| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Report |
1 Department of Radiologic Pathology, Armed Forces Institute of Pathology,
Alaska and Fern Streets, NW, Washington, DC 20306-6000.
2 Department of Radiology and Nuclear Medicine, Uniformed Services University of
the Health Sciences, 4301 Jones Bridge Rd., Bethesda, MD 20814-4799.
3 Present address: Department of Medical Education, Washington Hospital Center,
Washington, DC 20010-2975.
4 Department of Radiology, University of Maryland School of Medicine, Baltimore,
MD 21201-1544.
5 Department of Hepatic and Gastrointestinal Pathology, Armed Forces Institute
of Pathology, Washington, DC 20306-6000.
6 Department of Soft Tissue Pathology, Armed Forces Institute of Pathology,
Washington, DC 20306-6000.
Received December 31, 2003;
accepted after revision March 5, 2004.
Address correspondence to A. D. Levy
(levy{at}afip.osd.mil)
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. Gastrointestinal stromal tumors that occur in patients with neurofibromatosis commonly originate from the proximal small intestine and are often multiple. The cross-sectional imaging appearance of gastrointestinal stromal tumors that occur in patients with neurofibromatosis is similar to that of gastrointestinal stromal tumors that occur in the general population.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Gastrointestinal abnormalities in patients with neurofibromatosis 1 are reported to occur in up to 10–25% of patients and consist of four groups of lesions: mesenchymal neoplasms; hyperplasias of intestinal neural tissue; neuroendocrine tumors of the duodenum; and, rarely, other gastrointestinal neoplasms such as adenocarcinomas [2]. Although neurofibromas are the most commonly encountered mesenchymal neoplasm of the gastrointestinal tract in patients with neurofibromatosis 1, gastrointestinal stromal tumors, leiomyomas, and leiomyosarcomas also occur [3].
Gastrointestinal stromal tumors usually occur in middle-aged patients with neurofibromatosis 1 years after the appearance of cutaneous manifestations [2]. Gastrointestinal stromal tumors in patients with neurofibromatosis 1 are histologically and immunophenotypically identical to those in patients without neurofibromatosis 1. Studies in the pathology literature have shown that gastrointestinal stromal tumors in patients with neurofibromatosis 1 tend to be multiple and to be located predominantly within the small intestine [4]. To our knowledge, no case reports or series that describe the radiologic and imaging features of gastrointestinal stromal tumors in patients with neurofibromatosis 1 have yet been published. The purpose of our study was to evaluate the clinical, pathologic, and imaging features of gastrointestinal stromal tumors in a series of six patients with neurofibromatosis 1.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Clinical Data and Pathology Review
Clinical, pathology, and surgical data were reviewed for patient age, sex,
and presenting signs and symptoms. All patients except one underwent complete
surgical excision of their tumors. Surgical and pathology records were
reviewed for tumor location and extension into adjacent organs. If patients
had additional neoplasms or manifestations of neurofibromatosis 1 removed at
surgery, these findings were also noted. Histopathology reports and slides
were available for all patients. All pathologic material was rereviewed by
experienced gastrointestinal and soft-tissue pathologists. Histopathology
findings and mitotic activity (number of mitoses per 50 consecutive high power
fields) in all six cases were assessed by a gastrointestinal and soft-tissue
pathologist. The pathology records of each patient were reviewed to establish
tumor size and immunoreactivity with CD117 (KIT) and CD34. Tumors were
considered probably benign if they were 2 cm or smaller and had no more than
five mitoses per 50 high-power fields. Tumors were considered malignant if
they were larger 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 larger than 2 cm but 5 cm or smaller and had no more than five
mitoses per 50 high-power fields
[5].
Imaging Studies and Review
Imaging studies were available in all six patients. Four patients underwent
abdominal CT as the primary imaging technique. One of the patients who
underwent CT also had abdominal MRI, and two other patients also had
conventional small-bowel barium examinations. Of the two patients who did not
undergo abdominal CT for the initial diagnosis, one had abdominal MRI and the
other had gastrointestinal bleeding scintigraphy, mesenteric angiography, and
enteroclysis. The imaging techniques were not standardized because our
patients were referred from many institutions. Their studies were performed
using a variety of equipment, and differing protocols were followed regarding
slice thickness and contrast injection.
