- Email: [email protected]
T-staging of gastric cancer of air-filling multidetector-row CT: Comparison with hydro-multidetector-row CT
European Journal of Radiology 81 (2012) 2953–2960
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
T-staging of gastric cancer of air-filling multidetector-row CT: Comparison with hydro-multidetector-row CT Seishi Kumano ∗ , Masahiro Okada, Taro Shimono, Masatomo Kuwabara, Yukinobu Yagyu, Izumi Imaoka, Ryuichiro Ashikaga, Kazunari Ishii, Takamichi Murakami Department of Radiology, Kinki University School of Medicine, Osakasayama, Osaka 589-8511, Japan
a r t i c l e
i n f o
Article history: Received 2 August 2011 Received in revised form 22 December 2011 Accepted 23 December 2011 Keywords: Gastric cancer T-staging Air Water MDCT
a b s t r a c t Purpose: The purpose of this study was to evaluate the accuracy of T-staging of gastric cancer by air-filling multidetector-row CT (air-MDCT) compared with water-filling MDCT (hydro-MDCT). Materials and methods: One hundred fifteen patients with histologically diagnosed gastric cancer were included in this study. Fifty-eight patients underwent air-MDCT, and the remaining 57 had hydro-MDCT using a 64-channel scanner. Based on the volumetric data of contrast-enhanced MDCT obtained about 75 s after intravenously injecting 525 mg iodine per kilogram patients weight (525 mgI/kg) nonionic contrast material at the rate of 2 ml/s, oblique coronal and oblique sagittal multi-planar reformatted images perpendicular to the stomach wall, including the tumor, were reconstructed on a workstation. Mural invasion of gastric cancer into the gastric wall, as visualized by CT, was classified according to the TNM classification, and the results of T-staging by MDCT were compared with those by pathologic analysis after surgery. Results: Correct assessment of T-staging by air-CT was achieved in 48 of 58 patients (83%), and that by hydro-MDCT was 49 of 57 patients (86%). The sensitivity, specificity, and accuracy of the technique in determining the invasion of serosa were 88%, 93%, and 91% for air-CT and 83%, 95%, and 91% for hydroCT. There were no significant differences between hydro-MDCT and air-MDCT in sensitivity (P = 0.73), specificity (P = 0.71) and accuracy (P = 0.98). Conclusion: Air-MDCT is a very valuable tool in T-staging of gastric cancer as well as hydro-MDCT. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Gastric cancer is one of the major causes of death from malignant disease worldwide [1,2]. In this situation, the early detection and accurate preoperative staging of gastric cancer is critical. In particular, evaluation of the local extent of gastric cancer (the so-called “T-stage”) is of pivotal importance in the choice of the optimal therapeutic strategy. In particular, depiction of serosal invasion is important, because serosal involvement has been demonstrated to be a poor prognostic factor [3,4]. Although double-contrast barium examination and endoscopy of the stomach are very useful to detect early gastric cancer [5], the T-stage of gastric cancer and the presence or absence of metastasis cannot be determined with these modalities; therefore, preoperative staging of gastric cancer is performed mainly
∗ Corresponding author. Tel.: +81 72 366 0221; fax: +81 72 367 1685. E-mail addresses: [email protected], [email protected] (S. Kumano). 0720-048X/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.12.039
with computed tomography (CT). Conventional and single-detector helical CT with water filling of the stomach has been reported to provide more detailed T-staging [6–9]. Currently, multi-detector row CT (MDCT) with thinner collimation offers near-isotropic imaging of the stomach and allows high-quality multiplanar reformation (MPR) [10–12]. With adequate distention of the stomach using water for negative contrast, water-filling MDCT (hydro-MDCT) images offer superior differentiation of tumor tissue from normal mucosa and this method avoids overshooting artifacts due to intraluminal air obscuring the normal gastric wall pattern [13–16]. On the other hand, with air distention of the stomach, virtual gastrography images allow evaluation of gastric endoluminal disease and there are several reports that air-filling MDCT (air-MDCT) is also useful for the detection and T-staging of gastric cancer [17–19]. To our knowledge, no study has compared hydro-MDCT and air-MDCT to assess the depth of mural invasion in gastric cancer. Thus, the purpose of this study was to evaluate the accuracy of air-MDCT for T-staging of gastric cancer compared with hydroMDCT.
2954
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
2. Materials and methods
Table 1 Comparison of patient features with hydro-MDCT versus air-MDCT.
2.1. Patients This prospective study received institutional review board approval and written informed consent to participate was obtained from all patients. Between December 2006 and September 2008, one hundred fifty-one consecutive patients, who had been diagnosed with gastric cancer histologically by endoscopic biopsy and had undergone MDCT for international TNM classification [20] of the tumor before surgery, were included in this study. Of these 151 patients, 20 were excluded because they did not undergo surgical resection, endoscopic mucosal resection or laparoscopic examination with cytology because of disseminated peritoneal seeding (n = 6), distant metastasis (n = 12) or direct invasion of an adjacent organ (n = 2). Of 131 patients, two were excluded because the interval between CT and surgery was over 40 days (49 days and 62 days). Of 129 patients, 14 were excluded because of negative histology on staging laparoscopy. The final study population consisted of 115 patients who had undergone surgery (n = 92), staging laparoscopy with positive histology (n = 18), and endoscopic mucosal resection (n = 5). They included 101 men and 28 women ranging in age from 33 to 81 years (mean: 64 years). The mean interval between CT and surgery, staging laparoscopy or endoscopic mucosal resection was 22 days (range, 1–37 days). In patients who underwent surgical resection or endoscopic mucosal resection, the resected stomach was submitted for pathologic study, and the area of gastric cancer along with the surrounding normal-appearing gastric wall was microscopically examined to determine the depth of tumor penetration. On the other hand, in the other patients who were finally diagnosed
No. Gender Male Female Mean age Surgery Endoscopic resection Staging laparoscopy with positive histology Interval T1 M SM1 SM2 T2–3 T2 T3 T4a T4b Tumor location Fundus Body Antrum
Hydro-CT
Air-CT
57
58
115
43 14 62 (33–79) 46 1 10
48 10 65 (50–80) 46 4 8
91 24 63 92 5 18
22 (1–37) 27 16 2 9 12 5 7 17 1
22 (4–36) 26 11 6 9 16 5 11 15 1
22 53 27 8 18 28 10 18 32 2
10 30 17
8 36 14
18 66 31
M: limited to mucosal layer, SM1: minimal invasion into submucosal layer, SM2 massive invasion into submucosal layer.
by staging laparoscopy, the serosal membrane was macroscopically examined by surgeons and also the cytology of peritoneal lavage fluid and peritoneal metastasis was examined by a pathologist [21–23]. In this study, staging laparoscopy was performed
Fig. 1. Images obtained in a 53-year-old man with stage T1 gastric tumor overstaged as T2–3. (a) and (b) Coronal and sagittal CT images show enhancement with entire wall thickening (arrow). This patient was overstaged as T2–3. (c) On histological micrograph, the tumor shows massive submucosal invasion (arrow), but not muscular propria invasion (hematoxylin–eosin stain; low-power magnification).
