged as pT1; 17 neoplasms (31%), as pT2; 27 neoplasms (49%), as pT3 (Fig 5); and two neoplasms (4%), as pT4 (Fig 6). The overall diagnostic accuracy of T staging with dynamic contrast-enhanced transverse images was 73% (40 of 55 neoplasms), which improved to 89% (49 of 55 neoplasms) with MPR images (P < .01). MPR images were superior to transverse images in terms of diagnostic accuracy (Figs 5–7). With transverse images, over- and understaging occurred in three (5%) and 12 (22%) patients, respectively. With MPR images, over- and understaging (Fig 8) occurred in three (5%) and three (5%) patients, respectively. Accuracy rates for each T stage with transverse and MPR images, respectively, were 89% and 96% for T1, 85% and 95% for T2, 84% and 93% for T3, and 96% and 98% for T4 (Table 3).


Figure 5
View larger version (149K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5: Advanced T3 gastric cancer (Borrmann type I + II) in 65-year-old man. A, Coronal oblique arterial phase MPR shows well-enhanced hyperintense mucosal tumor with a nodular outer border of the stomach and reticular strands (arrow) in the fat plane contiguous with the outer border of the tumor; these findings suggest pathologic stage T3. B, Coronal oblique delayed phase MPR image shows transmural hyperintense tumor with irregular outer border of the stomach and reticular strands (arrow) in the fat plane contiguous with the outer border of the tumor in antrum of stomach; these findings suggest pathologic stage T3. C, Gross and, D, histologic specimen show extraserosal invasion of gastric cancer (pT3). (Hematoxylin-eosin stain; original magnification, x5.) E, Virtual gastroscopy image shows protruding mass (arrow) with ulcerated tumor. F, Conventional gastroscopy showed a similar finding (arrow).

 

Figure 6
View larger version (139K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6: Advanced T4 gastric cancer in 59-year-old woman. A, Transverse CT image shows well-enhanced tumor with adjacent fat plane infiltration (arrow) adjacent to transverse colon with normal morphology; these findings suggest pathologic stage T3. B, Coronal oblique MPR shows obliteration of fat planes (arrow) between cancer and transverse colon and direct tumor invasion of the superior margin of transverse colon; these findings suggest pathologic stage T4. C, D, Gross specimens show direct colonic invasion (arrow) of gastric cancer (pT4).

 

Figure 7
View larger version (105K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7: Advanced T2 gastric cancer in 56-year-old man. A, Transverse CT image shows focal well-enhanced hyperintense mucosal tumor (arrow) with visible outer low-attenuation strip and a clear fat plane around the tumor; these findings suggest pathologic stage T1. B, Coronal oblique MPR shows focal transmural involvement of the hyperintense mucosal tumor (arrow) in the superior aspect of the gastric antrum; this finding suggests pathologic stage T2. This lesion proved to be a subserosal invasion of gastric cancer (pT2).

 

Figure 8
View larger version (154K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8: T3N3M1 cancer mistakenly classified as T2N3M0 in 51-year-old woman. A, Coronal MPR shows focal, poorly enhanced tumor (arrow) in stomach body and transmural involvement with a smooth outer border and clear perigastric fat plane; these findings led to staging of the tumor as T2. B, C, Photomicrographs, however, show smooth outer border (in B) and scattered malignant cells penetrating the serosal margin (in C). (Hematoxylin-eosin stain; original magnification, x5 in B and x40 in C.) Pathologic stage was T3. D, Coronal MPR shows two clusters of enhancing nodes (arrows) along bilateral peritoneal cavities that proved metastatic at histopathologic examination. Bilateral ovarian metastases were not diagnosed owing to poor soft-tissue contrast resolution and no morphologic change (not shown).