All CT examinations (n = 4) were performed on helical scanners with slice thickness ranging from 5 to 10 mm. Two patients underwent contrast-enhanced CT. One patient received IV contrast material, and both IV and oral contrast material were administered in another patient. The MR images (n = 2) were obtained with body coils. Axial T1-weighted, fast spin-echo T2-weighted, and fat-saturated gadolinium-enhanced T1-weighted images were available for review in both patients.
Two abdominal radiologists reviewed all images retrospectively, and the final interpretation was made by consensus. Only the final histopathologic diagnosis of gastrointestinal stromal tumor was known at the time of image interpretation. Information regarding the number of gastrointestinal stromal tumors and the presence of additional pathologic diagnoses was not known at the time of interpretation. The CT scans and MR images were evaluated for the presence or absence of a mass or abnormality in the stomach, intestine, mesentery, adrenal glands, pancreas, liver, bile ducts, retroperitoneum, pelvis, subcutaneous tissues, or visualized skeletal structures. The presence or absence of adenopathy and any abnormality in the lower chest were noted. If a mass was present, the size, shape (round, oval, infiltrating, or elongated), margins (well-defined or ill-defined), and enhancement pattern (no enhancement, homogeneous, heterogeneous, or rimlike) were documented. The mass was also evaluated for the presence or absence of cystic change, degeneration, cavity formation, and calcification.
The three small-bowel barium studies (conventional small-bowel examination [n = 2] and enteroclysis [n = 1]) were evaluated for tumor location (duodenum, jejunum, ileum); tumor location with respect to the intestinal wall (intraluminal, mural, extrinsic); ulceration; and additional findings, such as intussusception or obstruction.
After imaging review, clinical data, surgical records, pathology reports, and photographs of the gross pathologic specimen (n = 5) were analyzed with the available imaging studies. The gross appearance of the lesion on photography (size, shape, margins, presence of internal hemorrhage, and presence of cystic change) was compared with imaging appearance.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Histopathologic and Immunophenotypic Features
Five patients (83%) had multiple lesions
(Table 1). Preoperative imaging
identified multiple tumors in only one of these patients. Four of the five
patients had multiple small-intestinal gastrointestinal stromal tumors. The
dominant tumor in these patients was confirmed to be a gastrointestinal
stromal tumor. The other patient with multiple lesions had one
small-intestinal gastrointestinal stromal tumor and two tumors of other types.
In this patient, the dominant tumor was a 5.5-cm polypoid submucosal
small-intestinal ganglioneuroma, the gastrointestinal stromal tumor was a
2.0-cm serosal nodule, and the third tumor was a 3.5-cm small-intestinal
adenoma containing a focus of adenocarcinoma. The latter two neoplasms were
incidental findings at surgery and were not identified on preoperative
imaging.
|
All gastrointestinal stromal tumors originated from the small intestine.
Four of the six cases had documented coexpression of KIT and CD34. The
remaining two cases with no KIT data were included in the study because of the
histologic identity of the KIT-positive tumors. Four of the gastrointestinal
stromal tumors were classified as malignant: one, high-grade malignant (>
50 mitoses per 50 high-power fields) and three, low-grade malignant (
50
mitoses per high-power fields). Of the remaining two tumors, one was
considered to be of uncertain malignant potential, and the other was
classified as benign.
CT Features
Three of the four patients who underwent CT had round, well-defined,
heterogeneously enhancing masses on CT. The enhancement pattern in each of
these cases was heterogeneous rimlike with central areas of hypoattenuation.
One contained a gas-filled cavity and punctate calcification and was located
in the distal duodenum (Fig.
1A,
1B,
1C). The other two had
significant low-attenuation cystic change and were located in the mid jejunum
with extension into the small-bowel mesentery. All these masses were
histologically confirmed to be malignant gastrointestinal stromal tumors.
|
|
|
The fourth patient who underwent CT had an intraluminal, oval, homogeneously enhancing mass that was the lead point for an intussusception (Fig. 2A, 2B, 2C, 2D). Histologically, this mass corresponded to the ganglioneuroma of the patient whose gastrointestinal stromal tumor was an incidental finding at surgery and could not be identified on CT.