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
2955
Table 2a Comparison of T staging with hydro-MDCT versus T staging with surgical pathology. Pathologic findings
T1 T2–3 T4a T4b
MDCT findings T1
T2–3
T4a
25 1
2 9 3
2 14
T4b
1
in 26 patients with advanced gastric cancer and 8 patients underwent total or partial gastrectomy after staging laparoscopy. Six of eight patients had subserosal invasion, one patient had muscularis propria invasion, and the other had extraserosal invasion. Of 17 patients diagnosed as T4a in staging laparoscopy, all the patients had positive cytology and/or peritoneal metastasis. Pathologic T-staging was based on the international TNM classification [20] as follows: p-T1, tumor invades lamina propria, muscularis mucosae, or submucosal; p-T2, tumor invades muscularis propria; p-T3, tumor invades subserosa; p-T4a, tumor perforates serosa; p-T4b, tumor invades adjacent structures. According to the Japanese classification of gastric cancer [24], T1 gastric cancers were classified into three grades as follows: M (limited to mucosal layer), SM1 (minimal invasion into submucosal layer), SM2 (massive invasion into submucosal layer). Pathologically, the tumors were classified as pT1 in 53 patients (M;28, SM1;8, SM2;17), pT2 in 10 patients, pT3 in 18 patients, pT4a in 32 patients, and pT4b in 2 patients. Tumor locations were at fundus in 18 patients, body in 66 patients, and antrum in 31 patients. Patients were randomly assigned to one of two protocols; the water-filling method (n = 57) and air-filling method (n = 58). Between the two patient groups, there were no significant differences with respect to patient age (P = 0.17, by Student’s t-test), sex distribution, pathologic T-staging distribution, tumor location, number of surgeries, staging laparoscopy, endoscopic mucosal resection (P = 0.33, 0.87, 0.59, 0.37, respectively, by the chi-square test), or the mean interval of surgery, staging laparoscopy or endoscopic mucosal resection (P = 0.49, by Student’s t-test) (Table 1).
Fig. 2. Images obtained in a 80-year-old woman with stage T2–3 gastric tumor. (a) MPR CT image shows enhancement with entire wall thickening (white arrow). (b) Histological micrograph shows a tumor involving the muscular propria layers (black arrow) (hematoxylin–eosin stain; low-power magnification).
2.2. MDCT protocol CT examinations were performed using a commercially available MDCT scanner (LightSpeed VCT; GE Medical System, Milwaukee). Scopolamine-N-butyl bromide (Buscopan; Boehringer Ingelhein) was not used in any patients before the examination. CT scan, from the level of the dome of the right hemidiaphragm to the lower edge of the stomach to include the entire liver, was started 75 s after the initiation of intravenous injection of 525 mgI/kg (Iohexol Omnipaque, Tokyo, Japan, Iopamidol Iopamiron, Tokyo, Japan) nonionic contrast material at the rate of 2 ml/s using a power injector (Autoenhance; Nemoto-hyoundo, Tokyo, Japan). CT scan parameters were as follows: detector-row configuration of 0.625 mm × 64 mm, pitch of 0.984, reconstruction interval of 0.625 mm, tube current of 200 mA, tube voltage of
120 kV and tube rotation time, 0.4 s. Based on the volumetric data of MDCT, axial images of 0.625-mm section thickness at 0.625-mm intervals were reconstructed and sent to the workstation. 2.3. Water-filling MDCT protocol (hydro-MDCT) In the hydro-MDCT protocol, all patients were intended to drink 600 ml tap water to distend the stomach wall. When they could not ingest 600 ml water, the amount of water intake was recorded for these patients. All patients were placed on the scanning table in a prone position to avoid artifacts caused by air in the stomach [12]; however, when the lesion was in the gastric cardia or fundus, as determined by barium or endoscopic examinations, the patient was placed in the supine position for CT scanning.
Table 2b Comparison of T staging with air-MDCT versus T staging with surgical pathology. Pathologic findings
T1 T2–3 T4a T4b
MDCT findings T1
T2–3
T4a
22 1
4 12 2
3 13
T4b
1
Table 3 Results of diagnosing serosal invasion with MDCT.
Sensitivity Specificity Accuracy
Hydro-MDCT
Air-MDCT
83% 95% 91%
88% 93% 91%
2956
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
Fig. 3. Images in a 73-year-old man with stage T3 gastric tumor overstaged as T4a. (a) and (b) Transverse CT and coronal image shows that the entire stomach wall is enhanced abnormally accompanied by wall thickening (arrow), and that a linear or reticular structure is observed in the fat layer of the surrounding stomach. (c) Histological micrograph shows subserosal invasion, but not extraserosal invasion (arrow) (hematoxylin–eosin stain; low-power magnification).
2.4. Air-filling MDCT protocol (air-MDCT)
2.6. Image interpretation
In the air-MDCT protocol, all patients were given 5 g gasproducing crystals with a small amount of water and placed in the supine position.
Two experienced radiologists (T.S., M.K.) with 17 and 13 years of experience of abdominal CT, respectively, performed the image analysis. They were unaware of the results of the upper gastrointestinal series, endoscopy, and surgical or histopathological study of the resected specimen, and independently evaluated CT images at a workstation (Virtual Place; AZE, Tokyo, Japan). After confirming the tumor location, oblique coronal and oblique MPR images perpendicular to the stomach wall, including the tumor, were reconstructed on a workstation, and the depth of tumor invasion was evaluated with a projection vertical to the tumor to avoid partial volume effects. Differences in assessment were resolved by consensus. We evaluated inter-observer agreement from differences in the assessment for T-staging of gastric cancer.
2.5. CT staging criteria of gastric cancer Mural invasion of gastric cancer in the gastric wall visualized by CT was classified as follows [11]:
T1: Strong enhancement with focal thickening is present in the inner and/or middle layer, but the outer layer shows no enhancement. Enhancement is present in the stomach wall, but the stomach wall is not thickened. The entire stomach wall does not abnormally enhance and there is no evidence of transmural tumoral involvement. T2–3: The entire stomach wall is enhanced abnormally accompanied by wall thickening, but the surface of the outer layer in contact with the fatty layer surrounding the stomach is smooth. T4a: The entire stomach wall is enhanced abnormally accompanied by wall thickening, and linear or reticular structures are observed in the fatty layer surrounding the stomach. T4b: The changes described above extend to adjacent contiguous organs.
2.7. Statistical analysis Fisher’s exact test was used to compare the accuracy of T-staging by air-MDCT with hydro-MDCT. A P-value of 0.05 was considered significant.
3. Results In hydro-MDCT, 45 of 57 patients could ingest 600 ml tap water, but 12 patients could not. These 12 patients drank about 500 ml tap water on average and the minimum amount of ingested water was
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
2957
Fig. 4. Images obtained in a 45-year-old woman with T3 gastric cancer overstaged as T4a. (a) and (b) Transverse CT and coronal images show that the entire stomach wall is enhanced abnormally accompanied by wall thickening (arrow), and that linear or reticular structures are observed in the fat layer of the surrounding stomach. (c) Histological micrograph shows a wide subserosal layer and tumor invading the subserosal layer with an irregular surface to the fat layer, but not reaching the serosal membrane (hematoxylin–eosin stain; low-power magnification).
about 400 ml. On the other hand, in air-MDCT, all patients could ingest 5 g gas-producing crystals. Correct assessment of the depth of mural invasion of the gastric wall (T stage) was 83% and 86% for hydro-MDCT and air-MDCT, respectively, which means that 49 of 57 patients with hydroMDCT and 48 of 58 patients with air-MDCT were correctly staged (Tables 2a and 2b). There were no significant differences between the two techniques in accuracy (P = 0.80). The sensitivity, specificity, and accuracy of the technique in determining the invasion of serosa were 83%, 95%, and 91% for hydro-MDCT and 88%, 93%, and 91% for air-MDCT (Table 3). There were no significant differences between them in sensitivity (P = 0.73), in specificity (P = 0.71) and in accuracy (P = 0.98).
