Surgical Oncology net
Blog aimed at residents in surgery

e mail


Blog aimed at residents in surgery
 URL :             e mail
Staging in gastric cancer: imaging diagnosis.

Romeo Giuli MD, resident.
School of General and Emergency Surgery.
University of Siena.   Italy.

November 2001.     Review Article.

Endoscopic Ultrasonography and Hydrosonography
Conventional Radiology
Computerized Tomography
Magnetic Resonance Imaging
Positron Emission Tomography
Angiographic Techniques


At the beginning of any cancer therapy, the tumor stage must be evaluated. The choice of therapy depends on the patient's prognosis, which generally is determined by the specific tumor stage. Therefore the preoperative imaging diagnosis is the basis for a tumor-stage-adapted therapy of each patient.

EUS ( Endoluminal Ultrasound ) is performed with 7 to 12 MHz scanners. Five layers of the gastric wall can be visualized endosonographically. The inner layer is hyperechoic. Thin hyperechoic and hypoecoic layers are alternating. The second hypoecoic layer represents the lamina muscolaris propria, which exists throughout the whole gastrointestinal tract. If the carcinoma does not penetrate this second hypoecoic layer, it has to be classified as T1. If the third hyperechoic layer is infiltrated, the carcinoma is classified as uT3. The serosa, which is the critical pathohistological layer, cannot accurately be visualized by ultrasound because of its thinness.
A major problem affecting endosonograpy (EUS) and hydrosonography is caused by the definition of T2 (tumor infiltrating into the lamina muscolaris propria of the stomach) and T3 (tumor penetrating into or through the serosa) carcinomas in the TNM system. When the carcinoma infiltrates the subserosal fat tissue it still has to be classified as pT2 carcinoma. But sonographically the lamina muscolaris propria appears hypoechoic and the subserosal fat tissue and the serosa itself appear hyperechoic. Therefore a carcinoma which infiltrates the subserosal fat tissue is sonographically visualized as uT3.
Furthermore the peritumorous desmoplatic reaction may equally appear in sonography, simulating a uT3 carcinoma as well when, indeed, it is still a pT2 carcinoma.
Other shortcomings of EUS include understaging due to microscopic nodal metastases and subtle tumor infiltration of deeper layers, which can go undetected.

Lymph nodes are detectable when their diameter exceeds 3 mm. According to Kuntz's study tumor-infiltrated lymph nodes appear inhomogenueus and hypoechoic,similar to the primary gastric carcinoma, whereas inflammatory enlarged lymph nodes mostly appear homogeneous and hyperechoic ( 1 ).
In comparison with resected specimen EUS has proven to be very accurate in estimating the depth of tumor invasion and lymph node involvement. The overall accuracy for T staging is in the vicinity of 85%. Lymph node involvement is detected with a sensitivity ranging from 55 to 80% and a specificity approaching 90% ( 2 ).

Recently miniprobe ultrasonography has shown an overall accuracy of 80% for T staging, but with more dependance on T stage. Therefore miniprobe US is a necessary device for those patients with early gastric cancer to whom mucosectomy should be offered, since differentiation of T1m and T1sm is possible (and necessary), and lymph node metastases occur in 3.3% (T1m) and 19.6% (T1sm) respectively. EUS and miniprobe ultrasonography correctly classifies lesions that are limited to the mucosa in 92% of cases. The combination of both techniques is therefore the method of choise for identifying mucosal carcinoma ( 3 ).

EUS is useful in detecting destruction of the gastric wall due to lymphoma, as well as linitis plastica and other disorders. EUS is the method of choice for staging infiltrative gastric wall disorders. Differential diagnosis of gastric fold thickening (Menetrier's disease, linitis plastica and lymphoma) is sometimes difficult, or even impossible, if no histologic abnormalities are found. In those cases, large biopsy forceps may increase diagnostic yeld, or EUS-guided FNA may be considered ( EUS is helpful in guiding needless precisely through the gut wall into surrounding structures to obtain bioptic samples ). MALT lymphoma can be assessed by EUS and EUS can be useful in assessing the response to Helicobacter eradication. Also the ultrasonic miniprobe can be recommended as part of routine care in patients with gastric MALT lymphoma, both initially and during the follow up period ( 3, 4 ).

Submucosal lesions of the gastrointestinal tract are best diagnosed by EUS. EUS can reliably distinguish between solid intramural lesions and extramural compressions. Furthermore EUS can suggest the nature of the tumor by determining the origin of the tumor and the corresponding layer ( e.g. a hypoecoic lesion in the fourth layer is pathognomonic for a stromal cell tumor ).Nevertheless different experts had low accuracy ( ranging from 56% to 77% ) in predicting malignancy ( 3 ).

