Pediatric-type Myoid Neoplasms of Somatic Soft Tissue: A Clinicopathological and Molecular Genetic Study of 78 Tumors, Highlighting Indolent Clinical Behavior and Frequent SRF Gene Rearrangements

Abstract

Soft tissue tumors with smooth muscle differentiation are rare in pediatric patients. Despite often showing morphologic features sufficient for classification as “leiomyosarcoma” in adults (eg, high cellularity and mitotic activity), clinical follow-up has shown only indolent behavior. The pathological features of recently reported SRF-rearranged “cellular myofibromas/myopericytomas,” typically occurring in children, overlap with those of true smooth muscle tumors. We studied a large series of pediatric tumors with morphologic and immunohistochemical evidence of smooth muscle differentiation, with the goals of better understanding their natural history and molecular genetic features. Seventy-eight tumors were identified in 45 males and 33 females, with a median age of 10 years. Clinical follow-up (50 patients; median, 45.5 months) disclosed local recurrence in 7 patients (15%). No metastases or deaths because of disease occurred. Group 1 (73/78) tumors consisted of cellular fascicles of mildly to at most moderately atypical, bland, ovoid to spindled cells with distinctly eosinophilic cytoplasm, appreciable mitotic activity (median, 5/50 high-power fields), and no necrosis. Group 2 tumors (5/78) showed greater cellularity, significant nuclear pleomorphism, and brisk mitotic activity (median, 59/50 high-power fields). Subsets of group 1 tumors harbored SRF rearrangements (16/47), and all group 2 tumors showed TP53 biallelic inactivation (5/5). SRF fusion partners included CITED1, NCOA2, C3orf62, RELA, ARGFXP1, ARNTI2, ICA1L, and unknown (n = 1). We conclude that the prognosis for pediatric tumors with smooth muscle differentiation that fall into group 1 is excellent. SRF rearrangements are present in a significant minority of tumors, typically showing features of smooth muscle rather than myopericytic differentiation. A smaller subset with more worrisome morphologic features harbor biallelic inactivation of TP53. To emphasize their unique features, we propose the term “pediatric-type myoid neoplasms of somatic soft tissue” rather than simply “leiomyoma” or “leiomyosarcoma” for group 1 tumors, and the designation of leiomyosarcoma in children should be limited to group 2 tumors.