3.1. T1 stage tumors Correct assessment of T1 tumor for hydro-MDCT and air-MDCT was 25/27 (92%) and 22/26 (85%), respectively. Four cases in airMDCT and two cases in hydro-MDCT were overstaged as T2–3. One T1 tumor in air-MDCT was a large protruding mass with contrast enhancement at MDCT. This tumor showed massive tumor formation of mucosa and minimal invasion of submucosa (SM1), but not invasion into the muscular layers on pathology. The other five (two in hydro-MDCT and three in air-MDCT) T1 tumors had entire enhancement accompanied by wall thickening on CT (Fig. 1). These five T1 tumors overstaged as T2–3 showed massive invasion into the submucosal layer (SM2) on pathology. However there were no significant differences in the staging of T1 gastric cancer between
the two methods (P = 0.42), the overstaged rate of the T1 tumor as T2–3 in air-MDCT tended to be higher than in hydro-MDCT. 3.2. T2–3 stage tumors Correct assessment of T2–3 tumors by hydro-MDCT and airMDCT was 9/12 (75%) and 12/16 (75%), respectively (Fig. 2), with no significant differences between the two methods (P = 1). One T3 tumor understaged as T1 on hydro-MDCT showed invasion into subserosa on histology. Even retrospectively on CT images, we thought that this tumor was very difficult to stage correctly as T3 because it had patchy enhancement of the inner and middle layers, but did not have entire wall thickening and enhancement. One T2 tumor understaged as T1 on air-MDCT existed at the pylorus, but could not be detected by two readers because they misinterpreted the pylorus tumor as normal gastric wall thickening. The two T3 tumors in hydro-MDCT and the three T3 tumors in air-MDCT were overstaged as T4a (Figs. 3 and 4). All tumors showed invasion into the subserosal layer. There were no T2 cases overstaged as T4a in either method. In these tumors, strandlike structures extending from the outer gastric surface into the surrounding fat plane were seen and considered suggestive of serosal involvement. At pathology, two T3 tumors had inflammatory change at the serosal membrane with an irregular outer surface, two T3 tumors had a relatively wide subserosal fat layer and tumor invaded this subserosal fat tissue with interface irregularity between the tumor and subserosal fat, but the tumor was surrounded by the serosal membrane. One T3
2958
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
Fig. 5. Images obtained in a 73-year-old man with stage T4a gastric tumor. (a) and (b) Transverse CT and coronal images show that the entire stomach wall is enhanced abnormally accompanied by wall thickening (arrow), and linear or reticular structures are observed in the fat layer of the surrounding stomach. (c) Histological micrograph shows extraserosal invasion (arrow) (hematoxylin–eosin stain; low-power magnification).
tumor had small lymph node metastasis adjacent to the serosal membrane.
3.3. T4a stage tumors Correct assessment of T4a tumor by hydro-MDCT and air-MDCT was 14/17 (82%) and 13/15 (87%), respectively (Fig. 5). There were no significant differences between the two methods (P = 0.74). Three T4a tumors in hydro-MDCT and two T4a tumors in airMDCT were understaged as T2–3 (Figs. 6 and 7). Of three T4a tumors understaged T2–3 in hydro-MDCT, 2 patients who had these tumors cannot ingest 600 ml water (average about 500 ml). By CT, the surface of the outer layer of the stomach wall, including the lesion, was smooth. These T4a tumors understaged as T2–3 often showed small enhanced lesions at CT. Pathological findings revealed that these tumors had minimal infiltration of cancer cells into the perigastric adipose tissue without macroscopic mass formation. In the five T4a tumors understaged as T2–3, all five tumors could be evaluated histologically.
3.4. T4b stage tumors One T4b tumor in hydro-MDCT and one T4b tumor in air-MDCT were both correctly staged because of the absence of normal fat cleavage planes with surrounding organs at CT. The T4b tumor in hydro-MDCT showed invasion into the left diaphragm by open surgery and the other in air-MDCT had confirmed invasion into the colon mesentrium and pancreas head by staging laparoscopy.
3.5. Interobserver agreement We used blinded consensus reading to resolve disagreements between the interpreting radiologists. In the evaluation of Tstaging, there was agreement in 48 of 58 patients (81%) with air-MDCT and in 46 of 57 patients (81%) with hydro-MDCT. 4. Discussion In preoperative CT examinations for T-staging of gastric cancer, there are many reports that hydro-CT and -MDCT are useful because of their excellent depiction of the gastric wall as three layers [6–13]. Air-MDCT allows 3D virtual gastrography, which improves the detectability of gastric cancer, especially early gastric cancer [17,18,25], but it is said that this technique is limited for evaluating T-staging of gastric cancer due to overshooting artifacts between the gastric wall and air in the stomach [14]. However, as the overshooting artifacts are expected to be reduced thanks to the advancement of multiplication of CT detector row, we compared air-MDCT and hydro-MDCT using a 64-channel system. In our present study, the diagnostic accuracy of T-staging of gastric cancer by air-MDCT was as good as by hydro-MDCT, which could explain why 64-channel MDCT can provide not only high spatial resolution of the z-axis but also high time resolution. The diagnostic accuracy of serosal invasion by air-MDCT was almost equal to that by hydro-MDCT and the sensitivity of serosal invasion by air-MDCT was superior to hydro-MDCT, although there were no significant differences between the two techniques. Diagnosis of serosal invasion is also very important, because serosal invasion of gastric cancer is a poor prognosis factor. Serosal invasion of gastric cancer underwent a trial treatment of
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
Fig. 6. Images obtained in a 62-year-old man with T4a gastric tumor understaged as T2–3. (a) Transverse and coronal CT images show gastric cancer on cardia. The tumor (arrow) was incorrectly staged as T2–3 because the surface of the outer layer appears smooth. (b) Histological microgram shows minimal infiltration of gastric cancer cells (arrow) into perigastric adipose tissue (hematoxylin–eosin stain; low-power magnification).
preoperative neoadjuvant chemotherapy for downstaging and curative operation. Staging laparoscopy has been widely used for the diagnosis of serosal invasion of gastric cancer [21–23], but this procedure is invasive and complicated. Our study showed that air-MDCT has the potential to indicate whether surgeons should perform staging laparoscopy. In our study, hydro-MDCT understaged T4a tumors as T2–3 more frequently than air-MDCT. T4a tumors, understaged as T2–3, showed that the outer surface of the stomach wall was smooth and regular, but there was minimal infiltration of tumor cells in the
2959
Fig. 7. Images obtained in a 2-year-old boy with T4a gastric tumor understaged as T2–3. (a) Transverse and coronal CT images show gastric cancer on cardia. Tumor (arrow) was incorrectly staged as T2–3 because the surface of the outer layer appears smooth. (b) Histological microgram shows minimal infiltration of gastric cancer cells (arrow) into perigastric adipose tissue (hematoxylin–eosin stain; low-power magnification).
