Laboratory Studies
No laboratory test can specifically confirm or rule out the presence of a gastrointestinal stromal tumor (GIST). The following tests are generally ordered in the workup of patients who present with nonspecific abdominal symptoms; abdominal pain; or findings that may be due to complications of GISTs, such as hemorrhage, obstruction, or perforation:
-
Complete blood cell count
-
Coagulation profile
-
Serum chemistry studies
-
BUN and creatinine
-
Liver function tests
-
Amylase and lipase values
-
Blood type, screen, and crossmatch
-
Serum albumin
GISTs are not associated with elevation of any serum tumor markers. However, depending on the location, size, and appearance of the tumor on imaging studies, performing tumor marker assays such as the following for other abdominal neoplasms may be appropriate:
-
Cancer antigen 19-9 (CA19-9)
-
Carcinoembryonic antigen (CEA)
-
CA-125
-
Alpha fetoprotein (AFP)
Imaging Studies
Imaging studies used in the workup of GISTs include the following:
-
Radiography
-
Ultrasonography
-
Computed tomography (CT)
-
Magnetic resonance imaging (MRI)
-
Positron emission tomography (PET)
Radiography
Plain abdominal radiography is nonspecific but may be ordered as part of the workup of a patient presenting emergently with a possible bowel obstruction or perforation. Abnormal gas patterns, including dilated loops of bowel or free extraluminal air, are examples of findings that may be detected in these clinical situations. [42]
Barium and air (double-contrast) series
Double-contrast radiographic series can usually detect GISTs that have grown to a size sufficient to produce symptoms. The choice of performing a barium swallow, barium enema, or both depends on the patient's clinical presentation. For example, patients whose primary symptomatology includes dysphagia should have a barium swallow. Those presenting with constipation, decreased stool caliber, or other signs and symptoms referable to the colon should have a barium enema.
In these contrast studies, GISTs appear as a filling defect that is sharply demarcated and is elevated compared with the surrounding mucosa. [43] Typically, the contour of the overlying mucosa is smooth unless ulceration has developed because of growth of the underlying tumor.
While these studies can produce striking images, frequently the information they provide is limited. Other modalities listed below have equal or greater sensitivity and can provide more information about the status of the surrounding structures.
Enteroclysis
As is frequently the case with other small bowel lesions, GISTs in the small intestine can be difficult to diagnose and localize. Enteroclysis allows delivery of contrast into the small bowel so it does not become too diluted by the time it reaches the area in question. This may help to better define small intestinal GISTs.
Ultrasonography
The ultrasonographic appearance of GISTs varies depending on the size of the lesion and the presence or absence of necrosis within the mass. In addition, ulceration or necrosis of the overlying mucosa can change the ultrasonographic characteristics of the tumor.
Transcutaneous ultrasonography is probably not the optimum choice for imaging these lesions unless the mass has reached quite a large size. Because GISTs are associated with air-filled viscera, image quality is often degraded by intervening bowel gas.
The best ultrasonographic images of these lesions are acquired during endoscopic ultrasonography (see Procedures).
Computed tomography
CT scanning with intravenous and oral contrast material is a necessary step in the diagnosis and staging of GISTs. [25] It provides comprehensive information regarding the size and location of the tumor and its relationship to the adjacent structures. CT scanning can also be used to detect the presence of multiple tumors and can provide evidence of metastatic spread.
Ghanem and colleagues described the CT characteristics of GISTs on the basis of size, as follows [2] :
-
Small (< 5 cm): Sharply demarcated, homogeneous masses, mainly exhibiting intraluminal growth patterns
-
Intermediate (5-10 cm): Irregular shape, heterogeneous density, an intraluminal and extraluminal growth pattern, and signs of biological aggression, sometimes including adjacent organ infiltration
-
Large (>10 cm): Irregular margins, heterogeneous densities, locally aggressive behavior, and distant and peritoneal metastases
In patients with gastric GISTS, the gastric mass can be detected originating from the gastric wall, as depicted in the image below. In some cases, however, the organ site of origin is not clear on CT scan. CT scanning can also be used to evaluate tumor invasion to adjacent structures and the presence of intra-abdominal metastasis. The identification of distant disease is important, as many as half of patients who initially present with a GIST have distant metastases (two-thirds of those have hepatic involvement).