|
|
|
|
Of the four patients with multiple gastrointestinal stromal tumors, preoperative CT showed multiple tumors in only one patient (Fig. 3A, 3B, 3C). This patient had two dominant gastrointestinal stromal tumors: one was a mixed solid–cystic mass and the other was predominantly cystic (Fig. 3B). A smaller, homogeneous-attenuation mass was noted in the proximal jejunum (Fig. 3A). Histologically, all these masses were gastrointestinal stromal tumors, and the patient also was found at surgery to have multifocal gastrointestinal stromal tumors carpeting the proximal jejunum (Fig. 3C).
|
|
|
Only one patient had additional CT features of neurofibromatosis. These findings included a histologically proven adrenal gland pheochromocytoma and multiple subcutaneous, psoas, and pleural masses consistent with neurofibromas (Fig. 2A, 2B, 2C, 2D). No CT evidence of liver metastasis, ascites, or peritoneal metastasis was seen in any of the patients.
MRI Features
MRI showed a well-defined, round to oval mass located primarily in the
small-bowel mesentery in the two patients who underwent MRI. In each case, the
small-bowel origin of the mass was not evident on MRI because the masses
appeared to be located predominantly in the small-bowel mesentery (Fig.
4A,
4B,
4C). One patient had a
predominantly solid mass that was low signal on T1-weighted images,
heterogeneously high signal on T2-weighted images, and heterogeneously
enhanced with gadolinium. The second patient had a cystic mass with an
irregular soft-tissue rim that was low signal on T1-weighted images, high
signal on T2-weighted images, and heterogeneously enhanced with gadolinium
(Figs. 4A and
4B).
|
|
|
Barium Features
In two of the three patients with small-bowel barium examinations, smoothly
marginated intraluminal filling defects were seen within the jejunum. In one
patient, the mass intussuscepted during the examination. The intussuscepting
mass was shown to be a ganglioneuroma at pathology. No evidence of ulceration
was found in either of these cases. The third patient had a cavitary, mural
mass in the fourth portion of the duodenum that contained a sinus tract
between the cavity and duodenal lumen (Fig.
1C).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mesenchymal neoplasms, most commonly neurofibromas, are known to occur with increased frequency in patients with neurofibromatosis 1 [2]. The association of gastrointestinal stromal tumors in patients with neurofibromatosis 1 has also been noted in the literature [4, 8–10]. The occurrence of gastrointestinal stromal tumors in the neurofibromatosis 1 population deserves attention because the distribution and multiplicity of gastrointestinal stromal tumors differ from those occurring in patients without neurofibromatosis 1. Moreover, gastrointestinal stromal tumors have been reported to be the initial clinical presentation of an adult patient with neurofibromatosis 1 who had subtle cutaneous and neurologic stigmata of the disease [8].
Although our study population was small in number, the clinical and pathologic features of our patients support the findings that have been previously reported in the literature [8–10]. Our patients presented during middle age (mean age, 43 years). Abdominal pain, followed by gastrointestinal bleeding, was the most common presenting clinical manifestation. The gastrointestinal stromal tumors in all our patients were localized to the small intestine. In contrast, the stomach is the most common location for gastrointestinal stromal tumors overall [7]. Finally, four of our patients had multiple gastrointestinal stromal tumors, and one of our patients had a gastrointestinal stromal tumor in addition to multiple other neoplasms. Multiplicity and an association with other neoplasms are consistent with the previously reported cases [4, 8–10].
Our series of patients was heterogeneous in terms of prognosis, with both benign and malignant tumors. Four of our patients had a dominant malignant gastrointestinal stromal tumor; one, uncertain malignant potential; and one, benign. The CT scans and MR images did not show evidence of ascites or peritoneal or liver metastasis in any of our patients with malignant gastrointestinal stromal tumors.
The imaging features of gastrointestinal stromal tumors have been recently reported [11, 12]. Our patients had tumors characterized by CT and MRI findings of heterogeneously enhancing masses with focal areas of CT hypoattenuation or cystic change. Pathologically, the gastrointestinal stromal tumors in our patients originated from the intestinal wall with or without extension into the small-bowel mesentery. These findings are consistent with the previously described features of gastrointestinal stromal tumors.
We had a single case in which the patient had an intussuscepting ganglioneuroma, a gastrointestinal stromal tumor, and an adenocarcinoma. In this patient, the gastrointestinal stromal tumor and adenocarcinoma were not identified on the preoperative CT or small-bowel barium examination, even in retrospect. Notably, the CT features of the ganglioneuroma (homogeneous, nonenhancing, intraluminal mass) are not consistent with the typical features of a gastrointestinal stromal tumor.