 

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

  		Table 3. Diagnostic Accuracy of Contrast-enhanced Multi–Detector Row CT for Each T Stage with Histopathologic Results as Reference Standard

 
Nodal Involvement
At histopathologic examination, 18 neoplasms (33%) were staged as pN0; 26 neoplasms (47%), as pN1; seven neoplasms (13%), as pN2; and four neoplasms (7%), as pN3. Overall accuracy of N staging with enhanced transverse images was 71% (39 of 55 neoplasms). Overall accuracy of N staging improved to 78% (43 of 55 neoplasms) with MPR images (Fig 9). However, these differences were not statistically significant (P = .103). Over- and understaging with transverse images occurred in six (11%) and 10 (18%) patients, respectively. With MPR images, over- and understaging occurred in seven (13%) and five (9%) patients, respectively. Nodal metastases were detected in 32 of 37 patients (86%) with transverse images and in 34 of 37 patients (92%) with MPR images. Among patients without nodal metastases, 14 of 18 (78%) received a correct classification with transverse images; 13 of 18 (72%) received a correct classification with MPR images. Accuracy rates for each N stage with transverse and MPR images, respectively, were 84% and 85% for N0, 75% and 80% for N1, 85% and 91% for N2, and 98% and 100% for N3 (Table 4).


Figure 9
View larger version (187K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9: Advanced gastric cancer in 63-year-old woman. Coronal oblique MPR shows cluster of enhancing small nodes (arrows) around infrapyloric region. Histopathologic examination revealed metastatic nodes. One enlarged solitary node with poorly enhancing ovoid shape (arrowhead) is seen along greater curvature. However, pathologic examination showed a reactive pattern.

 

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

  		Table 4. Diagnostic Accuracy of Contrast-enhanced Multi–Detector Row CT for Each N Stage with Histopathologic Results as Reference Standard

 
Distant Metastasis and Peritoneal Involvement
Dynamic contrast-enhanced multi–detector row CT with MPRs acquired during the portal venous phase correctly depicted the presence or absence of solid organ metastases in 54 of 55 patients (98%). Disease was understaged in one patient with bilateral ovarian metastases. In this false-negative case, the patient had mild enlargement of the right ovary and a normal-sized left ovary—both with normal attenuation—at CT. At surgery, the ovaries were hard and showed color change. Regarding peritoneal involvement, eight patients had omental metastases and peritoneal seeding at surgery. Accuracy for peritoneal involvement was 96% (53 of 55 neoplasms).

Interobserver Agreement
We used blinded consensus reading to resolve disagreements between the interpreting radiologists. In the evaluation of T staging, there was agreement in 51 of 55 patients (93%) with transverse images and in 54 of 55 patients (98%) with MPR images. In the evaluation of lymph node involvement, there was agreement in 50 of 55 patients (91%) with transverse images and in 52 of 55 patients (95%) with MPR images. With regard to extragastric metastases, the readers showed agreement in all patients.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
Our prospective study combined virtual gastroscopy and dynamic contrast-enhanced MPRs, a method that may overcome limitations in gastric cancer staging by revealing subtle mucosal changes with virtual gastroscopy and by providing detailed information on hemodynamic changes in gastric tumors, a choice of optimal imaging plane, and thinner section thicknesses with avoidance of partial-volume averaging effects. Indeed, our tumor detection rates with virtual gastroscopy and MPR images and our accuracy rates in T and N staging were superior to those previously reported (5,17,18,21,22); this fact should help clinicians choose more optimal treatment modalities for individual cases.