Magnetic resonance imaging
Like CT scanning, MRI can depict the tumor or tumors and provide information about surrounding structures. It can also be used to detect the presence of multiple tumors and metastases.
MRI can be an especially helpful adjunct to CT in the evaluation of large tumors that have necrotic and hemorrhagic components. Solid tumor portions show low intensity on T1 images and high intensity on T2 images, with enhancement of the mass when intravenous gadolinium is given. Signal intensity of hemorrhagic components of the tumor can vary from high to low, depending on the age of the hemorrhage.
Positron emission tomography
PET scanning with 2-[F-18]-fluoro-2-deoxy-D-glucose (18-FDG PET) has been recommended for detecting metastatic disease in patients with GISTs. However, it is principally useful for early detection of the tumor response to adjuvant therapies such as imatinib mesylate. [44]
Procedures
Endoscopy
Endoscopy is frequently performed early in the workup of patients with GISTs, to evaluate GI bleeding, abdominal pain, or GI obstructive symptoms. However, endoscopy is not generally required in the workup of patients with suggestive lesions on CT.
Endoscopic features of GISTs include the suggestion of a firm, smooth, yellowish submucosal mass displacing the overlying mucosa. Some tumors may be associated with ulceration or bleeding of the overlying mucosa from pressure necrosis. See the image below.

These tumors can be missed on endoscopy because of their frequent submucosal and extraluminal growth. If the diagnosis is suspected prior to endoscopy, an endoscopic ultrasound scan can be performed to further characterize and help confirm the origin of the lesion (when the organ site of the tumor is not clearly evident on CT scan), even if not visible endoscopically.
Endoscopic ultrasonography
Endoscopic ultrasonography (EUS) can be a valuable tool in the diagnosis and preoperative assessment of gastric GISTs when the diagnosis or location is in doubt, and EUS is generally the preferred modality to facilitate biopsy of the lesion, when biopsy is indicated. However, EUS is not generally required for preoperative workup. [4, 45]
If the location of the lesion is in doubt, the EUS can help plan the operative approach (eg, demonstrating that a proximal gastric lesion on CT scan is far enough away from the gastroesophageal junction to allow local wedge resection as opposed to total gastrectomy). EUS can also demonstrate the submucosal location of the tumor and can define its size, borders, and echoic pattern.
In general, ultrasonic features of a mass suspicious for malignancy are as follows:
-
Large size
-
Irregular extraluminal borders
-
Presence of cystic spaces and echogenic foci
Preoperative biopsy
While the diagnosis can often be made using ultrasonography-guided biopsy, the use of biopsy is controversial in an otherwise primary, resectable lesion suspicious for GIST. Generally, biopsy is not recommended in this setting. [39] Biopsy of a GIST, which tends to be soft and fragile, may cause intratumoral hemorrhage or even rupture and may increase the risk for tumor dissemination. Irrespective of the biopsy results, surgical resection is typically required for treatment and for definitive diagnosis.
However, biopsy is required in the setting of suspected metastatic disease or when neoadjuvant treatment of a borderline resectable GIST is being entertained. Biopsy may also be important when the diagnosis of GIST in is question, such as when the submucosal nature of the tumor is in doubt or when the tumor shows atypical characteristics on endoscopy or endoscopic ultrasonography. In specific patients, such as those at high operative risk with small benign-appearing lesions and minimal or no symptoms, tissue diagnosis may guide decision making.
