Advertisement

Extragastrointestinal Stromal Tumors (EGISTs)

  • Quan Jiang
  • Weiqi Lu
Chapter

Abstract

Until 2002, stromal tumor has been re-recognized and correctly named, which was confused with leiomyomas for such a long time. A breakthrough has been made on stromal tumor over this short span of a decade. Especially the use of imatinib, a tyrosine kinase inhibitor (TKI), allows stromal tumor to become a model for targeted therapy. Gastrointestinal stromal tumor (GIST) is the most common gastrointestinal mesenchymal tumor, accounting for 2% of all gastrointestinal tumors. GIST often occurs in the stomach (40–70%), small intestine (20–40%), colon (5–15%), and rectum (5%). On the contrary, primary extragastrointestinal stromal tumor (EGIST) accounts only a very small percentage of all stromal tumors. EGIST occurs frequently in the mesentery, omentum, and peritoneum but rarely in the retroperitoneum. Up to now, only 59 cases have been reported. EGIST is commonly seen in young and middle-aged people, with an incidence comparable between men and women. EGIST often grows very large before it is detected incidentally, resulting in a low rate of complete resection and a high rate of recurrence (even after complete surgical resection). Although located in different primary sites, EGIST is substantially similar to GIST in terms of pathology and immune phenotype. Under microscope, tumor cells are short spindle, nest, sheet, or vesicle shaped, with translucent cytoplasm. Immunohistochemistry reveals CD117 positive, as well as actin, calponin, and S100 positive or weakly positive. C-kit gene mutation is detectable. Under electron microscope, tumor cells appear as irregular fusiform with intertwined pseudopodia of varying lengths extending toward the surrounding area and rich in original intercellular connection. CT and MRI identify “abdominal mass” in the majority of patients with EGIST, which is diagnosed based on postoperative pathology. Currently, surgery remains the most common treatment for EGISTs. Although the use of imatinib has revolutionized treatment and improved outcomes for patients with GIST, little is known with regard to its efficacy in EGIST.

1 Introduction

Until 2002, stromal tumor has been re-recognized and correctly named, which was confused with leiomyomas for such a long time. A breakthrough has been made on stromal tumor over this short span of a decade. Especially the use of imatinib, a tyrosine kinase inhibitor (TKI), allows stromal tumor to become a model for targeted therapy. Gastrointestinal stromal tumor (GIST) is the most common gastrointestinal mesenchymal tumor, accounting for 2% of all gastrointestinal tumors. GIST often occurs in the stomach (40–70%), small intestine (20–40%), colon (5–15%), and rectum (5%). On the contrary, primary extragastrointestinal stromal tumor (EGIST) accounts only a very small percentage of all stromal tumors. EGIST occurs frequently in the mesentery, omentum, and peritoneum but rarely in the retroperitoneum. Up to now, only 59 cases have been reported. EGIST is commonly seen in young and middle-aged people, with an incidence comparable between men and women. EGIST often grows very large before it is detected incidentally, resulting in a low rate of complete resection and a high rate of recurrence (even after complete surgical resection). Although located in different primary sites, EGIST is substantially similar to GIST in terms of pathology and immune phenotype. Under microscope, tumor cells are short spindle, nest, sheet, or vesicle shaped, with translucent cytoplasm. Immunohistochemistry reveals CD117 positive, as well as actin, calponin, and S100 positive or weakly positive. C-kit gene mutation is detectable. Under electron microscope, tumor cells appear as irregular fusiform with intertwined pseudopodia of varying lengths extending toward the surrounding area and rich in original intercellular connection. CT and MRI identify “abdominal mass” in the majority of patients with EGIST, which is diagnosed based on postoperative pathology. Currently, surgery remains the most common treatment for EGISTs. Although the use of imatinib has revolutionized treatment and improved outcomes for patients with GIST, little is known with regard to its efficacy in EGIST.

2 Etiology and Pathogenesis

As noted above, genetic alteration and immunohistochemical characteristics of EGIST are substantially consistent with GIST, suggesting that both diseases may share common molecular mechanisms. Regarding the cellular origin, it is most widely accepted that GIST arises from mesenchymal progenitor cells or immature stem cells that are differentiated toward the ICC (interstitial cell of Cajal). ICC located intramurally are pacemaker cells to create the bioelectrical slow wave potential that leads to contraction of the smooth muscle. ICC is crucial to initiation and functional regulation of gastrointestinal motility. ICC is considered as the origin of GIST based on the following facts: (a) the activation of intestinal ICC requires the participation of C-kit; (b) the positive expression of CD117 is a classic immunophenotype of ICC; and (c) GIST cells highly resemble ICC under electron microscope.

