Although separating
gastrointestinal stromal tumor (GIST) from mesenteric
fibromatosis and
sclerosing mesenteritis is clinically important, this distinction sometimes poses problems for practicing pathologists. In the
STI571 (
Gleevec,
Imatinib) era, the problem may be further compounded when protocol-driven staining for CD117 (c-kit) is performed on spindle cell proliferations presenting in the bowel wall and mesentery using an antibody known to react with the majority of mesenteric
fibromatoses when other
antibodies are more specific. Because most mesenteric
fibromatoses have mutations in the pathway and hence have abnormal nuclear accumulation of
beta-catenin protein, we studied
beta-catenin expression among a panel of other immunohistochemical stains to distinguish mesenteric
fibromatosis, GIST, and
sclerosing mesenteritis. Examples of
gastrointestinal stromal tumors (GIST, 11),
sclerosing mesenteritis (5), and mesenteric
fibromatosis (10) were retrieved from the archives of our institutions. Cases were studied with an immunohistochemical panel consisting of CD117,
beta-catenin, CD34, smooth muscle actin,
desmin,
keratin, and
S-100 protein. Cases were scored as "negative," "focally positive," or "diffusely positive." In evaluating
beta-catenin, nuclear accumulation was required. GIST all had CD117 (11 of 11, diffuse) and CD34 (11 of 11, diffuse) with variable actin (5 of 11, focal) and negative
desmin,
keratin,
S-100 protein. All GIST lacked
beta-catenin (0 of 11). Mesenteric
fibromatosis had CD117 (6 of 10, 3 focal, 3 diffuse), typically expressed more weakly than in GIST, actin (5 of 9, focal), and
desmin (3 of 8, focal) in keeping with myofibroblastic differentiation but lacked CD34, S-100, and
keratin. CD117 staining was not eliminated by use of a non-
avidin-
biotin technique. Nuclear
beta-catenin was detected in 9 of 10
fibromatoses, including one case associated with
familial adenomatous polyposis. Two of five
sclerosing mesenteritis cases focally expressed CD117. None of the
sclerosing mesenteritis cases had nuclear
beta-catenin.
Sclerosing mesenteritis cases were otherwise fibroblastic and myofibroblastic with focal actin in 5 of 5 and negative
desmin,
keratin, and
S-100 protein but one had CD34 (1 of 5, focal). With increasing protocol-driven interest in evaluating bowel wall and mesenteric spindle cell lesions using CD117 (c-kit)
antibodies, it is important for practicing pathologists to be aware that lesions other than GISTs are likely to express this
antigen using certain
antibodies.
beta-Catenin staining identifies lesions that are, instead, mesenteric
fibromatoses.