perigastric adipose tissue at pathology. We speculated that insufficient distension of the stomach wall due to inappropriate intake of water made it difficult to evaluate invasion of the outer layer. In our study, approximately one fourth of patients could not drink 600 ml tap water. Good distension of the gastric wall is necessary to evaluate the T stage of gastric cancer, and there is a report that 800–1000 ml tap water was given to each patient [12], but realistically it is not easy for all patients to drink that much water. Air-MDCT is advantageous for T4a tumor staging due to good gastric wall distension. It is simpler and easier to perform air-MDCT than hydro-MDCT without the need to ingest water, and minimal
2960
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
invasion into the extraserosa seems to be limited by using 64channel MDCT in either hydro-MDCT or air-MDCT. In our series, three and two T3 tumors were overstaged as T4a in air-MDCT and in hydro-MDCT, respectively. Shimizu et al. reported that deformity of the outer surface of the gastric wall can occur as a result of factors other than serosal invasion, such as compression of the serosa by a tumor [10]. In T3 tumors overstaged as T4a gastric cancer, histologically, serosal inflammation without cancer cells in serosa was seen in 1 case, invasion into the wide subserosal layer with an irregular interface was seen in 2 cases, and small lymph node metastasis adjacent to extraserosa was seen in 1 case. These pathologic features of the T3 tumor overstaged as T4a were thought to be the most likely cause of deformity of the outer surface of the gastric wall. All T2–3 tumors overstaged as T4a were T3 tumor (subserosa (SS) invasion and no mascularis propria (MP) invasion). Either by air-MDCT or hydroMDCT, MDCT could correctly stage T2 tumors with invasion, but it might be limited to differentiate T3 tumors with SS and T4a tumors. Although there were no significant differences between the two techniques, air-MDCT tended to overstage T1 tumor as T2–3 compared to hydro-MDCT CT in our study. Hydro-MDCT can depict the three layers of the gastric wall more clearly than air-MDCT, because a gastric wall well-distended by air made the middle layer, consisting of the submucosal layer, unclear in air-MDCT. Mass-forming T1 gastric cancer with invasion into the submucosal layer tends to be overstaged as T2–3 in either hydro-MDCT or air-MDCT. One T2 tumors were understaged as T1 in our study, the case existed at the pylorus, which could not be distinguished from normal gastric wall thickening. There were problems evaluating the tumor at the antrum, which had a thickened wall due to insufficient distension by air-CT. Patient position might need to be changed on the CT table to the right side for good distension of the antrum. Our study has several limitations. First, we could not evaluate the two methods (hydro-MDCT and air-MDCT) in the same patients. Our investigation was a clinical study, so it was impossible to compare the two methods in the same patients; however, there were no significant differences in the background of the two patient groups. Second, we did not obtain a resected specimen in all T4a cases. However some of diagnosis was made by staging laparoscopy, all of the laparoscopic examinations staged for T4a, including the cytological evaluation of peritoneal lavage fluid and peritoneal metastasis, gave pathological confirmation. Moreover, there were no significant differences in the two patient groups in the number of T4a tumors diagnosed by staging laparoscopy, and we considered that our study could minimize the disadvantages. Third, we did not use 3D gastrography in air-MDCT. There have been reports that 3D gastrography can detect early gastric cancer more than 2D images [25]; however, our aim was to investigate the ability of MDCT to evaluate the depth of mural invasion, but not to show the detectability of gastric cancer on MDCT. This cannot be construed as a major limitation of our study. Forth, we did not perform the dynamic study and did not obtain the arterial phase, therefore we adopted the injection rate for 2 ml/s. The arterial phase is important to visualize the mucosal distortion, in the present study, it might underestimate the evaluation of early gastric cancer. In conclusion, air-MDCT is a valid examination for preoperative staging of patients with gastric cancer, and gave especially high accuracy in the assessment of serosal invasion of gastric cancer.
Conflict of interest None. References [1] Devesa SS, Blot WJ, Fraumeni Jr JF. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer 1998;83(10):2049–53. [2] Mishima Y, Hirayama R. The role of lymph node surgery in gastric cancer. World J Surg 1987;11(4):406–11. [3] Boku T, Nakane Y, Minoura T, Takada H, Yamamura M, Yamamoto M. Prognostic significance of serosal invasion and free intraperitoneal cancer cells in gastric cancer. Br J Surg 1990;77(4):436–9. [4] Abe S, Shiraishi M, Nagaoka S, Yoshimura H, Dhar DK, Nakamura T. Serosal invasion as the single prognostic indicator in stage IIIA (T3N1M0) gastric cancer. Surgery 1991;109(5):582–8. [5] Kurtz RC, Sherlock P. Carcinoma of the stomach. In: Berk JD, editor. Gastroenterology. Philadelphia: Saunders; 1985. p. 1278–304. [6] 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(1):69–75. [7] 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(1):173–8. [8] Habermann CR, Weiss F, Riecken R, et al. Preoperative staging of gastric adenocarcinoma: comparison of helical CT and endoscopic US. Radiology 2004;230(2):465–71. [9] Rossi M, Broglia L, Graziano P, et al. Local invasion of gastric cancer: CT findings and pathologic correlation using 5-mm incremental scanning, hypotonia, and water filling. AJR Am J Roentgenol 1999;172(2): 383–8. [10] Shimizu K, Ito K, Matsunaga N, Shimizu A, Kawakami Y. Diagnosis of gastric cancer with MDCT using the water-filling method and multiplanar reconstruction: CT–histologic correlation. AJR Am J Roentgenol 2005;185(5): 1152–8. [11] Kumano S, Murakami T, Kim T, et al. T staging of gastric cancer: role of multidetector row CT. Radiology 2005;237(3):961–6. [12] Chen CY, Hsu JS, Wu DC, et al. Gastric cancer: preoperative local staging with 3D multi-detector row CT—correlation with surgical and histopathologic results. Radiology 2007;242(2):472–82. [13] Kadowaki K, Murakami T, Yoshioka H, et al. Helical CT imaging of gastric cancer: normal wall appearance and the potential for staging. Radiat Med 2000;18(1):47–54. [14] 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(2):149–51. [15] Ba-Ssalamah A, Prokop M, Uffmann M, Pokieser P, Teleky B, Lechner G. Dedicated multidetector CT of the stomach: spectrum of diseases. Radiographics 2003;23(May–June (3)):625–44. [16] Park KJ, Lee MW, Koo JH. Detection of early gastric cancer using hydro-stomach CT: blinded vs unblinded analysis. World J Gastroenterol 2011;17(February (8)):1051–7. [17] Shin KS, Kim SH, Han JK, et al. Three-dimensional MDCT gastrography compared with axial CT for the detection of early gastric cancer. J Comput Assist Tomogr 2007;31(5):741–9. [18] Yang DM, Kim HC, Jin W, et al. 64 multidetector-row computed tomography for preoperative evaluation of gastric cancer: histological correlation. J Comput Assist Tomogr 2007;31(1):98–103. [19] 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(September (3)):879–85. [20] Sobin LH, Wittekind C, editors. TNM classification of malignant tumors. International Union Against Cancer (UICC). 7th ed. New York: Wiley; 2009. [21] D’Ugo DM, Persiani R, Caracciolo F, Ronconi P, Coco C, Picciocchi A. Selection of locally advanced gastric carcinoma by preoperative staging laparoscopy. Surg Endosc 1997;11(12):1159–62. [22] Blackshaw GRJC, Barry JD, Edwards P, Allison MC, Thomas GV, Lewis WG. Laparoscopy significantly improves the perceived preoperative stage of gastric cancer. Gastric Cancer 2003;6(4):225–9. [23] Chag L, Stefanidis D, Richardson WS, Earle DB, Fanelli RD. The role of staging laparoscopy for intraabdominal cancers: an evidence-based review. Surg Endosc 2009;23(2):231–41. [24] Japanese Gastric Cancer Association. Japanese classification of gastric carcinoma. 2nd English edition. Gastric Cancer 1998;1:10–24. [25] Shen Y, Kang HK, Jeong YY, et al. Evaluation of early gastric cancer at multi-detector CT with multiplanar reformation and virtual endoscopy. Radiographics 2011;31:189–99.
