Gastrointestinal stromal tumor versus intra-abdominal fibromatosis of the bowel wall: a clinically important differential diagnosis.

Intra-abdominal fibromatosis (IAF) is an uncommon benign neoplasm that usually occurs in the mesentery or retroperitoneum and may, on occasion, mimic a gastrointestinal stromal tumor (GIST). Differentiating between these two entities is important clinically because IAF is a benign tumor whereas GISTs frequently have malignant potential. In this study, the authors identified 13 cases of IAF with prominent involvement of the bowel wall as well as 35 GISTs of the small intestine, colon, or mesentery and analyzed their clinical, gross, histologic, immunophenotypic, and ultrastructural characteristics to identify important distinguishing features. Patients with IAF were younger (mean, 34 yrs) than patients with GIST (mean, 54 yrs). Both types of tumors tended to be large, but GISTs were soft and lobulated with hemorrhage, necrosis, or cystification whereas IAFs were firm, tan, and homogeneous. Histologic features characteristic of GIST included the presence of spindle or epithelioid cells with variable architecture, mitotic activity (range, <1-95 mitoses/50 high-power fields [hpf]; mean, 15 mitoses/50 hpf), nuclear atypia, and myxoid or hyalinized stroma. Necrosis and hemorrhage were seen in 16 and 25 tumors, respectively. In contrast, IAFs were composed of broad, sweeping fascicles of monotonous spindle cells with mitotic activity (range, <3-11 mitoses/50 hpf; mean, 4 mitoses/50 hpf), bland nuclear features, and finely collagenous stroma. Necrosis, hemorrhage, and myxoid degeneration were not seen. Immunohistochemical studies performed on a limited number of GISTs and IAFs demonstrated that cells expressed vimentin (100% GIST and IAF), CD117 (88% GIST and 75% IAF), CD34 (42% GIST and 0% IAF), smooth muscle actin (63% GIST and 75% IAF), muscle actin (75% GIST and 75% IAF), desmin (8% GIST and 50% IAF), and S-100 protein (16% GIST and 0% IAF). Ultrastructural analysis of 21 GISTs revealed incomplete smooth muscle differentiation in some tumors whereas IAFs were shown to have complete myofibroblastic/fibroblastic differentiation. Information regarding clinical outcome was available on 29 patients and revealed that three patients with histologically benign GISTs were alive with no evidence of disease at 5 months to 6 years (mean, 3.5 yrs) and one patient with a histologically benign tumor died of disease after 7 years. Of patients with histologically malignant GIST, one died of surgical complications, 10 were alive without disease at I to 13 years (mean, 5.4 yrs), four were alive with disease at 4 months to 15 years (mean, 3.8 yrs), three had disseminated disease at operation, and seven were dead of disease at 10 months to 3 years (mean, 2.2 yrs). Follow up of eight patients with IAF demonstrated that five were alive without disease at 4 months to 15 years (mean, 5.3 yrs) and three had recurrences at 1 (two patients) and 2 years (one patient). In summary, IAFs can have many features (large size, infiltration of adjacent structures, mitotic activity) that can cause diagnostic confusion with GISTs and, importantly, the degree of mitotic activity present in IAFs may overlap that seen in malignant GISTs. These entities can be distinguished primarily by their light microscopic and ultrastructural features but there is a notable overlap in their immunohistochemical profiles. The distinction between these neoplasms is important because there are important clinical implications for the patient.