Overwhelming amount of data have proved the close relationship between ICC and GIST; however, the cellular origin of EGIST is yet to be elucidated. From the view of anatomical position, ICC can be definitively ruled out as the cellular origin of EGIST due to their location within muscle layers of the gastrointestinal tract wall. A considerable part of EGIST cases reported in previous literatures might be originated from GIST and grow extraluminally. In the absence of visible anatomical connection to primary gastrointestinal tract, GIST is misdiagnosed as EGIST clinically. There is another possibility that EGIST may have deviated from the original site of the gastrointestinal tract in the process of tumorigenesis. It is also suggested that EGIST may originate from mesenchymal pluripotent stem cells of multiple differentiation potential. In a retrospective analysis conducted in 14 cases of EGIST, the stump of seromuscular layer is observed on the margin of tumor pseudocapsule in 8 cases after HE staining, which was further confirmed as desmin positive by immunohistochemistry. Two out of 14cases were finally diagnosed as extensive metastases of GIST into abdominal cavity. Interestingly, among them, one case was diagnosed with multiple celiac EGIST. Although the presence of GIST was ruled out, this patient had partial resection of the small intestine 5 years ago due to an abdominal mass in another hospital, and postoperatively histological findings suggested solitary fibromatosis, as mentioned in his medical history records. At that time, the diagnosis of GIST relied more on CD34, so that GIST might easily be misdiagnosed as solitary fibromatosis.

Recently a patient with large abdominal mass was admitted to Zhongshan Hospital of Fudan University. Endoscopy did not identify occupying lesions of stomach; however, CT image displayed a large mass in peritoneal cavity, indicating mesenchymal tumor of unknown origin (Fig. 21.1). Intraoperatively, a fairly large mass was found in the omentum, which was irregularly shaped, soft in nature, and easy bleeding by touch. After complete removal of the mass from the omentum, a 5-cm-long and 0.5-cm-wide (at the basal part) cord linking to the greater curvature of the stomach was identified. Postoperative pathology suggested GIST. Intraoperatively, the mass was easily detached from the cord with a little force. If the tumor was confused with the attached cord to be separated, it would be postoperatively misdiagnosed as EGIST (Fig. 21.2). This experience suggests that the incidence of EGIST may be much lower than expected. How to correctly define GIST is worthy of further discussion. In the clinical diagnosis, efforts should be taken to seek for sufficient evidence, thus providing accurate guidance for the postoperative treatment of those patients.
Fig. 21.1

CT image displays a huge mass, mesenchymal tumor of unknown origin in peritoneal cavity

Fig. 21.2

Intraoperatively the mass is easily detached from the cord under a little force, or it is frequently confused with attached cord to be separated and misdiagnosed as EGIST postoperatively

3 Pathology

EGIST mostly appears as a solid mass in gray-pink color, surrounded by pseudo-fibrous capsule, in round or lobulated shape, with cross section in gray or reddish brown color. Unlike leiomyoma, it lacks common spiral structure, occasionally accompanied by mucus degeneration, hemorrhage, necrosis, and cystic degeneration. Calcification is sporadically observed, relatively common in large tumors but not in small ones. Under microscope, the tissue structure is complex and cell morphology varies, similar to GIST. Based on cell morphology, it can be divided into three categories: (a) spindle cell type (70%), (b) epithelial cell type (20%), and (c) mixed (10%). Immunohistochemical markers associated with GIST are the most important tool for the diagnosis. Among them, the relatively specific parameters to be detected include CD117, DOG-1, and CD34, whereas nonspecific parameters that are mainly attributable to differential diagnosis include SMA, S-100, caldesmon, desmin, vimentin, and nestin (Table 21.1).
Table 21.1

Differential diagnosis of GIST with other tumors: immunohistochemical and molecular markers

Tumor

KIT

DOG1

CD34

SMA

CALDES

DES

S-100

Other

GIST

+

+

+

±

+

 

Smooth muscle tumor

+

+

+

 

Solitary fibromatosis

+

CD99, BCL-2

Inflammatory myofibroblastic tumor

+

+

ALK-1

Schwann cell tumor

+

 

Malignant peripheral nerve sheath tumor

+

GFAP

4 Clinical Manifestation and Diagnosis

Clinical manifestations of EGIST are similar to general retroperitoneal tumors. Chief complaints include (a) abdominal mass, lower back pain, abdominal pain, and bloating; (b) difficulty in urination and defecation and lower limb pain, if it occurs in retroperitoneum and pelvis; and (c) fever, if the tumor is necrotic. Some patients feel no obvious discomfort and consult doctors only because of anemia and fatigue. Due to deep location, large space from surrounding tissue, and late presentation of clinical symptoms, the volume of retroperitoneal EGISTs often grows large, with the maximal diameter up to 32 cm at the time of diagnosis.

