| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Report |
1 Department of Radiologic Pathology, Armed Forces Institute of Pathology, 6825
16th St., N.W., Washington, DC 20306-6000.
2 Department of Radiology and Nuclear Medicine, Uniformed Services University of
the Health Sciences, 4301 Jones Bridge Rd., Bethesda, MD 20814.
3 Department of Hepatic and Gastrointestinal Pathology, Armed Forces Institute
of Pathology, Washington, DC 20306-6000.
4 Department of Radiology, Duke University, Box 3808, Durham, NC 27710.
5 Department of Soft Tissue Pathology, Armed Forces Institute of Pathology,
Washington, DC 20306-6000.
Received September 30, 2002;
accepted after revision November 13, 2002.
Address correspondence to A. D. Levy
(levya{at}afip.osd.mil).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. Anorectal gastrointestinal stromal tumors are mesenchymal neoplasms that typically arise in the muscularis propria of the intestinal wall. The cross-sectional imaging appearance is that of a well-defined mural mass that may have an exophytic component and may invade adjacent structures. A prominent intraluminal component is a rare feature.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the earlier medical literature, gastrointestinal stromal tumors were not distinguished from other mesenchymal neoplasms and therefore were classified as smooth muscle tumors, leiomyomas, leiomyosarcomas, epithelioid leiomyosarcomas, and leiomyoblastomas. KIT immunoreactivity not only distinguishes gastrointestinal stromal tumors as unique from these other mesenchymal neoplasms but also is the target for KIT-inhibitor therapy. Recent availability of KIT-tyrosine kinase inhibitor (STI-571, imatinib [Gleevec], Novartis, Basel, Switzerland) for successful treatment of malignant gastrointestinal stromal tumors mandates a high level of awareness of gastrointestinal stromal tumors in all stages of tumor diagnosis and therapy [4, 5].
Gastrointestinal stromal tumors are the most common mesenchymal neoplasms of the gastrointestinal tract [6]. The term "gastrointestinal stromal tumor" does not encompass leiomyomas and leiomyosarcomas. Leiomyomas occur most frequently in the esophagus as intramural lesions and less commonly in the colorectum as diminutive lesions involving the muscularis mucosae. Leiomyosarcomas are rare in the gastrointestinal tract. Gastrointestinal stromal tumors are most frequently seen in the stomach (70%), followed by the small intestine (20–30%), anorectum (7%), colon (5%), and esophagus (<5%) [6]. Gastrointestinal stromal tumors may also occur primarily in the omentum, mesentery, and retroperitoneum [7]. They occur most often in patients more than 50 years old, and they are rare before the age of 40 [6].
Although a number of studies on rectal and anorectal gastrointestinal stromal tumors have been published in the pathology literature [8, 9, 10], there are only single patient reports of the imaging appearance of anorectal gastrointestinal stromal tumors [11, 12]. The purpose of our study was to describe the CT and MR imaging features of anorectal gastrointestinal stromal tumors with clinical and pathologic correlation. Knowledge of these features may permit imaging differentiation of these tumors from more common rectal tumors such as adenocarcinoma and lymphoma. To our knowledge, ours is the largest study of anorectal gastrointestinal stromal tumors in the radiology literature.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Clinical data were reviewed for patient age, sex, and presenting signs and symptoms. Histopathology and mitotic activity (number of mitoses per 50 consecutive high-power fields) in all six patients were assessed by a gastrointestinal pathologist. The pathologic records of each patient were reviewed to establish tumor size and immunoreactivity with KIT and CD34. Tumors were considered probably benign if they were less than or equal to 2 cm and had no more than five mitoses per 50 high-power fields. Tumors were considered malignant if they were greater than 5 cm or had more than five mitoses per 50 high-power fields. Tumors were considered to have uncertain or low malignant potential if they were greater than 2 cm but less than or equal to 5 cm and had no more than 5 mitoses per 50 high-power fields [13]. Photographs of gross specimens (n = 4) were evaluated for evidence of hemorrhage, necrosis, and cyst formation.