Endoscopic biopsy is the recommended procedure for obtaining a preoperative histologic diagnosis. Endoscopic biopsy with ultrasound guidance can achieve a correct histologic diagnosis in more than 80% of cases. Without ultrasound guidance, endoscopic biopsy leads to a correct diagnosis in less than 50% of patients; unguided biopsies may miss the tumor and show only mucosal tissue, or provide samples that contain too little tumor tissue to establish whether the tumor is malignant.
Biopsy specimens can also be obtained percutaneously under CT or ultrasonographic guidance. [46] Consider this approach in selected patients when endoscopic biopsy is impossible to perform or the results are negative.
Histologic Findings
Cellular morphology as visualized by light microscopy can be variable. Most often, the tumors are highly cellular and composed of spindle-shaped cells that resemble smooth-muscle tissue, as depicted in the first image below. However, this histologic appearance is not uniform. A similar tumor with a predominant epithelioid component was historically diagnosed as leiomyoblastoma, as depicted in the second image below. This variant is occasionally associated with a well-defined condition called Carney syndrome.


Important histologic factors to consider in evaluating these tumors are as follows:
-
Mitotic index
-
Cellularity
-
Necrosis
-
Nuclear atypia and nuclear-cytoplasmic ratio
-
Cell shape
-
Amount of stroma
-
Vascularity
Investigations of GISTs by immunohistochemistry and electron microscopy (ultrastructural parameters) reveal phenotype variability that includes myoid, neural, and indeterminate characteristics. [47] Study of GISTs by immunohistochemistry methods reveals expression of CD117 and other various antigens, such as the following:
-
Nestin (90-100% positivity)
-
CD34 (70% positivity)
-
CD44 [48]
-
Vimentin
-
Desmin
-
Muscle-specific actin
-
Smooth-muscle actin
-
S-100 protein
-
Neurofilament
-
Neuron-specific enolase
-
Protein gene product 9.5 (PGP9.5)
CD117 plays an important role in the latest specific diagnostic criteria for GISTs. CD117 (c-kit protein) is a growth factor receptor with tyrosine-kinase activity and is a product of the proto-oncogene c-kit. CD117, although not tumor-specific, is expressed in all GISTs but not in true smooth muscle tumors and neural tumors.
CD117 has become a very important tool in the differentiation of GIST from other GI mesenchymal tumors. [49, 50] Positive CD117 staining in a spindle-shaped cell GI tumor is diagnostic for GIST, as depicted in the image below. CD34 is another important diagnostic marker. It is detected in approximately 70% of GISTs, and its presence may indicate a higher probability for a malignant phenotype. CD44 is variably expressed by GISTs, but its expression has been demonstrated to correlate with a better prognosis.
Staging
No consensus has been reached regarding a uniform staging system, and none of the currently used classifications is fully satisfactory. Most staging systems employ the American Joint Committee on Cancer (AJCC) tumor-node-metastasis (TNM) classification, which utilizes the three most important survival predictors: tumor size, histologic grade, and presence or absence of distant metastatic disease. [51]
The NCCN criteria for risk stratification of primary GIST have not been incorporated into the AJCC staging but may be more helpful in determining individual risk for progressive disease, after margin-negative resection. The stratification is by mitotic index (≤5 versus > 5 per 50 high-power field [HPF]) and then further divided by tumor size (≤2 cm vs > 2 cm; ≤5 cm vs > 5 cm; ≤10 cm vs > 10 cm) and tumor location (gastric and non-gastric). [39]
Gastric GISTs greater than 10 cm but with a mitotic index of 5/50 HPF or less have only a 12% risk of progressive disease, compared with 34-52% risk of progressive disease in the other tumor locations. Gastric GISTs greater than 10 cm with a high mitotic index (> 5/50 HPF), however, have an equally high risk of progressive disease (86%) as GISTs in other locations (71-90%). [39]
Many studies have shown that tumor diameter greater than 5 cm is associated with increased risk for malignancy. However, the relation of size to malignant potential may be gradual, with no clear cut-off point.