Due to limited number of patients, the imaging data are scarce, and only a few studies on imaging features of EGIST have been reported. Preoperative imaging plays an important role in determining the origin of tumor, and the following manifestation contributes to predict the accurate location of tumor arising from the retroperitoneum: (a) the center or the maximum axis diameter of the tumor located in the retroperitoneum; (b) forward displacement of kidneys, pancreas, intestine, and other organs under compression; (c) embedded tumor pushing large abdominal vessels forward or laterally; and (d) fat space existing between the tumor and abdominal organs. Any tumor necrosis or hemorrhage, if occurs, will seriously affect the imaging manifestation of tumor. The diagnosis of EGISTs should be made based on specific imaging features combined with the general radiological characteristics of GIST through individual analysis, rather than mechanical imitation of imaging study of GIST. A case of primary retroperitoneal EGIST showed a well-defined and heterogeneous peripancreatic mass, about 16 × 15 × 18 cm3 in size, with annular calcification on CT image. Enhanced CT identified a heterogeneous mass, progressive contrast enhancement. The tumor compressed but did not invade the pancreas, stomach, or spleen. MRI showed a retroperitoneal mass with heterogeneous intensity signals on T1- and T2-weighted images and heterogeneous enhancement of the tumor on gadolinium-enhanced T1-weighted image with fat suppression. Generally, a small tumor is more homogeneous than a large one accompanied with calcification, necrosis, and other degenerative lesions. Malignant tumors are often surrounded by soft tissue, 86% of which having well-defined boundary. The density or signal strength of tumor is lower than that of enhanced liver during the same period. PET-CT may be considered for multiple EGISTs, although its clinical significance requires further investigation.

5 Risk Stratification System

Accurate assessment of the patient’s condition and choosing the optimal treatment regimen for individual patient have become the focus of personalized therapy. Since stromal tumor was not recognized until 2002, clinicians (Fletcher et al. 2002a, b) reached a consensus on the criteria for classification of GIST, which is now the most commonly accepted in clinical work. GIST usually presents malignant potential, while benign GIST does not exist at all. Therefore, malignant risk of GIST is classified into minimal, minor, intermediate, and high levels. Tumor size and mitotic figure are used as the specific parameters to identify GIST. However, the biggest problem arising from the cutoff value of mitotic figures at 50/HPFs in clinics is ignoring the significant effect of tumor location and progressive rupture on the assessment. Furthermore, such parameter cannot simply apply to EGIST. For this reason, such criteria have been improved. Franquemont et al. (1992, 1995) reported that R1 resection and rupture of stromal tumor are closely related to poor prognosis. Peritoneal implantation, tumor metastasis, and invasion of adjacent organs are considered as predictors for high-degree malignancy by Fletcher et al. (2002a, b). In their opinion, EGIST with the above risk factors, regardless of tumor size and mitotic activity, is classified as high-degree malignancy. On the basis of Fletcher’s opinion, Joensuu proposed new criteria including the presence of tumor rupture as a high-risk factor independent of size and mitotic count. Further modification in Joensuu’s criteria reclassified non-gastric tumors in the NIH intermediate category to the high-risk group (Table 21.2). According to NIH criteria, EGIST should be classified as high-risk tumor; however, due to lack of sufficient cases of EGIST, further studies are guaranteed to verify the applicability of such criteria in evaluation of EGIST.
Table 21.2

NIH risk stratification system for primary GIST after resection

Classification of risk degree

Tumor size (cm)

Mitotic index (50/HPF)

Primary site

Minimal

≤2.0

≤10

Any

Minor

2.1–5.0

≤5

Any

Intermediate

2.1–5.0

>5

Stomach

 

<5.0

6_10

Any

 

5.1–10

≤5

Stomach

High

Any

Any

Tumor rupture

 

>10

Any

Any

 

Any

>10

Any

 

>5.0

>5

Any

 

2.1–5.0

>5

Non-primary gastric

 

5.1–10

≤5

Non-primary gastric

6 Treatment

Surgery remains the most effective treatment for EGIST. The basic principles are similar to those for general retroperitoneal tumors, namely, striving for R0 resection. To avoid tumor capsule rupture and intraperitoneal spread intraoperatively, if the tumor invades adjacent organs, the involved organs should be en bloc resected together with the tumor. Laparotomy is a common surgical procedure for these patients. If the tumor is small (<2 to 5 cm) in size, laparoscopic surgery would be an option. Sometimes, in order to protect vital organs from being damaged, R1 resection is acceptable, but palliative surgery is not recommended. EGIST, especially retroperitoneal GIST, often grows very large and closely relates to surrounding organs and large blood vessels at diagnosis. Definite diagnosis allows patients to be treated with targeted drugs to shrink the tumor preoperatively. C-KIT exon 11 mutations have been identified in most of EGIST cases, so imatinib may be a preoperatively adjuvant therapy appropriate for them from this perspective. A patient with inoperable multiple peritoneal stromal tumors had experienced a significant reduction in tumor size after preoperatively receiving imatinib as neoadjuvant therapy and ultimately underwent R0 resection. As noted previously, the origin of EGIST remains controversial. If the tumor arises from GIST that grows extraluminally, this hypothesis would be convincible, further supporting the effectiveness of imatinib therapy. Collectively, targeted therapy can be used as an alternative strategy.