Cross-sectional imaging studies of the anorectal region were available in all patients. Each patient (n = 6) underwent CT scanning. Three examinations were enhanced and three were unenhanced. No patient had both enhanced and unenhanced studies. Three patients underwent MR imaging (all three patients had both T1-weighted and fast spin-echo T2-weighted images). Two examinations were gadolinium-enhanced and one was unenhanced. All MR imaging was performed with body coils. Because our patients are referred from many institutions, studies were performed on a variety of equipment, and differing protocols were followed, so techniques were not standardized. Two radiologists reviewed all radiologic studies retrospectively, with final interpretations by consensus. After review of the radiologic studies, correlation with pathology reports in all patients and photographs of gross specimens (n = 4) was performed.
Findings of CT and MR imaging were evaluated for the location of the tumor with respect to the bowel wall. Each tumor was assessed for the presence of intraluminal, intramural, and extramural components. The adjacent organs, fat planes, and pelvic sidewalls were assessed for evidence of invasion, which was suspected when there was focal enlargement of the organ or anatomic structure in direct continuity with the tumor mass. Extension of tumor into the fat of the supralevator space and ischiorectal fossa was suspected when these areas contained focal soft-tissue density or soft-tissue stranding in direct continuity with the tumor mass. These spaces were also evaluated for the presence of lymphadenopathy. The tumors were assessed for areas of low attenuation on CT and the pattern of attenuation during the administration of contrast material. Likewise, findings on MR images were assessed for tumor signal intensity and enhancement patterns. CT and MR imaging findings were also evaluated for evidence of a tumor capsule, calcification, cyst formation, and secondary findings such as abdominal adenopathy and liver metastasis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Pathologic Findings
All patients were confirmed as having gastrointestinal stromal tumors
histologically, and findings showed a spindle cell pattern on light microscopy
(Fig. 1). In three of the six
patients, the tumor involved the rectal wall only; in the remaining three
patients, tumor involved the wall of both the rectum and anus.
Immunohistochemical studies were documented in four of the six tumors. In
these four tumors, there was coexpression of KIT and CD34. The remaining two
patients with no KIT data were included on the basis of the histologic
identity with the KIT-positive tumors.
|
Tumor size ranged from 5 to 12 cm in greatest dimension (mean, 7.8 cm). Mitotic activity was seen in all tumors. All tumors were classified as malignant on the basis of their size or mitotic activity. In two patients, the tumors that had fewer than five mitoses per 50 high-power fields were classified as malignant on the basis of their size of greater than 5 cm.
Photographs of gross specimens (n = 4) showed well-defined tan or white masses with central areas of hemorrhage (Figs. 2A, 2B, 2C, 2D, 2E). No evidence of necrosis or capsule formation was seen in any of the tumors.
|
|
|
|
|
Imaging Findings
CT findings in all patients showed an eccentric mural mass and well-defined
tumor margins (Fig. 3).
Findings in four patients showed an intraluminal component that created
irregular mucosal margins. Although the remaining two patients did not have a
definite intraluminal component, findings did show smooth deformity of the
overlying rectal mucosae. Extramural extension was present in four of the six
patients. Findings in three patients showed extension into the ischiorectal
fossa (Figs. 2A,
2B,
2C,
2D,
2E) and, in one patient, showed
invasion into the prostate gland (Fig.
4).
|
|
The CT attenuation was heterogeneous in all patients, with areas of central low attenuation that corresponded to areas of remote hemorrhage in those patients with photographs of gross specimens (Figs. 2A, 2B, 2C, 2D, 2E). The three largest tumors (> 8.0 cm) had larger areas of central low attenuation on CT (Fig. 4) when compared with those in the remaining three patients. The areas of low attenuation were evident on unenhanced and enhanced images. In three patients, the low-attenuation areas resembled well-defined cystic spaces (Figs. 2A, 2B, 2C, 2D, 2E and 5A, 5B, 5C, 5D). The addition of IV contrast material (n = 3) made the areas of low attenuation more conspicuous because only the solid portions of the tumor showed enhancement.
|
|
|
|
No evidence of calcification or a tumor capsule was seen in any patient. CT findings did not reveal perirectal, pelvic, or upper abdominal adenopathy. In addition, no imaging evidence of liver metastasis was seen in any of our patients.