With EUS ascites in the bursa omentalis can be detected as an indirect sign of peritoneal metastases. In contrast to hydrosonography, liver metastases or peritoneal metastases distant from the stomach cannot adequately be detected by EUS.

Further repeated tumor evaluations during preoperative chemotherapy are easily performed with hydrosonography when the tumor has been correctly visualized once before. On the other hand small gastric tumors (pT1, pT2 ) can be better differentiated with the high resolution capacity of EUS ( 12 MHz ) than with hydrosonography ( 1 ).
Hydrosonography of the stomach may be performed simply, with oral ingestion of 500 ml of an aqueous solution. To avoid air bubble formation and rapid absorption, an isotonic polyetilenglycole solution may be preferable. A lesion detection of 75% with an overall T-staging accuracy rate of 50% can be achieved, with relatively lower values for the cardias and fundus regions, which are located in a deeper position with respect of the US beam. The N staging accuracy is 61% as compared with 86% of endosonography ( 5 ).

Conventional radiology is particularly useful in scirrhous carcinoma characterized by diffuse, predominantly submucosal infiltration. It is manifested by irregular narrowing of the lumen and rigidity due to marked desmoplastic reaction, the so called linitis plastica. In advanced neplasms of this type, extension of the tumor to the transverse colon through the gastrocolic ligament may be manifested on barium enema studies by irregular tethering of the mucosal surface contour or luminal narrowing. When peritoneal dissemination occurs, an abdominal plain film can disclose the presence of associated ascites or separation, or narrowing and angulation of the mesenteric gas-filled bowel loops. Distant skeletal and pulmonary metastases from gastric cancer can be seen on skeletal survey and chest radiographs, respectively ( 6 ).

In the preoperative staging of gastric carcinoma the criteria to be assessed with CT include extension of the tumor along the wall and adjacent areas, lymph node metastases, and distant metastases ( 7 ).
There are different CT techniques used: incremental or helical CT, negative or positive intraluminal contrast medium, or intravenous contrast medium ( 1 ).
For better evaluation, filling the stomach with water as negative oral contrast material can reduce artifacts. Peristaltic interference may be further avoided with drug-induced hypotonia of the stomach. Helical CT has been another advance. By shortening scanning time, motion artifacts have been reduced and, using rapid intravenous contrast administration has enhanced the detail of the gastric wall and primary tumor ( 2 ).
The results are still disappointing for recognition of neoplasms confined to mucosa and submucosa ( early gastric cancer ), as the diagnostic accuracy values range from only 23% to 56%. However, the accuracy values increases for evaluation of tumors in more advanced stages: 88% to 95% for T4. Recognition of infiltration of the serosa was found to have accuracy values of 70% to 80% in the same case series ( 7 ).
Regarding lymph node detection, Fukuya et al. note that helical CT identifie only 1% of nodes <5 mm, 45% of nodes between 5 and 9 mm and over 70% of nodes larger than 9 mm in size. Lymph node size correlates with metastases, and over 80% of nodes greater than 14 mm containe disease. However, not all historically proven lymph node metastases were found in enlarged glands. Tumor was noted in 5% of lymph nodes less than 5 mm, 21% of nodes 59 mm, and 23% of nodes 1014 mm. Overall, the authors note the difficulty in diagnosing metastases in lymph nodes less than 14 mm ( 8 ).
By far the most common staging tool to diagnose distant disease is the abdominal and pelvic CT scan. Carcinomatosis with diffuse peritoneal seeding, malignant ascites, and adnexal and pelvic metastases is well depicted by CT. In a prospective trial, CT had a 79% accuracy at diagnosing hepatic metastases and a 81% accuracy at diagnosing peritoneal metastases. All metastases were histologically confirmed by laparoscopy or laparotomy. In a study in which all patients underwent laparotomy, Adachi et al. noted that CT accurately staged the liver for metastases in 96% of cases but missed 27% of the cases with peritoneal metastases. Obviously, the accuracy of CT is dependent on the bulk of metastatic tumor. CT will miss the majority of hepatic metastases <1 cm and small volume peritoneal disease ( 2, 9, 10 ).

The sensitivity of US in identifying metastatic liver nodules is inversely proportional to the size of the tumor and can be optimized, as already discussed, with the introduction of contrast media in conjunction with the utilization of harmonic tissue imaging. Moreover under US guidance biopsies can be performed. Finally, intraoperative US may be considered the gold standard in liver nodule detection and may alter the preoperative plan in one third of the patients. ( 5 ).