Because we achieved good distention of the stomach with air and water and used rapid imaging, patients did not require intravenous spasmolytic agents before multi–detector row CT. Virtual gastroscopy from images obtained in air-distended stomachs provides an excellent overview of abnormal mucosal lesions within the stomach lumen (23). It is helpful for detection of type IIa, IIc, and III early gastric cancer mucosal lesions. These lesions are usually missed at CT due to absence of thickening of the gastric wall. One limitation of virtual gastroscopy is that it could not depict flat type IIb early gastric tumors owing to the absence of morphologic change in the gastric mucosa and a lack of color change in the flat lesions; color change can be demonstrated on images from conventional but not virtual gastroscopy. Lee and Ko (10) reported that three-dimensional imaging with the surface-shaded display technique improved early gastric cancer detection rates from 64.5% to 93.5%. Recently, Kim et al (18) reported that seven of 14 early gastric cancer lesions were seen only at virtual gastroscopy; in our study, the detection rate with virtual gastroscopy and MPR images was 78% (seven of nine lesions) and was superior to that achieved with dynamic contrast-enhanced transverse images alone (44% [four of nine lesions]).

Owing to their hypervascularity (neovascularity) on dynamic contrast-enhanced CT images, most gastric cancers are seen as enhancing lesions (24,25). According to Lee et al (8), helical CT with two-phase scanning, particularly the mucosal phase (38–45 seconds), is effective for identifying the unique enhancement patterns of early gastric cancer. Takao et al (26) showed that the earlier phase (45 seconds) is accurate for determining the depth of tumor invasion through the wall; the later phase (3 minutes) did not provide additional information. In our study, contrast-enhanced images obtained during the late arterial phase (40 seconds) proved useful for tumor detection—especially those tumors without an obviously thickened wall. The delayed phase (150 seconds) was particularly helpful for distinguishing T1 from T2 tumors. In this study, good contrast enhancement of the lymph nodes was achieved in the portal venous phase (70 seconds). Hundt et al (27) reported that contrast enhancement of the lymph nodes was better shown in the later phase (60 seconds). In the studies of Kim et al (18,28), to assess tumor staging, including distant metastases, scanning was performed 72 seconds after initiation of the contrast material injection, which corresponds to the venous phase of the liver. Therefore, contrast-enhanced CT images obtained during the portal venous phase (70 seconds) are helpful for N and M staging.

The use of MPRs allows the radiologist to choose an optimal imaging plane to accurately evaluate the depth of tumor invasion of the gastric wall and to identify the thin fat plane between the tumor and adjacent organs to avoid the partial-volume averaging effect. In our study, MPRs were superior to transverse images in T staging not only because of thinner sections but also because of optimal imaging planes. In our study, three cases were missed on transverse images, even when sections as thin as 1.25 mm were used. However, the MPR images clearly identified these lesions. The accuracy of our dynamic contrast-enhanced MPR images (overall accuracy in T staging, 89% [50 of 55 neoplasms]) was superior to accuracy values in previous reports (5,17,18,21,22,26,27). Gastric cancers can be located on the horizontally oriented portions of the gastric wall, such as the lesser or greater curvature. Cancers on the superior or inferior wall of the gastric antrum are difficult to detect and to stage correctly because of poor z-axis resolution and partial-volume averaging effects (26,27). Although MPRs are useful for tumor staging, MPRs may have intrinsic limitations, such as in imaging T1 cancers with massive submucosal invasion of cancer cells, T1 cancers in the absence of a thickened gastric wall, T2 cancers with perigastric inflammation or vascular or lymphatic engorgement, and T3 cancers with minimal infiltration of cancer cells into the perigastric adipose tissue (28).

Regarding N staging, we used the new criteria of the American Joint Committee on Cancer and the International Union Against Cancer (20), which differ from criteria used by earlier investigators (5,14,17,23,26–28), who followed Japanese guidelines defined in the General Rules for Gastric Cancer Study in Surgery and Pathology (29,30). Ichikura et al (31) reported that the American Joint Committee on Cancer/International Union Against Cancer system appeared to be a superior staging system. Prognosis in gastric cancer depends on lymph node involvement and adequate lymphadenectomy (2,32,33). CT is relatively insensitive and also nonspecific for detecting nodal metastases due to its inability to detect microscopic nodal invasion, which is common in gastric cancer, and the presence of reactive nodes that may be greater than 10 mm (34).