MR images (n = 3) showed low-signal-intensity tumor on T1-weighted images with slightly higher signal-intensity areas focally in the tumor (Figs. 2A, 2B, 2C, 2D, 2E and 5A, 5B, 5C, 5D) that corresponded to areas of hemorrhage on gross pathology (n = 3). Solid areas of tumor without hemorrhage showed high signal intensity on T2-weighted images. The solid portions of the tumor enhanced with gadolinium in the two patients in whom it was used.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although findings in all our patients showed intramural masses, the presence of significant exophytic tumor may make it difficult to determine the organ of origin on cross-sectional imaging alone. For example, anterior extension of tumor into the prostate gland in male patients makes it difficult to differentiate a rectal gastrointestinal stromal tumor from a prostatic neoplasm. Histologically, rectal gastrointestinal stromal tumors extending into the prostate should be differentiated from leiomyosarcomas of the prostate. Those tumors that had significant extension into the ischiorectal fossa (and in one of our patients, extension to the perineum and base of the penis) closely resembled the appearance of malignancies arising from the ischiorectal fossa (such as soft-tissue sarcomas) and the perineum or sarcomas of the corpus cavernosum in men.
Clinically, most of our patients were older men (83%), as were the two patients described in the patient reports by Hama et al. [12] and van den Berg et al. [11]. A male predominance in anorectal gastrointestinal stromal tumors has also been reported in larger pathologic series [9, 10]. The most common clinical presentation in our patients was pain, likely due to the fact that most of the tumors in our series had an exophytic rather than endoluminal growth pattern. This finding may suggest that pain is more likely associated with gastrointestinal stromal tumors than with adenocarcinoma or lymphoma, which classically presents with bleeding. However, our series is small and likely underrepresents the full clinical spectrum of anorectal gastrointestinal stromal tumors. In large pathology series, anorectal gastrointestinal stromal tumors most commonly present with bleeding [9, 10].
The biologic behavior of gastrointestinal stromal tumors is often difficult to predict. Tumor size and mitotic rate are most frequently used to predict malignant behavior [13]. However, because benign-appearing gastrointestinal stromal tumors (small size and absent mitoses or low mitotic rate) may recur or metastasize, a recently published consensus statement suggests classifying gastrointestinal stromal tumors into very low–, low-, intermediate-, and high-risk categories, rather than as benign or malignant [14]. In the largest pathologic series of anorectal gastrointestinal stomal tumors with long-term follow up, Miettinen et al. [13] established that 70% of rectal gastrointestinal stromal tumors less than or equal to 2 cm in maximal dimension and having no more than five mitoses per 50 high-power fields showed malignant behavior and were associated with significant tumor-related mortality rates [9]. Tumors in all our patients were considered malignant by these criteria. Only one of our patients, however, at presentation had imaging features that suggested malignancy by the tumor's invasive growth. This patient had unequivocal evidence of prostatic invasion. It is not surprising that only one patient had imaging evidence of malignancy because the malignant behavior of gastrointestinal stromal tumors usually does not become clinically apparent until liver metastasis develops, often years after the primary lesion is surgically removed.