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
T-staging of gastric cancer of air-filling multidetector-row CT: Comparison with hydro-multidetector-row CT Seishi Kumano ∗ , Masahiro Okada, Taro Shimono, Masatomo Kuwabara, Yukinobu Yagyu, Izumi Imaoka, Ryuichiro Ashikaga, Kazunari Ishii, Takamichi Murakami Department of Radiology, Kinki University School of Medicine, Osakasayama, Osaka 589-8511, Japan
a r t i c l e
i n f o
Article history: Received 2 August 2011 Received in revised form 22 December 2011 Accepted 23 December 2011 Keywords: Gastric cancer T-staging Air Water MDCT
a b s t r a c t Purpose: The purpose of this study was to evaluate the accuracy of T-staging of gastric cancer by air-filling multidetector-row CT (air-MDCT) compared with water-filling MDCT (hydro-MDCT). Materials and methods: One hundred fifteen patients with histologically diagnosed gastric cancer were included in this study. Fifty-eight patients underwent air-MDCT, and the remaining 57 had hydro-MDCT using a 64-channel scanner. Based on the volumetric data of contrast-enhanced MDCT obtained about 75 s after intravenously injecting 525 mg iodine per kilogram patients weight (525 mgI/kg) nonionic contrast material at the rate of 2 ml/s, oblique coronal and oblique sagittal multi-planar reformatted images perpendicular to the stomach wall, including the tumor, were reconstructed on a workstation. Mural invasion of gastric cancer into the gastric wall, as visualized by CT, was classified according to the TNM classification, and the results of T-staging by MDCT were compared with those by pathologic analysis after surgery. Results: Correct assessment of T-staging by air-CT was achieved in 48 of 58 patients (83%), and that by hydro-MDCT was 49 of 57 patients (86%). The sensitivity, specificity, and accuracy of the technique in determining the invasion of serosa were 88%, 93%, and 91% for air-CT and 83%, 95%, and 91% for hydroCT. There were no significant differences between hydro-MDCT and air-MDCT in sensitivity (P = 0.73), specificity (P = 0.71) and accuracy (P = 0.98). Conclusion: Air-MDCT is a very valuable tool in T-staging of gastric cancer as well as hydro-MDCT. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Gastric cancer is one of the major causes of death from malignant disease worldwide [1,2]. In this situation, the early detection and accurate preoperative staging of gastric cancer is critical. In particular, evaluation of the local extent of gastric cancer (the so-called “T-stage”) is of pivotal importance in the choice of the optimal therapeutic strategy. In particular, depiction of serosal invasion is important, because serosal involvement has been demonstrated to be a poor prognostic factor [3,4]. Although double-contrast barium examination and endoscopy of the stomach are very useful to detect early gastric cancer [5], the T-stage of gastric cancer and the presence or absence of metastasis cannot be determined with these modalities; therefore, preoperative staging of gastric cancer is performed mainly
∗ Corresponding author. Tel.: +81 72 366 0221; fax: +81 72 367 1685. E-mail addresses: [email protected], [email protected] (S. Kumano). 0720-048X/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.12.039
with computed tomography (CT). Conventional and single-detector helical CT with water filling of the stomach has been reported to provide more detailed T-staging [6–9]. Currently, multi-detector row CT (MDCT) with thinner collimation offers near-isotropic imaging of the stomach and allows high-quality multiplanar reformation (MPR) [10–12]. With adequate distention of the stomach using water for negative contrast, water-filling MDCT (hydro-MDCT) images offer superior differentiation of tumor tissue from normal mucosa and this method avoids overshooting artifacts due to intraluminal air obscuring the normal gastric wall pattern [13–16]. On the other hand, with air distention of the stomach, virtual gastrography images allow evaluation of gastric endoluminal disease and there are several reports that air-filling MDCT (air-MDCT) is also useful for the detection and T-staging of gastric cancer [17–19]. To our knowledge, no study has compared hydro-MDCT and air-MDCT to assess the depth of mural invasion in gastric cancer. Thus, the purpose of this study was to evaluate the accuracy of air-MDCT for T-staging of gastric cancer compared with hydroMDCT.
2954
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
2. Materials and methods
Table 1 Comparison of patient features with hydro-MDCT versus air-MDCT.
2.1. Patients This prospective study received institutional review board approval and written informed consent to participate was obtained from all patients. Between December 2006 and September 2008, one hundred fifty-one consecutive patients, who had been diagnosed with gastric cancer histologically by endoscopic biopsy and had undergone MDCT for international TNM classification [20] of the tumor before surgery, were included in this study. Of these 151 patients, 20 were excluded because they did not undergo surgical resection, endoscopic mucosal resection or laparoscopic examination with cytology because of disseminated peritoneal seeding (n = 6), distant metastasis (n = 12) or direct invasion of an adjacent organ (n = 2). Of 131 patients, two were excluded because the interval between CT and surgery was over 40 days (49 days and 62 days). Of 129 patients, 14 were excluded because of negative histology on staging laparoscopy. The final study population consisted of 115 patients who had undergone surgery (n = 92), staging laparoscopy with positive histology (n = 18), and endoscopic mucosal resection (n = 5). They included 101 men and 28 women ranging in age from 33 to 81 years (mean: 64 years). The mean interval between CT and surgery, staging laparoscopy or endoscopic mucosal resection was 22 days (range, 1–37 days). In patients who underwent surgical resection or endoscopic mucosal resection, the resected stomach was submitted for pathologic study, and the area of gastric cancer along with the surrounding normal-appearing gastric wall was microscopically examined to determine the depth of tumor penetration. On the other hand, in the other patients who were finally diagnosed
No. Gender Male Female Mean age Surgery Endoscopic resection Staging laparoscopy with positive histology Interval T1 M SM1 SM2 T2–3 T2 T3 T4a T4b Tumor location Fundus Body Antrum
Hydro-CT
Air-CT
57
58
115
43 14 62 (33–79) 46 1 10
48 10 65 (50–80) 46 4 8
91 24 63 92 5 18
22 (1–37) 27 16 2 9 12 5 7 17 1
22 (4–36) 26 11 6 9 16 5 11 15 1
22 53 27 8 18 28 10 18 32 2
10 30 17
8 36 14
18 66 31
M: limited to mucosal layer, SM1: minimal invasion into submucosal layer, SM2 massive invasion into submucosal layer.
by staging laparoscopy, the serosal membrane was macroscopically examined by surgeons and also the cytology of peritoneal lavage fluid and peritoneal metastasis was examined by a pathologist [21–23]. In this study, staging laparoscopy was performed
Fig. 1. Images obtained in a 53-year-old man with stage T1 gastric tumor overstaged as T2–3. (a) and (b) Coronal and sagittal CT images show enhancement with entire wall thickening (arrow). This patient was overstaged as T2–3. (c) On histological micrograph, the tumor shows massive submucosal invasion (arrow), but not muscular propria invasion (hematoxylin–eosin stain; low-power magnification).