7 Efficacy and Prognosis Factors

It is now widely accepted that CHOI or RECIST criteria for GIST targeted therapy also apply to the efficacy evaluation for EGIST. According to RECIST criteria, a decrease of more than 30% in the maximum diameter of tumor size is considered as response. Patients with GIST who respond to targeted therapy tend to present varying changes in tissues in early stage, which are characterized by necrosis, hemorrhage, cystic degeneration, and mucus. The volume of tumor may not be obviously shrunk and even becomes slightly larger because of hemorrhagic necrosis. Therefore, Choi et al. (2007) proposed a new evaluation criteria—CHOI for GISTs in combination with dynamic change rate of strengthened CT value (Table 21.3)—which may provide a reference for retroperitoneal EGIST. The prognosis of EGIST is affected by many factors such as primary tumor site, size, mitotic figures, and necrosis, all of which are considered to be important predictors of clinical outcome. EGIST originating in retina results in better prognosis than that in mesentery. It is also reported that the survival period is shorter in patients aged <50 years old than those aged ≥50 years old. Additionally, Joensuu et al. (2002, 2012) demonstrated that active mitosis, large tumor size, and the presence of tumor necrosis could predict poor prognosis.
Table 21.3

CHOI criteria for efficacy of retroperitoneal extragastrointestinal stromal tumor

Efficacy

Definition

CR

Disappearance of all lesions. No new lesions

PR

≥10% decrease in maximum diameter of tumor size and/or a ≥15% decrease in tumor density (HU). No new lesions. No obvious progression of non-measurable disease

SD

Does not meet criteria for CR, PR, or PD. No symptomatic deterioration attributed to tumor progression

PD

≥10% increase in maximum diameter of tumor size; change in HU doesn’t meet PR criteria; the appearance of new lesions and intratumoral nodules or an increase in size of the existing intratumoral nodules

References

  1. Choi H, Charnsangavej C, Faria S, et al. Correlation of computed tomography and positron emission tomography in patients with metastatic gastrointestinal stromal tumor treated at a single institution with imatinibmesylate: proposal of new computed tomography response criteria. J Clin Oncol. 2007;25:1753–9.CrossRefPubMedGoogle Scholar
  2. Fletcher CD, Berman JJ, Corless C, Gorstein F, Lasota J, Longley BJ, Miettinen M, O’Leary TJ, Remotti H, Rubin BP, Shmookler B, Sobin LH, Weiss SW. Diagnosis of gastrointestinal stromal tumors: a consensus approach. Hum Pathol. 2002a;33:459–65.CrossRefPubMedGoogle Scholar
  3. Fletcher CDM, Berman JJ, Corless C, Gorstein F, Lasota J, Longley BJ, et al. Diagnosis of gastrointestinal stromal tumors: a consensus approach. Hum Pathol. 2002b;33:459–65.CrossRefPubMedGoogle Scholar
  4. Franquemont DW. Differentiation and risk assessmentof gastrointestinalstromal tumors. Am J Clin Pathol. 1995;103:41–7.CrossRefGoogle Scholar
  5. Franquemont DW, Frierson HF Jr. Muscle differentiation and clinicopathologic features of gastrointestinal stromal tumors. Am J Surg Pathol. 1992;16:947–54.CrossRefGoogle Scholar
  6. Joensuu H, Fletcher C, Dimitrijevic S, Silberman S, Roberts P, Demetri G. Management of malignant gastrointestinal stromal tumours. Lancet Oncol. 2002;3(11):655–44.CrossRefPubMedGoogle Scholar
  7. Joensuu H, Vehtari A, Riihimäki J. Risk of recurrence of gastrointestinal stromal tumour after surgery: an analysis of pooled population-based cohorts. Lancet Oncol. 2012;13(3):265–74.CrossRefPubMedGoogle Scholar

Copyright information

© Springer Science+Business Media B.V. 2018

Authors and Affiliations

  1. 1.Zhongshan Hospital of Fudan UniversityShanghaiChina

Personalised recommendations