The imaging differential diagnosis of anorectal masses includes epithelial neoplasms (carcinomas arising from the rectum and anal canal), lymphoma, melanoma, carcinoid, and other mesenchymal neoplasms. Our study shows that anorectal gastrointestinal stromal tumors tend to have smooth margins without true capsules. The tumors in our series were large and contained evidence of remote hemorrhages that were manifested by low attenuation and cystic spaces on CT. In contrast, adenocarcinomas of the rectum tend to have irregular margins with strands of soft tissue extending from the tumor margin into the fat of the ischiorectal fossa or supralevator space. Perirectal lymphadenopathy is commonly present in adenocarcinomas of the rectum and was not observed in any of our patients with anorectal gastrointestinal stromal tumors. In addition, the presence of detectable hemorrhage on MR imaging is a helpful differentiating feature that favors gastrointestinal stromal tumors.
Primary anorectal lymphoma may mimic anorectal gastrointestinal stromal tumors because both neoplasms may present as mural masses and have overlying mucosal ulcerations. However, anorectal lymphoma occurs primarily in the HIV-positive population. Other imaging features that may be seen in primary anorectal lymphoma include tumor homogeneity, concentric wall thickening, intraluminal polypoid masses, fistula formation, thickening of the adjacent levator ani muscles, and adenopathy [15]. Therefore, the findings of a large circumscribed heterogeneous anorectal mass with evidence of remote hemorrhage and no surrounding adenopathy favors a gastrointestinal stromal tumor rather than adenocarcinoma or lymphoma.
There are several inherent limitations to our study. Imaging of our patients was performed at various institutions with different equipment and protocols. The lack of standardization of imaging parameters and injection techniques cannot be overcome in a study population of this type and may influence final radiologic interpretation. We had the advantage, however, of reviewing resected gross specimens and pathologic records to help ensure that radiologic interpretation was correct. In addition, a natural selection bias in referred patients limits the approximation of our study population to a naturally occurring population of patients. Finally, our assessment of the intraluminal component of the tumor was limited by the degree of luminal distention, which varied from study to study. CT and MR imaging are not optimal modalities to evaluate the mucosal surface characteristics of a tumor originating in the gastrointestinal tract wall. The addition of barium examinations and endoscopic correlation would have strengthened this portion of our study.
In conclusion, anorectal gastrointestinal stromal tumors are uncommon mesenchymal neoplasms. However, gastrointestinal stromal tumor should be suggested as the diagnosis in a patient who has a large, well-marginated anorectal mass that contains evidence of remote hemorrhage and lacks perirectal adenopathy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H J KIM, G H LEE, and Y W KIM Imaging features of perianal leiomyoma Br. J. Radiol., August 1, 2009; 82(980): e168 - e170. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Pickhardt, D. H. Kim, C. O. Menias, D. V. Gopal, G. M. Arluk, and C. P. Heise Evaluation of Submucosal Lesions of the Large Intestine: Part 1. Neoplasms RadioGraphics, November 1, 2007; 27(6): 1681 - 1692. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. C. Hoeffel, L. Azizi, N. Mourra, M. Lewin, L. Arrive, and J.-M. Tubiana MRI of rectal disorders. Am. J. Roentgenol., September 1, 2006; 187(3): W275 - W284. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-C. Kim, J. M. Lee, K. W. Kim, S. H. Park, S. H. Kim, J. Y. Lee, J. K. Han, and B. I. Choi Gastrointestinal Stromal Tumors of the Stomach: CT Findings and Prediction of Malignancy Am. J. Roentgenol., October 1, 2004; 183(4): 893 - 898. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-C. Kim, J. M. Lee, K. R. Son, S. H. Kim, K. H. Lee, K. W. Kim, M. Lee, J. K. Han, and B. I. Choi Gastrointestinal Stromal Tumors of the Duodenum: CT and Barium Study Findings Am. J. Roentgenol., August 1, 2004; 183(2): 415 - 419. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-C. Kim, J. M. Lee, S. H. Kim, K. W. Kim, M. Lee, Y. J. Kim, J. K. Han, and B. I. Choi Primary Gastrointestinal Stromal Tumors in the Omentum and Mesentery: CT Findings and Pathologic Correlations Am. J. Roentgenol., June 1, 2004; 182(6): 1463 - 1467. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