Our MPR images seemed slightly better than transverse images for N staging, although this difference was not statistically significant. MPR images can provide more accurate measurement of lymph node size, and better differentiation between lymph nodes and small perigastric vessels is possible. Our study results showed that there was good correlation between pathologic and CT findings in N staging; this was probably due to our using a modified N staging system. Metastatic lymph nodes were more common, and those with small short-to-long axis ratios demonstrated increased attenuation on contrast-enhanced images (19).

An analysis of 1082 nodes resected in patients with gastric cancer who were examined preoperatively with spiral CT (5-mm section thickness) (19) revealed that 21% of nodes 5–9 mm, 23% of nodes 10–14 mm, and 82% of nodes greater than 14 mm were positive. However, another study of nodes resected at surgery (35) indicated that 55% of metastatic lymph nodes were less than 5 mm in diameter. Clinical studies (33,36,37) revealed that extended resection for locally advanced gastric cancers offers better control and survival. The use of our criteria would improve sensitivity but reduce specificity for detecting diseased lymph nodes. We believe that high sensitivity is more important because understaging lymph node involvement may lead to undertreatment, which will result in increased morbidity and mortality. In our study, the transverse and MPR images, respectively, had overall sensitivities of 86% (32 of 37 neoplasms) and 92% (34 of 37 neoplasms), with specificities of 78% (14 of 18 neoplasms) and 72% (13 of 18 neoplasms), in depicting pathologic lymph nodes.

Our study had some limitations. First, the triple-phase dynamic technique would somewhat increase radiation dose. To minimize radiation dose, we limited the scan ranges just enough to cover the lesion sites in postcontrast late arterial and delayed phases. Second, we used 1.25-mm near-isotropic voxels instead of 0.625-mm isotropic voxels. The use of isotropic voxels may improve MPR images, as well as virtual gastroscopy images. However, we found that using 1.25-mm near-isotropic voxels produced sufficient image quality with our CT protocol. Moreover, the scanning time and the radiation dose to the patient can be reduced. Last, the patient population in this study was relatively small, so further studies with a larger number of patients are needed.

In conclusion, use of the combination of virtual gastroscopy and dynamic contrast-enhanced MPR images obtained at multi–detector row CT after air and water distention of the stomach can improve tumor detection rates as well as accuracy rates in T and N staging of gastric cancers.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
References
 


   FOOTNOTES
 

Abbreviations: MPR = multiplanar reformation

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, C.Y.C., G.C.L.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, all authors; clinical studies, C.Y.C., J.S.H., D.C.W., W.Y.K., J.S.H., T.S.J., G.C.L.; statistical analysis, C.Y.C., M.T.W.; and manuscript editing, C.Y.C., G.C.L.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 

