Uterine Inflammatory Myofibroblastic Tumors: p16 as a Surrogate for CDKN2A Deletion and Predictor of Aggressive Behavior.

ABSTRACT

Uterine inflammatory myofibroblastic tumors (IMTs) are rare mesenchymal neoplasms of uncertain malignant potential. Aside from the recently described risk stratification score, which has not been validated by other studies, and rare reports of aberrant p16 expression in malignant tumors, there are no criteria to reliably predict behavior. Herein, we evaluated the clinicopathologic features and p16 expression patterns in 30 IMTs, with genomic profiling performed in a subset (13 malignant, 3 benign). Fifteen patients had malignant IMTs, defined by extrauterine disease at diagnosis (n=5) or recurrence (n=10; median 24 mo). Patients ranged from 8 to 65 (median 51) years and tumors from 6 to 22 (median 12.5) cm. In primary tumors (n=13), infiltrative borders were noted in 10, moderate/severe cytologic atypia in 9, tumor cell necrosis in 7, and lymphovascular invasion in 6, while mitoses ranged from 0 to 21 (median 7) per 10 high-power fields. In contrast, 15 patients with benign IMTs ranged from 28 to 65 (median 44) years, with follow-up of 18 to 114 (median 41) months. Tumors ranged from 1.9 to 8.5 (median 5.5) cm, 2 demonstrated infiltrative borders, and 1 had moderate cytologic atypia. No other high-risk histologic features were observed. Application of the previously described clinicopathologic risk stratification score in all primary IMTs with complete data (n=18) classified 8 as high-risk (all malignant), 8 as intermediate-risk (3 malignant, 5 benign), and 2 as low-risk (benign). p16 was aberrant in all malignant IMTs, with <1% expression noted in 10, overexpression (>90%) in 4, and subclonal loss in 1; all benign tumors had patchy staining (20% to 80%; median 50%). Molecular analysis detected CDKN2A deletions in 8 of 9 tumors with <1% p16 expression, while the other harbored a TERT promoter mutation. TERT promoter mutations were also identified in 2 of 3 IMTs with p16 overexpression. Neither of these alterations was detected in the 3 sequenced benign IMTs. Thus, we recommend performing p16 on all uterine IMTs, which, combined with the risk stratification score, is a promising and cost-effective tool for predicting CDKN2A status and outcome in these patients. It may be particularly useful for tumors with incomplete information for risk stratification (ie, morcellated tumors) and for further stratifying intermediate-risk IMTs when sequencing is unavailable.