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
2955
Table 2a Comparison of T staging with hydro-MDCT versus T staging with surgical pathology. Pathologic findings
T1 T2–3 T4a T4b
MDCT findings T1
T2–3
T4a
25 1
2 9 3
2 14
T4b
1
in 26 patients with advanced gastric cancer and 8 patients underwent total or partial gastrectomy after staging laparoscopy. Six of eight patients had subserosal invasion, one patient had muscularis propria invasion, and the other had extraserosal invasion. Of 17 patients diagnosed as T4a in staging laparoscopy, all the patients had positive cytology and/or peritoneal metastasis. Pathologic T-staging was based on the international TNM classification [20] as follows: p-T1, tumor invades lamina propria, muscularis mucosae, or submucosal; p-T2, tumor invades muscularis propria; p-T3, tumor invades subserosa; p-T4a, tumor perforates serosa; p-T4b, tumor invades adjacent structures. According to the Japanese classification of gastric cancer [24], T1 gastric cancers were classified into three grades as follows: M (limited to mucosal layer), SM1 (minimal invasion into submucosal layer), SM2 (massive invasion into submucosal layer). Pathologically, the tumors were classified as pT1 in 53 patients (M;28, SM1;8, SM2;17), pT2 in 10 patients, pT3 in 18 patients, pT4a in 32 patients, and pT4b in 2 patients. Tumor locations were at fundus in 18 patients, body in 66 patients, and antrum in 31 patients. Patients were randomly assigned to one of two protocols; the water-filling method (n = 57) and air-filling method (n = 58). Between the two patient groups, there were no significant differences with respect to patient age (P = 0.17, by Student’s t-test), sex distribution, pathologic T-staging distribution, tumor location, number of surgeries, staging laparoscopy, endoscopic mucosal resection (P = 0.33, 0.87, 0.59, 0.37, respectively, by the chi-square test), or the mean interval of surgery, staging laparoscopy or endoscopic mucosal resection (P = 0.49, by Student’s t-test) (Table 1).
Fig. 2. Images obtained in a 80-year-old woman with stage T2–3 gastric tumor. (a) MPR CT image shows enhancement with entire wall thickening (white arrow). (b) Histological micrograph shows a tumor involving the muscular propria layers (black arrow) (hematoxylin–eosin stain; low-power magnification).
2.2. MDCT protocol CT examinations were performed using a commercially available MDCT scanner (LightSpeed VCT; GE Medical System, Milwaukee). Scopolamine-N-butyl bromide (Buscopan; Boehringer Ingelhein) was not used in any patients before the examination. CT scan, from the level of the dome of the right hemidiaphragm to the lower edge of the stomach to include the entire liver, was started 75 s after the initiation of intravenous injection of 525 mgI/kg (Iohexol Omnipaque, Tokyo, Japan, Iopamidol Iopamiron, Tokyo, Japan) nonionic contrast material at the rate of 2 ml/s using a power injector (Autoenhance; Nemoto-hyoundo, Tokyo, Japan). CT scan parameters were as follows: detector-row configuration of 0.625 mm × 64 mm, pitch of 0.984, reconstruction interval of 0.625 mm, tube current of 200 mA, tube voltage of
120 kV and tube rotation time, 0.4 s. Based on the volumetric data of MDCT, axial images of 0.625-mm section thickness at 0.625-mm intervals were reconstructed and sent to the workstation. 2.3. Water-filling MDCT protocol (hydro-MDCT) In the hydro-MDCT protocol, all patients were intended to drink 600 ml tap water to distend the stomach wall. When they could not ingest 600 ml water, the amount of water intake was recorded for these patients. All patients were placed on the scanning table in a prone position to avoid artifacts caused by air in the stomach [12]; however, when the lesion was in the gastric cardia or fundus, as determined by barium or endoscopic examinations, the patient was placed in the supine position for CT scanning.
Table 2b Comparison of T staging with air-MDCT versus T staging with surgical pathology. Pathologic findings
T1 T2–3 T4a T4b
MDCT findings T1
T2–3
T4a
22 1
4 12 2
3 13
T4b
1
Table 3 Results of diagnosing serosal invasion with MDCT.
Sensitivity Specificity Accuracy
Hydro-MDCT
Air-MDCT
83% 95% 91%
88% 93% 91%
2956
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
Fig. 3. Images in a 73-year-old man with stage T3 gastric tumor overstaged as T4a. (a) and (b) Transverse CT and coronal image shows that the entire stomach wall is enhanced abnormally accompanied by wall thickening (arrow), and that a linear or reticular structure is observed in the fat layer of the surrounding stomach. (c) Histological micrograph shows subserosal invasion, but not extraserosal invasion (arrow) (hematoxylin–eosin stain; low-power magnification).
2.4. Air-filling MDCT protocol (air-MDCT)
2.6. Image interpretation
In the air-MDCT protocol, all patients were given 5 g gasproducing crystals with a small amount of water and placed in the supine position.
Two experienced radiologists (T.S., M.K.) with 17 and 13 years of experience of abdominal CT, respectively, performed the image analysis. They were unaware of the results of the upper gastrointestinal series, endoscopy, and surgical or histopathological study of the resected specimen, and independently evaluated CT images at a workstation (Virtual Place; AZE, Tokyo, Japan). After confirming the tumor location, oblique coronal and oblique MPR images perpendicular to the stomach wall, including the tumor, were reconstructed on a workstation, and the depth of tumor invasion was evaluated with a projection vertical to the tumor to avoid partial volume effects. Differences in assessment were resolved by consensus. We evaluated inter-observer agreement from differences in the assessment for T-staging of gastric cancer.
2.5. CT staging criteria of gastric cancer Mural invasion of gastric cancer in the gastric wall visualized by CT was classified as follows [11]:
T1: Strong enhancement with focal thickening is present in the inner and/or middle layer, but the outer layer shows no enhancement. Enhancement is present in the stomach wall, but the stomach wall is not thickened. The entire stomach wall does not abnormally enhance and there is no evidence of transmural tumoral involvement. T2–3: The entire stomach wall is enhanced abnormally accompanied by wall thickening, but the surface of the outer layer in contact with the fatty layer surrounding the stomach is smooth. T4a: The entire stomach wall is enhanced abnormally accompanied by wall thickening, and linear or reticular structures are observed in the fatty layer surrounding the stomach. T4b: The changes described above extend to adjacent contiguous organs.
2.7. Statistical analysis Fisher’s exact test was used to compare the accuracy of T-staging by air-MDCT with hydro-MDCT. A P-value of 0.05 was considered significant.
3. Results In hydro-MDCT, 45 of 57 patients could ingest 600 ml tap water, but 12 patients could not. These 12 patients drank about 500 ml tap water on average and the minimum amount of ingested water was
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
2957
Fig. 4. Images obtained in a 45-year-old woman with T3 gastric cancer overstaged as T4a. (a) and (b) Transverse CT and coronal images show that the entire stomach wall is enhanced abnormally accompanied by wall thickening (arrow), and that linear or reticular structures are observed in the fat layer of the surrounding stomach. (c) Histological micrograph shows a wide subserosal layer and tumor invading the subserosal layer with an irregular surface to the fat layer, but not reaching the serosal membrane (hematoxylin–eosin stain; low-power magnification).
about 400 ml. On the other hand, in air-MDCT, all patients could ingest 5 g gas-producing crystals. Correct assessment of the depth of mural invasion of the gastric wall (T stage) was 83% and 86% for hydro-MDCT and air-MDCT, respectively, which means that 49 of 57 patients with hydroMDCT and 48 of 58 patients with air-MDCT were correctly staged (Tables 2a and 2b). There were no significant differences between the two techniques in accuracy (P = 0.80). The sensitivity, specificity, and accuracy of the technique in determining the invasion of serosa were 83%, 95%, and 91% for hydro-MDCT and 88%, 93%, and 91% for air-MDCT (Table 3). There were no significant differences between them in sensitivity (P = 0.73), in specificity (P = 0.71) and in accuracy (P = 0.98).
