The MDCT images of 11 IMTs, which primarily developed in the GI tract, were reviewed in this study. The MDCT findings of GI IMTs can be classified into two patterns: Wall-thickening and solitary mass-forming patterns. Each growth pattern was well matched with a characteristic pathological pattern: Wall-thickening pattern with the myxoid-vascular pattern and solitary mass-forming pattern with compact spindle cell pattern. In the present study, the GI IMTs met the requirements of only two of the three main pathological patterns, that is, compact spindle cell pattern and myxoid vascular pattern.
Tumors with the myxoid vascular pattern showed similar findings to those of GI tract lymphoma, which could present as infiltrative wall thickening or a well-circumscribed mass unassociated with intestinal obstruction, even if the mass is large. Tumors with a compact spindle cell pattern showed similar findings to those of subepithelial tumors, such as GIST, but they did not exhibit hypervascularity. There was no bowel obstruction in either type of GI IMT despite its large size, except two cases, which showed intraluminal growth and small bowel intussusception.
Since GI IMT is generally a rare disease, imaging findings are not well-established, and their relationship with disease prognosis has not been sufficiently studied. In this regard, Lee et al. examined five IMTs of the stomach and reported that gastric IMT appeared with two features, including a well-defined subepithelial tumor or irregular wall thickening, with a commonly strong enhancement (
23). This appearance is consistent with the results of the present study on the IMT of the entire GI tract. It has been reported that larger lesions may result in central necrosis (
4). There was no significant difference between the two pathological patterns of IMTs. In a study investigating abdominal IMT imaging findings, including the bowel, liver, and mesentery, it was reported that the compact cell type and hypocellular fibrous type formed a mass with discrete margins, which is also consistent with our findings (
11).
Inflammatory myofibroblastic tumor has emerged as a distinct disease entity with characteristic clinical, pathological, and molecular features, although the term “inflammatory pseudotumor” has been used to describe a wide range of reactive and neoplastic lesions (
2). It is known that chromosomal translocations leading to the activation of ALK tyrosine kinase can be detected in approximately half of IMTs; these features are known to suggest neoplasms of clonal change rather than reactive changes caused by inflammation (
2). Coffin et al. reported that ALK-negative IMTs were associated with older age and a higher rate of distant metastasis (
15). In the current study, there were no ALK-positive IMT patients. This result is because of the fact that patients in this study were relatively older than those with the usual features of IMT; these findings are consistent with the results of the previous study by Coffin et al. (
15).
Additionally, it is known that a large part of inflammatory pseudotumor involves immunoglubulin G4 (IgG4) infiltration in the serum and tissues, which is classified as an inflammatory pseudotumor associated with IgG4-related sclerosing disease (
24). Immunoglubulin G4-related inflammatory pseudotumors show typical microscopic features, including infiltration of many lymphocytes, IgG4-positive plasma cells, fibrosclerosis, and obstructive phlebitis. Compared to IMT, IgG4-related inflammatory pseudotumor usually develops in older patients and is commonly associated with autoimmune pancreatitis and sclerosing cholangitis. However, in the present study, IgG4 immunofluorescence stain for histological specimens and IgG4 serum level test were not performed for all cases. In this study, findings, such as diffuse infiltration of inflammatory cells with prominent plasma cells and low mitotic rate without atypical forms, were consistent with features suggestive of IMT.
The IMTs are known to be tumors of intermediate malignancy potential according to the WHO classification due to their tendency for local recurrence and a low risk of distant metastasis (
2). The recurrence of IMT following resection has been reported in 18% to 40% of cases and appears to be higher in extrapulmonary lesions, which are larger than 8 cm or locally invasive (
22). In our patients, hepatic metastasis was found in gastric IMT with a compact spindle cell pattern. There was no tumor recurrence in IMTs with a compact spindle cell pattern during the follow-up. Two patients with IMTs of myxoid vascular pattern showed perienteric infiltration and regional lymphadenopathy, one of whom experienced local recurrence after surgery.
The usual treatment for IMT is surgical resection, and additional chemotherapy is required if the surrounding invasion is severe, or if there is metastasis. However, some studies have reported improvements by administering anti-inflammatory drugs, including steroids and nonsteroidal anti-inflammatory drugs (
25,
26). However, it is not known which medical therapy is effective for which pathological type. It is assumed that medical treatment using anti-inflammatory drugs can be more helpful for the treatment of tumors containing various inflammatory cells, with severe infiltration into the surrounding area. However, further study on a larger sample size is needed. In the present study, steroid treatment caused improvements without any need for surgery; the tumor was a compact cell type with a solitary mass-forming pattern.
This study has several limitations. First, the sample size was limited. Only 11 patients were included in this study because of the rarity of GI IMTs. Therefore, all cases showed negative results for ALK, and no hypocellular fibrous pattern was included. Also, the clinical significance of each immunohistochemical study could not be evaluated; therefore, further research should be conducted on a sufficient sample size. Second, the follow-up period was relatively short. The duration of follow-up ranged from seven months to seven years (mean, 5 years); consequently, we cannot rule out the possibility of late-onset metastasis after treatment. Third, there may be conflicts regarding the pathological type of IMT. In some cases, it was difficult to determine the histological subtype, because there were overlapping areas of two histological subtypes; the main histological subtype was determined according to its composition. Finally, the CT protocol was inconsistent, because multicenter cases over a long period were collected, and quantitative analyses, such as texture analysis or dynamic enhancement pattern, could not be performed.
In conclusion, the imaging findings of GI IMTs in adults on CT scans can be classified into two patterns: Wall-thickening and solitary mass-forming patterns. Each growth pattern was well-matched with a characteristic pathological subtype (myxoid vascular pattern or compact spindle cell pattern), which could help explain the tumor behavior. Based on the findings, GI IMTs might show different biological potentials according to the histopathological subtype.