Inflammatory myofibroblastic tumor: comparison of clinicopathologic, histologic, and immunohistochemical features including ALK expression in atypical and aggressive cases.

Inflammatory myofibroblastic tumor (IMT) is a neoplasm of intermediate biologic potential. In this study, we report a subset of IMTs with histologic atypia and/or clinical aggressiveness that were analyzed for clinicopathologic features, outcome, and immunohistochemical expression of anaplastic lymphoma kinase (ALK) and other markers to identify potential pathologic prognostic features. Fifty-nine IMTs with classic morphology (5 cases), atypical histologic features (21 cases), local recurrence (27 cases), and/or metastasis (6 cases) were studied. Immunohistochemistry was performed for ALK1 and other markers (Mib-1, c-Myc, cyclin D1, caspase 3, Bcl-2, Mcl-1, survivin, p27, CD56, p53, MDM-2) using standard techniques. The 59 IMTs had an age at diagnosis ranging from 3 weeks to 74 years (mean 13.2 y, median 11 y, 44% in the first decade). The mean tumor size was 7.8 cm. Sites included the abdomen or pelvis in 64%, lung in 22%, head and neck in 8%, and extremities in 5%. The follow-up ranged from 3 months to 11 years, with a mean of 3.6 years and a median of 3 years. Thirty-three patients had local recurrences, including 13 with multiple local recurrences and 6 patients with both local recurrences and distant metastases. Six patients died of disease, 5 with local recurrences, and 1 with distant metastases. Histologic evolution to a more pleomorphic cellular, spindled, polygonal, or round cell morphologic pattern was observed in 7 cases. Abdominal and pelvic IMTs had a recurrence rate of 85%. Recurrent and metastatic IMTs were larger, with mean diameters of 8.7 and 11 cm, respectively. Cytoplasmic ALK reactivity was seen in 56%. ALK-negative IMTs occurred in older patients (mean age 20.1) years and had greater nuclear pleomorphism, atypia, and atypical mitoses. All 6 metastatic IMTs were ALK-negative. Nuclear expression of p53 was detected in 80% of IMTs overall, but in only 25% of the metastatic subset. There were no significant differences among the subgroups for c-Myc, cyclin D1, MDM-2, Mcl-1, Bcl-2, CD56, p27, caspase 3, or survivin expression. In conclusion, among these 59 IMTs, ALK reactivity was associated with local recurrence, but not distant metastasis, which was confined to ALK-negative lesions. Absent ALK expression was associated with a higher age overall, subtle histologic differences, and death from disease or distant metastases (in a younger subset). Other proliferative, apoptotic, and prognostic markers did not correlate well with morphology or outcome. Thus, ALK reactivity may be a favorable prognostic indicator in IMT and abdominopelvic IMTs recur more frequently.