3.1. T1 stage tumors Correct assessment of T1 tumor for hydro-MDCT and air-MDCT was 25/27 (92%) and 22/26 (85%), respectively. Four cases in airMDCT and two cases in hydro-MDCT were overstaged as T2–3. One T1 tumor in air-MDCT was a large protruding mass with contrast enhancement at MDCT. This tumor showed massive tumor formation of mucosa and minimal invasion of submucosa (SM1), but not invasion into the muscular layers on pathology. The other five (two in hydro-MDCT and three in air-MDCT) T1 tumors had entire enhancement accompanied by wall thickening on CT (Fig. 1). These five T1 tumors overstaged as T2–3 showed massive invasion into the submucosal layer (SM2) on pathology. However there were no significant differences in the staging of T1 gastric cancer between
the two methods (P = 0.42), the overstaged rate of the T1 tumor as T2–3 in air-MDCT tended to be higher than in hydro-MDCT. 3.2. T2–3 stage tumors Correct assessment of T2–3 tumors by hydro-MDCT and airMDCT was 9/12 (75%) and 12/16 (75%), respectively (Fig. 2), with no significant differences between the two methods (P = 1). One T3 tumor understaged as T1 on hydro-MDCT showed invasion into subserosa on histology. Even retrospectively on CT images, we thought that this tumor was very difficult to stage correctly as T3 because it had patchy enhancement of the inner and middle layers, but did not have entire wall thickening and enhancement. One T2 tumor understaged as T1 on air-MDCT existed at the pylorus, but could not be detected by two readers because they misinterpreted the pylorus tumor as normal gastric wall thickening. The two T3 tumors in hydro-MDCT and the three T3 tumors in air-MDCT were overstaged as T4a (Figs. 3 and 4). All tumors showed invasion into the subserosal layer. There were no T2 cases overstaged as T4a in either method. In these tumors, strandlike structures extending from the outer gastric surface into the surrounding fat plane were seen and considered suggestive of serosal involvement. At pathology, two T3 tumors had inflammatory change at the serosal membrane with an irregular outer surface, two T3 tumors had a relatively wide subserosal fat layer and tumor invaded this subserosal fat tissue with interface irregularity between the tumor and subserosal fat, but the tumor was surrounded by the serosal membrane. One T3
2958
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
Fig. 5. Images obtained in a 73-year-old man with stage T4a gastric tumor. (a) and (b) Transverse CT and coronal images show that the entire stomach wall is enhanced abnormally accompanied by wall thickening (arrow), and linear or reticular structures are observed in the fat layer of the surrounding stomach. (c) Histological micrograph shows extraserosal invasion (arrow) (hematoxylin–eosin stain; low-power magnification).
tumor had small lymph node metastasis adjacent to the serosal membrane.
3.3. T4a stage tumors Correct assessment of T4a tumor by hydro-MDCT and air-MDCT was 14/17 (82%) and 13/15 (87%), respectively (Fig. 5). There were no significant differences between the two methods (P = 0.74). Three T4a tumors in hydro-MDCT and two T4a tumors in airMDCT were understaged as T2–3 (Figs. 6 and 7). Of three T4a tumors understaged T2–3 in hydro-MDCT, 2 patients who had these tumors cannot ingest 600 ml water (average about 500 ml). By CT, the surface of the outer layer of the stomach wall, including the lesion, was smooth. These T4a tumors understaged as T2–3 often showed small enhanced lesions at CT. Pathological findings revealed that these tumors had minimal infiltration of cancer cells into the perigastric adipose tissue without macroscopic mass formation. In the five T4a tumors understaged as T2–3, all five tumors could be evaluated histologically.
3.4. T4b stage tumors One T4b tumor in hydro-MDCT and one T4b tumor in air-MDCT were both correctly staged because of the absence of normal fat cleavage planes with surrounding organs at CT. The T4b tumor in hydro-MDCT showed invasion into the left diaphragm by open surgery and the other in air-MDCT had confirmed invasion into the colon mesentrium and pancreas head by staging laparoscopy.
3.5. Interobserver agreement We used blinded consensus reading to resolve disagreements between the interpreting radiologists. In the evaluation of Tstaging, there was agreement in 48 of 58 patients (81%) with air-MDCT and in 46 of 57 patients (81%) with hydro-MDCT. 4. Discussion In preoperative CT examinations for T-staging of gastric cancer, there are many reports that hydro-CT and -MDCT are useful because of their excellent depiction of the gastric wall as three layers [6–13]. Air-MDCT allows 3D virtual gastrography, which improves the detectability of gastric cancer, especially early gastric cancer [17,18,25], but it is said that this technique is limited for evaluating T-staging of gastric cancer due to overshooting artifacts between the gastric wall and air in the stomach [14]. However, as the overshooting artifacts are expected to be reduced thanks to the advancement of multiplication of CT detector row, we compared air-MDCT and hydro-MDCT using a 64-channel system. In our present study, the diagnostic accuracy of T-staging of gastric cancer by air-MDCT was as good as by hydro-MDCT, which could explain why 64-channel MDCT can provide not only high spatial resolution of the z-axis but also high time resolution. The diagnostic accuracy of serosal invasion by air-MDCT was almost equal to that by hydro-MDCT and the sensitivity of serosal invasion by air-MDCT was superior to hydro-MDCT, although there were no significant differences between the two techniques. Diagnosis of serosal invasion is also very important, because serosal invasion of gastric cancer is a poor prognosis factor. Serosal invasion of gastric cancer underwent a trial treatment of
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
Fig. 6. Images obtained in a 62-year-old man with T4a gastric tumor understaged as T2–3. (a) Transverse and coronal CT images show gastric cancer on cardia. The tumor (arrow) was incorrectly staged as T2–3 because the surface of the outer layer appears smooth. (b) Histological microgram shows minimal infiltration of gastric cancer cells (arrow) into perigastric adipose tissue (hematoxylin–eosin stain; low-power magnification).
preoperative neoadjuvant chemotherapy for downstaging and curative operation. Staging laparoscopy has been widely used for the diagnosis of serosal invasion of gastric cancer [21–23], but this procedure is invasive and complicated. Our study showed that air-MDCT has the potential to indicate whether surgeons should perform staging laparoscopy. In our study, hydro-MDCT understaged T4a tumors as T2–3 more frequently than air-MDCT. T4a tumors, understaged as T2–3, showed that the outer surface of the stomach wall was smooth and regular, but there was minimal infiltration of tumor cells in the
2959
Fig. 7. Images obtained in a 2-year-old boy with T4a gastric tumor understaged as T2–3. (a) Transverse and coronal CT images show gastric cancer on cardia. Tumor (arrow) was incorrectly staged as T2–3 because the surface of the outer layer appears smooth. (b) Histological microgram shows minimal infiltration of gastric cancer cells (arrow) into perigastric adipose tissue (hematoxylin–eosin stain; low-power magnification).
