No abstract

Solitary fibrous tumor (SFT) is characterized by the inv12(q13q13)-derived NAB2-STAT6 fusion, which exhibits variable breakpoints and drives STAT6 nuclear expression. The implications of NAB2-STAT6 fusion variants in pathological features and clinical behavior remain to be characterized in a large cohort of SFTs. We investigated the clinicopathological correlates of this genetic hallmark and analyzed STAT6 immunoexpression in 28 intrathoracic, 37 extrathoracic, and 23 meningeal SFTs. These 88 tumors were designated as histologically nonmalignant in 75 cases and malignant in 13, including 1 dedifferentiated SFT. Eighty cases had formalin-fixed and/or fresh samples to extract assessable RNAs for RT-PCR assay, which revealed NAB2-STAT6 fusion variants comprising 12 types of junction breakpoints in 73 fusion-positive cases, with 65 (89%) falling into 3 major types. The predominant NAB2ex4-STAT6ex2 (n = 33) showed constant breakpoints at the ends of involved exons, whereas the NAB2ex6-STAT6ex16 (n = 16) and NAB2ex6-STAT6ex17 (n = 16) might exhibit variable breakpoints and incorporate NAB2 or STAT6 intronic sequence. Including 73 fusion-positive and 7 CD34-negative SFTs, STAT6 distinctively labeled 87 (99%) SFTs in nuclei, exhibited diffuse reactivity in 73, but did not decorate 98 mimics tested. In seven fusion-negative cases, 6 were STAT6-positive, suggesting rare fusion variants not covered by RT-PCR assay. Regardless of histological subtypes, intrathoracic SFTs affected older patients (P = 0.035) and tended to be larger in size (P = 0.073). Compared with other variants, NAB2ex4-STAT6ex2/4 fusions were significantly predominant in the SFTs characterised by intrathoracic location (Po 0.001), older age (P = 0.005), decreased mitoses (P = 0.0028), and multifocal or diffuse STAT6 staining (P = 0.013), but not found to correlate with disease-free survival. Conclusively, STAT6 nuclear expression was distinctive in the vast majority of SFTs, including all fusion-positive tumors, and exploitable as a robust diagnostics of CD34-negative cases. Despite the associations of NAB2-STAT6 fusion variants with several clincopathological factors, their prognostic relevance should be further validated in large-scale prospective studies of SFTs.

Endometrial stromal tumors (ESTs), including low-grade endometrial stromal sarcomas (LGESSs) and endometrial stromal nodules (ESNs) of classic histology, exhibit characteristic morphologic features and contain the nonrandom t(7;17)(p15; q21), which results in the fusion of two novel genes, JAZF1 and JJAZ1. ESTs may pose diagnostic challenges when they involve extrauterine sites, present as metastases, or display variant histologic appearances. The aim of this study was to evaluate the frequency of the JAZF1-JJAZ1 gene fusion among primary uterine, metastatic, and primary extrauterine ESTs of various histologic types and its role as a possible diagnostic adjunct. Using a nonnested reverse transcriptase-polymerase chain reaction approach, we assayed for JAZF1-JJAZ1 gene fusion transcripts in 10 cases with available fresh-frozen tissue. These included five primary uterine (two classic, one mixed smooth muscle, and one epithelioid LGESS; one classic ESN), four metastatic (two fibromyxoid, one classic, and one epithelioid LGESS), and one extrauterine (classic LGESS) tumor. The same primer set and assay conditions were used on five additional paraffin-embedded cases with adequate RNA, including three primary uterine (one fibromyxoid and one mixed smooth muscle LGESS; 1 mixed smooth muscle ESN) and two intraabdominal recurrent (two mixed smooth muscle LGESSs) ESTs. Two cellular leiomyomas and one ESS cell line (ESS-1) without the t(7;17) at the cytogenetic level were run in parallel as controls. JAZF1-JJAZ1 gene fusion transcripts were detected in five (33%) of 15 ESTs, including three of eight primary uterine, one of four metastatic, one of one extrauterine, and none of two recurrent cases. Most ESTs of classic histology showed evidence of JAZF1-JJAZ1 fusion (4 of 5 cases), whereas only one mixed smooth muscle ESN of 10 variant cases was positive. Positivity for JAZF1-JJAZ1 fusion transcripts was found in four of 10 fresh-frozen samples and in one of five paraffin-embedded ESTs. The control specimens were all negative. In conclusion, our data suggest that ESTs are genetically heterogeneous, with the prevalence of the JAZF1-JJAZ1 fusion being highest among ESTs of classic histology. Hence, the diagnostic utility of a JAZF1-JJAZ1 fusion transcript assay in ESTs may be limited to the classic histologic subset.