In some cases MRI has demonstrated the ability to depict a primary tumor and define the degree of its invasion depth; it is a valid means of evaluating intraperitoneal extension. The administration of intravenous contrast medium increases, in most cases, the conspicuousness of gastric lesions, and with T1/ fat suppressions sequences facilitates the identification of an eventual intraperitoneal diffusion of the lesion. The regional lymph nodes and the peritoneal seedings enhance intensely if the fat signals is suppressed.
It is important to note that at this time MRI is equal to CT in evaluating lymph nodes ( 11 ). MRI is moreover helpful characterization of liver lesions ( 2 ).

The potential distribution of the anatomical sites of metastases, via the large venous and lymphatic drainage, can make FDG-PET very useful in the evaluation of nodal involvement in the upper to middle body. No large clinical studies have applied PET in the staging of gastric cancer. The diagnostic performance is encouraging in detecting lymph node invasion and is superior to the conventional imaging techniques. Changing in the 18 F-FDG uptake have been observed in tumors after chemotherapeutic treatments, and the rate of reduction of FDG uptake is correlated with prolonged survival. Further clinical studies are needed to investigate both the usefulness of the FDG-PET in gastric cancer staging and in predicting a favourable response to chemotherapy in advanced disease ( 12, 13, 15 ).

Selective celiac and mesenteric angiography is not routinely performed before gastroduodenal resections. However, preoperative knowledge of the detailed vascular anatomy may be useful in patients who have undergone previous surgical gastrointestinal resections to identify major feeding vessels to the organs. Moreover gastric bleeding that is refractory to systemic and endoscopic control may be the first sign of a gastric cancer, or may occur during the follow up of an advanced or recurrent tumor. In these cases, angiographic superselective embolization may successfully stop the bleeding ( 14 ).


1) C Kuntz, C Herfarth. Imaging diagnosis for staging of gastric cancer. Seminars in Surgical Oncology 1999; 17: 96-102.

2) J Tschmelitsch, M R Weise, M S Karpeh. Modern staging in gastric cancer. Surgical Oncology, 2000; 9, 23-30.

3) Helmut Messmann, Klaus Schlottmann. Role of endoscopy in the staging of esophageal and gastric cancer. Seminars in Surgical Oncology 2001; 20: 78-81.

4) N Lugering et al. Impact of miniprobes compared to conventional endosonography in the staging of low-grade gastric MALT lymphoma. Endoscopy 2001; 33 (10): 832-837.

5) CE Neumaier et al. Role of ultrasonography in the staging of gastrointestinal neoplasms. Seminars in Surgical Oncology 2001; 20: 86-90.

6) N Gourtsoyannis et al. Role of conventional radiology in the diagnosis and staging of gastrointestinal tract neoplasms. Seminars in Surgical Oncology 2001; 20: 91-108.

7) G Angelelli et al. Role of computerized tomography in the staging of gastrointestinal neoplasms. Seminars in Surgical Oncology 2001; 20: 109-121.

8) FukayaT et al. Lymph node metastases: efficacy of detection with helical CT in patients with gastric cancer. Radiology 1995; 197: 705-11.

9) Stell DA et al. Prospective comparison of laparoscopy, ultrasonography and computed tomography in the staging of gastric cancer. British Journal of Surgery 1996; 83 ( 9 ): 1260-62.

10) Y. Adachi, I. Sakino, T. Matsumata, Y. Iso, R. Yoh, S. Kitano et al., Preoperative assessment of advanced gastric carcinoma using computed tomography. American Journal of Gastroenterology 92 5 (1997), pp. 872875.

11) R Ferraris et al. Role of Magnetic Resonance Imaging in the staging of gastrointestinal neoplasms. Seminars in Surgical Oncology 2001; 20: 122-129.

12) Histopathologic validation of lymph node staging with FDG-PET scan in cancer of the esophagus and gastroesophageal junction. Annals of Surgery,2000, vol 232, n 6, 743-752.

13) E Bombardieri et al. Positron emission tomography and other nuclear medicine modalities in staging gastrointestinal cancer. Seminars in Surgical Oncology 2001; 20: 134-146.

14) G De Caro, G Cittadini. Role of angiographic techniques in the preoperative staging and management of gastrointestinal neoplasms. Seminars in Surgical Oncology 2001; 20: 130-133.

15) HJ Stein, A Sendler, U Fink and JR Siewert.Multidisciplinary approach to esophageal and gastric cancer. Surgical Clinics of North America, 2000, vol 80, n 2, 659-681.

TO HOME PAGE next article

Surgical Oncology net