Significantly, in four (80%) of the five patients with multiple lesions,
only one tumor was identified on preoperative CT, MRI, or barium examination.
In several instances, the tumors were small ( 1.0 cm) and may have been
obscured by the dominant mass or by suboptimal opacification of the small
bowel with oral contrast material on CT. In two patients, however, the
secondary gastrointestinal stromal tumors were larger (2.0 and 2.6 cm) and
should have been detected on CT or small-bowel barium examination. One of
these was likely obscured by the intussusception complex on both CT and barium
evaluations. In the other patient, a small-bowel barium examination was
performed but overlapping intestinal segments on the images available for
review may have obscured the lesion.
Several limitations are inherent to our study because our patients were referred from many different institutions. The lack of standardization of imaging parameters and injection techniques may have influenced the final radiologic interpretations. We had the advantage, however, of being able to review surgical and pathology records in all patients and photographs of resected gross specimens in five patients to ensure that the radiologic interpretations were correct. For those patients who had small-bowel barium examinations, our interpretations were limited because we did not have the advantage of real-time fluoroscopy.
The differential diagnosis of mesenchymal neoplasms arising from the wall of the small bowel in a patient with neurofibromatosis 1 includes neurofibroma, gastrointestinal stromal tumor, leiomyosarcoma, and leiomyoma. All these tumors may extend into the small-bowel mesentery or have a dominant component within the small-bowel mesentery.
Gastrointestinal neurofibromas in patients with neurofibromatosis 1 may be solitary or multiple. Plexiform neurofibromas are considered pathognomonic of neurofibromatosis 1. On CT, neurofibromas appear as homogeneous soft tissue in the intestinal wall or mesentery [13]. They may originate in the mesentery and extend to encase segments of small bowel or colon [14]. Occasionally, neurofibromas may contain areas of low attenuation from myxoid degeneration within the tumor [15].
The differential diagnosis of small-bowel masses in patients with neurofibromatosis 1 should also include adenocarcinoma, carcinoid tumor, and lymphoma. These are the most common small-bowel neoplasms seen in the general population. Adenocarcinoma is the most common and typically manifests as an annular lesion in the proximal small bowel. The appearance of adenocarcinoma usually does not overlap with that of gastrointestinal stromal tumor. Lymphoma, however, may have a similar appearance to a gastrointestinal stromal tumor because lymphoma of the small intestine may ulcerate, cavitate, and extend into the adjacent mesentery. However, in most cases of lymphoma, features such as low-attenuation hemorrhage, degeneration, and cystic change are usually not present. Carcinoid tumors of the small bowel tend to be concentrated in the duodenum and ileum and may be multiple. They may manifest as a focal polypoid mass or mural thickening. Those occurring in the distal small bowel often have adjacent mesenteric involvement, which stimulates considerable desmoplasia and fibrosis. The result is an angulated or kinked small bowel that is prone to obstruction and ischemia.
Finally, lesions that originate in the mesentery such as mesenteric fibromatosis (desmoid tumor), inflammatory myofibroblastic tumor (inflammatory pseudotumor), sclerosing mesenteritis, and metastatic disease should be considered in the differential diagnosis. Mesenteric fibromatosis, in particular, may be mistaken for a gastrointestinal stromal tumor on imaging and at histology. Typically mesenteric fibromatosis is hypoattenuating on CT because of the presence of a myxoid stroma. It appears as a well-circumscribed mesenteric mass on CT that invades adjacent segments of small intestine.
In summary, gastrointestinal stromal tumors in patients with neurofibromatosis 1 tend to be multiple and located in the proximal small intestine. The imaging manifestations of gastrointestinal stromal tumors in patients with neurofibromatosis 1 are similar to gastrointestinal stromal tumors that occur in the general population. Our series indicates that solitary gastrointestinal stromal tumors may occur but are uncommon. Therefore, careful attention and meticulous technique should be practiced when interpreting CT scans and MR images in patients with neurofibromatosis and suspected small-bowel gastrointestinal stromal tumors.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A. D. Levy, N. Patel, N. Dow, R. M. Abbott, M. Miettinen, and L. H. Sobin From the Archives of the AFIP: Abdominal Neoplasms in Patients with Neurofibromatosis Type 1: Radiologic-Pathologic Correlation RadioGraphics, March 1, 2005; 25(2): 455 - 480. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