  1. Forman D, Goodman KJ. The epidemiology of stomach cancer: correlating the past with the present—socioeconomic influences in early life can influence mortality in adult life. BMJ 2000;320:1682–1683.[Free Full Text]
  2. Msika S. Lymph node excision in cancer of the stomach [in French]. J Chir (Paris) 1998;135:155–161.[Medline]
  3. Gill S, Shah A, Le N, Cook EF, Yoshida EM. Asian ethnicity-related differences in gastric cancer presentation and outcome among patients treated at a Canadian cancer center. J Clin Oncol 2003;21:2070–2076.[Abstract/Free Full Text]
  4. Ohashi S, Segawa K, Okamura S, et al. The utility of endoscopic ultrasonography and endoscopy in the endoscopic mucosal resection of early gastric cancer. Gut 1999;45:599–604.[Abstract/Free Full Text]
  5. Habermann CR, Weiss F, Riecken R, et al. Preoperative staging of gastric adenocarcinoma: comparison of helical CT and endoscopic US. Radiology 2004;230:465–471.[Abstract/Free Full Text]
  6. Winter TC, Ager JD, Nghiem HV, Hill RS, Harrison SD, Freeny PC. Upper gastrointestinal tract and abdomen: water as an orally administered contrast agent for helical CT. Radiology 1996;201:365–370.[Abstract/Free Full Text]
  7. Hahn FJ, Jacoby CG, Ehrhardt JC. The overshoot artifact as an aid in differentiating low densities on the computed tomographic scan. J Comput Tomogr 1978;2:149–151.[CrossRef][Medline]
  8. Lee JH, Jeong YK, Kim DH, et al. Two-phase helical CT for detection of early gastric carcinoma: importance of the mucosal phase for analysis of the abnormal mucosal layer. J Comput Assist Tomogr 2000;24:777–782.[CrossRef][Medline]
  9. Lee DH. Three-dimensional imaging of the stomach by spiral CT. J Comput Assist Tomogr 1998;22:52–58.[CrossRef][Medline]
  10. Lee DH, Ko YT. The role of 3D spiral CT in early gastric carcinoma. J Comput Assist Tomogr 1998;22:709–713.[CrossRef][Medline]
  11. Lee DH, Ko YT. Advanced gastric carcinoma: the role of three-dimensional and axial imaging by spiral CT. Abdom Imaging 1999;24:111–116.[CrossRef][Medline]
  12. Shin SS, Kang HK, Jeong YY, et al. Virtual gastroscopy using spiral CT in gastric lesions. J Korean Radiol Soc 1998;39:947–952.
  13. Rustgi AK. Gastrointestinal cancers: a companion to Sleisenger and Fordtran's gastrointestinal and liver disease. Edinburgh, Scotland: Saunders, 2003.
  14. Japanese Gastric Cancer Association. Japanese classification of gastric carcinoma—2nd english edition. Gastric Cancer 1998;1:10–24.
  15. Borrmann R. Geshwulste des magens und duodenums. In: Henke F, Lubarsch O, eds. Handbuch der Speziellen Pathologischen Anatomieund Histologis. Berlin, Germany: Springer-Verlag, 1926; 865.
  16. Minami M, Kawauchi N, Itai Y, Niki T, Sasaki Y. Gastric tumors: radiologic-pathologic correlation and accuracy of T staging with dynamic CT. Radiology 1992;185:173–178.[Abstract/Free Full Text]
  17. D'Elia F, Zingarelli A, Palli D, Grani M. Hydro-dynamic CT preoperative staging of gastric cancer: correlation with pathological findings—a prospective study of 107 cases. Eur Radiol 2000;10:1877–1885.
  18. Kim AY, Kim HJ, Ha HK. Gastric cancer by multidetector row CT: preoperative staging. Abdom Imaging 2005;30:465–472.[CrossRef][Medline]
  19. Fukuya T, Honda H, Hayashi T, et al. Lymph-node metastases: efficacy for detection with helical CT in patients with gastric cancer. Radiology 1995;197:705–711.[Abstract/Free Full Text]
  20. Greene FL, Page DL, Fleming ID, et al, eds. AJCC manual of staging of cancer. 6th ed. New York, NY: Springer-Verlag, 2002.
  21. Kim HS, Han HY, Choi JA, et al. Preoperative evaluation of gastric cancer: value of spiral CT during gastric arteriography (CTGA). Abdom Imaging 2001;26:123–130.[CrossRef][Medline]
  22. Fukuya T, Honda H, Kaneko K, et al. Efficacy of helical CT in T-staging of gastric cancer. J Comput Assist Tomogr 1997;21:73–81.[CrossRef][Medline]