perigastric adipose tissue at pathology. We speculated that insufficient distension of the stomach wall due to inappropriate intake of water made it difficult to evaluate invasion of the outer layer. In our study, approximately one fourth of patients could not drink 600 ml tap water. Good distension of the gastric wall is necessary to evaluate the T stage of gastric cancer, and there is a report that 800–1000 ml tap water was given to each patient [12], but realistically it is not easy for all patients to drink that much water. Air-MDCT is advantageous for T4a tumor staging due to good gastric wall distension. It is simpler and easier to perform air-MDCT than hydro-MDCT without the need to ingest water, and minimal
2960
S. Kumano et al. / European Journal of Radiology 81 (2012) 2953–2960
invasion into the extraserosa seems to be limited by using 64channel MDCT in either hydro-MDCT or air-MDCT. In our series, three and two T3 tumors were overstaged as T4a in air-MDCT and in hydro-MDCT, respectively. Shimizu et al. reported that deformity of the outer surface of the gastric wall can occur as a result of factors other than serosal invasion, such as compression of the serosa by a tumor [10]. In T3 tumors overstaged as T4a gastric cancer, histologically, serosal inflammation without cancer cells in serosa was seen in 1 case, invasion into the wide subserosal layer with an irregular interface was seen in 2 cases, and small lymph node metastasis adjacent to extraserosa was seen in 1 case. These pathologic features of the T3 tumor overstaged as T4a were thought to be the most likely cause of deformity of the outer surface of the gastric wall. All T2–3 tumors overstaged as T4a were T3 tumor (subserosa (SS) invasion and no mascularis propria (MP) invasion). Either by air-MDCT or hydroMDCT, MDCT could correctly stage T2 tumors with invasion, but it might be limited to differentiate T3 tumors with SS and T4a tumors. Although there were no significant differences between the two techniques, air-MDCT tended to overstage T1 tumor as T2–3 compared to hydro-MDCT CT in our study. Hydro-MDCT can depict the three layers of the gastric wall more clearly than air-MDCT, because a gastric wall well-distended by air made the middle layer, consisting of the submucosal layer, unclear in air-MDCT. Mass-forming T1 gastric cancer with invasion into the submucosal layer tends to be overstaged as T2–3 in either hydro-MDCT or air-MDCT. One T2 tumors were understaged as T1 in our study, the case existed at the pylorus, which could not be distinguished from normal gastric wall thickening. There were problems evaluating the tumor at the antrum, which had a thickened wall due to insufficient distension by air-CT. Patient position might need to be changed on the CT table to the right side for good distension of the antrum. Our study has several limitations. First, we could not evaluate the two methods (hydro-MDCT and air-MDCT) in the same patients. Our investigation was a clinical study, so it was impossible to compare the two methods in the same patients; however, there were no significant differences in the background of the two patient groups. Second, we did not obtain a resected specimen in all T4a cases. However some of diagnosis was made by staging laparoscopy, all of the laparoscopic examinations staged for T4a, including the cytological evaluation of peritoneal lavage fluid and peritoneal metastasis, gave pathological confirmation. Moreover, there were no significant differences in the two patient groups in the number of T4a tumors diagnosed by staging laparoscopy, and we considered that our study could minimize the disadvantages. Third, we did not use 3D gastrography in air-MDCT. There have been reports that 3D gastrography can detect early gastric cancer more than 2D images [25]; however, our aim was to investigate the ability of MDCT to evaluate the depth of mural invasion, but not to show the detectability of gastric cancer on MDCT. This cannot be construed as a major limitation of our study. Forth, we did not perform the dynamic study and did not obtain the arterial phase, therefore we adopted the injection rate for 2 ml/s. The arterial phase is important to visualize the mucosal distortion, in the present study, it might underestimate the evaluation of early gastric cancer. In conclusion, air-MDCT is a valid examination for preoperative staging of patients with gastric cancer, and gave especially high accuracy in the assessment of serosal invasion of gastric cancer.
Conflict of interest None. References [1] Devesa SS, Blot WJ, Fraumeni Jr JF. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer 1998;83(10):2049–53. [2] Mishima Y, Hirayama R. The role of lymph node surgery in gastric cancer. World J Surg 1987;11(4):406–11. [3] Boku T, Nakane Y, Minoura T, Takada H, Yamamura M, Yamamoto M. Prognostic significance of serosal invasion and free intraperitoneal cancer cells in gastric cancer. Br J Surg 1990;77(4):436–9. [4] Abe S, Shiraishi M, Nagaoka S, Yoshimura H, Dhar DK, Nakamura T. Serosal invasion as the single prognostic indicator in stage IIIA (T3N1M0) gastric cancer. Surgery 1991;109(5):582–8. [5] Kurtz RC, Sherlock P. Carcinoma of the stomach. In: Berk JD, editor. Gastroenterology. Philadelphia: Saunders; 1985. p. 1278–304. [6] 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(1):69–75. [7] 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(1):173–8. [8] Habermann CR, Weiss F, Riecken R, et al. Preoperative staging of gastric adenocarcinoma: comparison of helical CT and endoscopic US. Radiology 2004;230(2):465–71. [9] Rossi M, Broglia L, Graziano P, et al. Local invasion of gastric cancer: CT findings and pathologic correlation using 5-mm incremental scanning, hypotonia, and water filling. AJR Am J Roentgenol 1999;172(2): 383–8. [10] Shimizu K, Ito K, Matsunaga N, Shimizu A, Kawakami Y. Diagnosis of gastric cancer with MDCT using the water-filling method and multiplanar reconstruction: CT–histologic correlation. AJR Am J Roentgenol 2005;185(5): 1152–8. [11] Kumano S, Murakami T, Kim T, et al. T staging of gastric cancer: role of multidetector row CT. Radiology 2005;237(3):961–6. [12] Chen CY, Hsu JS, Wu DC, et al. Gastric cancer: preoperative local staging with 3D multi-detector row CT—correlation with surgical and histopathologic results. Radiology 2007;242(2):472–82. [13] Kadowaki K, Murakami T, Yoshioka H, et al. Helical CT imaging of gastric cancer: normal wall appearance and the potential for staging. Radiat Med 2000;18(1):47–54. [14] 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(2):149–51. [15] Ba-Ssalamah A, Prokop M, Uffmann M, Pokieser P, Teleky B, Lechner G. Dedicated multidetector CT of the stomach: spectrum of diseases. Radiographics 2003;23(May–June (3)):625–44. [16] Park KJ, Lee MW, Koo JH. Detection of early gastric cancer using hydro-stomach CT: blinded vs unblinded analysis. World J Gastroenterol 2011;17(February (8)):1051–7. [17] Shin KS, Kim SH, Han JK, et al. Three-dimensional MDCT gastrography compared with axial CT for the detection of early gastric cancer. J Comput Assist Tomogr 2007;31(5):741–9. [18] Yang DM, Kim HC, Jin W, et al. 64 multidetector-row computed tomography for preoperative evaluation of gastric cancer: histological correlation. J Comput Assist Tomogr 2007;31(1):98–103. [19] 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(September (3)):879–85. [20] Sobin LH, Wittekind C, editors. TNM classification of malignant tumors. International Union Against Cancer (UICC). 7th ed. New York: Wiley; 2009. [21] D’Ugo DM, Persiani R, Caracciolo F, Ronconi P, Coco C, Picciocchi A. Selection of locally advanced gastric carcinoma by preoperative staging laparoscopy. Surg Endosc 1997;11(12):1159–62. [22] Blackshaw GRJC, Barry JD, Edwards P, Allison MC, Thomas GV, Lewis WG. Laparoscopy significantly improves the perceived preoperative stage of gastric cancer. Gastric Cancer 2003;6(4):225–9. [23] Chag L, Stefanidis D, Richardson WS, Earle DB, Fanelli RD. The role of staging laparoscopy for intraabdominal cancers: an evidence-based review. Surg Endosc 2009;23(2):231–41. [24] Japanese Gastric Cancer Association. Japanese classification of gastric carcinoma. 2nd English edition. Gastric Cancer 1998;1:10–24. [25] Shen Y, Kang HK, Jeong YY, et al. Evaluation of early gastric cancer at multi-detector CT with multiplanar reformation and virtual endoscopy. Radiographics 2011;31:189–99.