A high mitotic index (> 5/50 HPF) usually signifies highly malignant disease. However, a low mitotic index is not always associated with a benign course. As many as 25% of tumors with a mitotic index of < 5/50 HPF may manifest an aggressive biological behavior. Other histologic parameters, such as cellularity, atypia, and necrosis, are also taken into consideration.
The TNM classification for GISTs is provided in Table 1, below.
Table 1. TNM Classification for Gastrointestinal Stromal Tumors (Open Table in a new window)
Primary tumor (T) |
|
TX |
Primary tumor cannot be assessed |
T0 |
No evidence of primary tumor |
T1 |
Tumor 2 cm or smaller |
T2 |
Tumor more than 2 cm but not more than 5 cm |
T3 |
Tumor more than 5 but not more than 10 cm |
T4 |
Tumor more than 10 cm in greatest dimension |
Regional lymph node (N) |
|
N0 |
No regional lymph node metastasis |
N1 |
Regional lymph node metastasis |
Metastasis (M) |
|
M0 |
No distant metastasis |
M1 |
Distant metastases |
The anatomic stage/prognostic groups for gastric and small-intestinal GISTs are shown in tables 2 and 3, below.
Table 2. Anatomic Stage/Prognostic Groups for Gastric GISTs (Open Table in a new window)
Group |
Tumor |
Node |
Metastasis |
Mitotic Rate* |
Stage IA |
T1 or T2 |
N0 |
M0 |
Low |
Stage IB |
T3 |
N0 |
M0 |
Low |
Stage II |
T1 |
N0 |
M0 |
High |
T2 |
N0 |
M0 |
High |
|
T4 |
N0 |
M0 |
Low |
|
Stage IIIA |
T3 |
N0 | M0 | High |
Stage IIIB | T4 | N0 | M0 | High |
Stage IV |
Any T |
N1 |
M0 |
Any rate |
Any T | Any N | M1 | Any rate | |
*Mitotic rate: Low (≤5 per 5 mm2 or per 50 HPF); High (>5 per 5 mm2 or per 50 HPF) |
Table 3. Anatomic Stage/Prognostic Groups for Small-Intestinal GISTs (Open Table in a new window)
Group |
Tumor |
Node |
Metastasis |
Mitotic Rate* |
Stage I |
T1 or T2 |
N0 |
M0 |
Low |
Stage II |
T3 |
N0 |
M0 |
Low |
Stage IIIA |
T1 |
N0 | M0 | High |
T4 | N0 | M0 | Low | |
Stage IIIB | T2 | N0 | M0 | High |
T3 |
N0 |
M0 |
High |
|
T4 |
N0 |
M0 |
High |
|
Stage IV |
Any T |
N1 |
M0 |
Any Rate |
Any T | Any N | M1 | Any Rate | |
*Mitotic rate: Low (≤5 per 5 mm2 or per 50 HPF); High (>5 per 5 mm2 or per 50 HPF) |
-
Shown here is a gastric gastrointestinal stromal tumor (GIST). This is a gross specimen following partial gastrectomy. Note the submucosal tumor mass with the classic features of central umbilication and ulceration.
-
CT scan of the abdomen with oral contrast in a 60-year-old woman with a gastric gastrointestinal stromal tumor (GIST). A huge mass with central necrosis is observed originating from the gastric wall and narrowing its lumen. An ulcer crater can be identified within the mass (arrow).
-
Photomicrograph of gastrointestinal stromal tumor (GIST) stained with hematoxylin and eosin (H&E) and magnified 40X. Note the solid sheet of spindle cells.
-
Photomicrograph of gastric gastrointestinal stromal tumor (GIST) stained with hematoxylin and eosin (H&E) and magnified 400X. This stromal tumor demonstrates spindle cells with epithelioid features.
-
Photomicrograph of gastrointestinal stromal tumor (GIST) with immunohistochemical staining for CD117. Note the strong positive staining of tumor cells with negative staining of the adjacent vessel. Positive stain for CD117 is diagnostic of GIST.