  23. Lee DH. Two-dimensional and three-dimensional imaging of gastric tumors using spiral CT. Abdom Imaging 2000;25:1–6.[CrossRef][Medline]
  24. Efsen F, Fischerman K. Angiography in gastric tumours. Acta Radiol Diagn (Stockh) 1974;15:193–197.[Medline]
  25. Reuter SR, Redman HC, Miller WJ, Hoskins PA. Gastric angiography. Radiology 1970;94:271–276.[Medline]
  26. Takao M, Fukuda T, Iwanaga S, Hayashi K, Kusano H, Okudaira S. Gastric cancer: evaluation of triphasic spiral CT and radiologic-pathologic correlation. J Comput Assist Tomogr 1998;22:288–294.[CrossRef][Medline]
  27. Hundt W, Braunschweig R, Reiser M. Assessment of gastric cancer: value of breathhold technique and two-phase spiral CT. Eur Radiol 1999;9:68–72.[CrossRef][Medline]
  28. Kim HJ, Kim AY, Oh ST, et al. Gastric cancer staging at multi-detector row CT gastrography: comparison of transverse and volumetric CT scanning. Radiology 2005;236:879–885.[Abstract/Free Full Text]
  29. Kajitani T. Japanese Research Society for Gastric Cancer: the general rules for the gastric cancer study in surgery and pathology. I. Clinical classification. Jpn J Surg 1981;11:127–139.
  30. Japanese Research Committee on Histological Classification of Gastric Cancer. The general rules for the gastric cancer study in surgery and pathology. II. Histological classification of gastric cancer. Jpn J Surg 1981;11:140–145.[CrossRef][Medline]
  31. Ichikura T, Tomimatsu S, Uefuji K, et al. Evaluation of the New American Joint Committee on Cancer/International Union Against Cancer classification of lymph node metastasis from gastric carcinoma in comparison with the Japanese classification. Cancer 1999;86:553–558.[CrossRef][Medline]
  32. Sierra A, Regueira FM, Hernandez-Lizoain JL, Pardo F, Martinez-Gonzalez MA, A-Cienfuegos J. Role of the extended lymphadenectomy in gastric cancer surgery: experience in a single institution. Ann Surg Oncol 2003;10:219–226.[Abstract/Free Full Text]
  33. Lee WJ, Lee WC, Houng SJ, et al. Survival after resection of gastric cancer and prognostic relevance of systematic lymph node dissection: twenty years experience in Taiwan. World J Surg 1995;19:707–713.[CrossRef][Medline]
  34. Cho JS, Kim JK, Rho SM, Lee HY, Jeong HY, Lee CS. Preoperative assessment of gastric carcinoma: value of two-phase dynamic CT with mechanical iv. injection of contrast material. AJR Am J Roentgenol 1994;163:69–75.
  35. Monig SP, Zirbes TK, Schroder W, et al. Staging of gastric cancer: correlation of lymph node size and metastatic infiltration. AJR Am J Roentgenol 1999;173:365–367.[Abstract/Free Full Text]
  36. Noguchi Y, Imada T, Matsumoto A, Coit DG, Brennan MF. Radical surgery for gastric cancer: a review of the Japanese experience. Cancer 1989;64:2053–2062.[CrossRef][Medline]
  37. Maruyama K, Gunven P, Okabayashi K, Sasako M, Kinoshita T. Lymph node metastases of gastric cancer: general pattern in 1931 patients. Ann Surg 1989;210:596–602.[Medline]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
C.-Y. Chen, D.-C. Wu, Y.-T. Kuo, C.-H. Lee, T.-S. Jaw, W.-Y. Kang, and J.-S. Hsu
MDCT for Differentiation of Category T1 and T2 Malignant Lesions from Benign Gastric Ulcers
Am. J. Roentgenol., June 1, 2008; 190(6): 1505 - 1511.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Roentgenol.Home page
S. J. Kim, J. M. Lee, J. Y. Lee, J. Y. Choi, S. H. Kim, J. K. Han, and B. I. Choi
Accuracy of Preoperative T-Staging of Gallbladder Carcinoma Using MDCT
Am. J. Roentgenol., January 1, 2008; 190(1): 74 - 80.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
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 Chen, C.-Y.
Right arrow Articles by Liu, G.-C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, C.-Y.
Right arrow Articles by Liu, G.-C.

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