Inflammatory Myofibroblastic Tumor of the Uterus: An Immunohistochemical Study of 23 Cases. (November 2020)

• Record Type: Journal Article
• Authors: Bennett, Jennifer A.; Croce, Sabrina; Pesci, Anna; Niu, Nifang; Van de Vijver, Koen; Burks, Eric J.; Burandt, Eike; Zannoni, Gian Franco; Rabban, Joseph T.; Oliva, Esther
• Is Part Of: American journal of surgical pathology. Volume 44:Number 11(2020)
• Publication Date: 2020-11
• Subjects: inflammatory myofibroblastic tumor -- uterus -- p16 -- BCOR -- IFITM1 -- transgelin -- CDKN2A; Pathology, Surgical -- Periodicals; 617.0705
• DOI: 10.1097/PAS.0000000000001525
• Languages: English
• ISSNs: 0147-5185

Abstract

Abstract : Inflammatory myofibroblastic tumors (IMT) of the uterus may be underrecognized as their morphology and immunophenotype may overlap with myxoid variants of uterine smooth muscle tumors and endometrial stromal tumors. Although ALK is a helpful biomarker, not all uterine IMTs are ALK -rearranged, and a small subset of myxoid leiomyosarcomas is ALK-positive. Herein, we evaluated a series of 23 IMTs for the novel endometrial stromal markers interferon-inducible transmembrane protein-1 (IFITM1) and BCOR, the novel myoid marker transgelin, and possible predictive markers p16 and p53 by immunohistochemistry to determine their expression profile and potential prognostic value. Patients' ages ranged from 8 to 59 (mean 39) years and tumors from 2 to 20 (mean 8.2) cm. Follow-up was available for 12/23 (52%) patients; 9/12 (75%) without evidence of disease, 2/12 (17%) alive with disease, and 1/12 (8%) dead from disease. Four IMTs were classified as malignant due to extrauterine disease at diagnosis and/or recurrence. IFITM1 was positive (combined score>2) in 19/23 (83%), BCOR in 8/20 (40%), and transgelin in 22/23 (96%) of tumors. IFITM1 and BCOR were more often expressed in the myxoid component, and transgelin in the compact areas. p16 expression was absent in 5/23 (22%) of IMTs, while p53 was wildtype in all tumors. p16-negative IMTs included all 4 classified as malignant and one where the patient was lost to follow-up. Molecular data were available in 2 malignant IMTs, both of which harbored CDKN2A deletions. We conclude that caution is advised when using IFITM1, BCOR, and transgelin as markers for endometrial and smooth muscle tumors, as these are commonly expressed in IMTs. However, we did identify an association among lack of p16 staining, CKDN2A deletions, and aggressive behavior that merits corroboration by other studies. As a result of this finding, we recommend the use of p16 in the diagnostic work-up of uterine IMTs due to its potential prognostic significance.