Search

The clinical evolution of solitary fibrous tumors (SFTs) is often uncertain and several risk stratification systems (RSS) have been proposed. The Demicco et al. RSS is the most frequently implemented. In this study we aim to validate two alternative RSS (Sugita et al. and G-Score) using results for the Demicco RSS from a previous study of 97 SFTs. In addition, we aim to determine whether reclassified cases had any distinctive molecular features. As the Sugita et al. system substitutes mitotic count with Ki-67 index we also investigated whether Ki-67 results for tissue microarrays are comparable to those obtained using whole tissue sections. In the present study we detected that many cases classified by Demicco RSS as low-risk were reclassified as intermediate risk using the new system (G-score RSS). Kaplan-Meier survival plots for G-Score RSS showed that the low-risk and intermediate-risk SFTs had a similar evolution that contrasted with the more aggressive high-risk group. Moreover, the similar evolution in both low and intermediate-risk groups occurred despite the G-score system being stricter in classifying low-risk tumors. We observed that Sugita RSS does not provide any better risk stratification in comparison with the Demicco RSS, and testing both RSS in our series produced similar Kaplan-Meier survival data. We found some discordant results when comparing whole sections and the corresponding tissue microarrays samples, finding the hotspot areas easier to locate in whole sections. Forty-one SFTs with initial low-risk assigned by the Demicco RSS were reclassified as intermediate-risk by G-score finding both TP53 and HTER mutations in four cases, only HTER mutation in 11 cases, and only TP53 mutation in 2 cases. All six cases of SFT classified as high-risk by both the Demicco and G-score RSS suffered recurrence/metastasis, and half showed both TP53 and HTER mutations. Five SFTs were categorized as low-risk by both Demicco and G-score, of which 4 cases revealed HTER mutation. Regarding the outcome of these 5 patients, two were lost to follow-up, and one of the remaining three patients suffered recurrence. We believe that although the presence of both TP53 and HTER mutations may confer or be related to poor evolution, the isolated presence of HTER mutation alone would not necessarily be related to poor outcome. The G-score RSS more accurately identified low-risk patients than the other two risk models evaluated in the present series. Late recurrence/metastasis may occasionally be observed even in low-risk SFTs categorized by stricter classification systems such as the G-score RSS. These findings support the possibility that additional, as yet unknown factors may influence the clinical evolution of SFTs. In conclusion, given the possibility of late recurrence, long-term follow-up is recommended for all SFT patients, even in cases classified as low risk by the stricter G-score system. An integration of clinical, radiological, pathological, and molecular findings may improve SFT risk stratification and better predict patient outcome.

Background Despite extensive immunohistochemical (IHC) and molecular studies combined with morphologic findings, a group of round/ovoid cell tumors histologically similar to Ewing sarcomas (ES) but lacking EWSR1-rearrangements may remain unclassifiable. Design We retrospectively analyzed 41 Ewing-like tumors (formalin-fixed, paraffin-embedded) previously determined as negative or non-informative for EWSR1-rearrangements by FISH and/or RT-PCR. A new histopathology revision and additional IHC and molecular analyses were carried out in order to investigate whether additional IHC and/or molecular testing in combination with the morphological findings may help in reaching a definitive diagnosis. Results Almost all the tumors (n = 40) involved soft tissue and/or bone and half the patients died of disease. In the archival cases all diagnoses were Ewing sarcoma (ES), Ewing-like sarcoma (ELS), myoepithelial tumor and undifferentiated sarcoma (US). In the new review all the tumors were re-classified as, ES (n = 16), Ewing-like tumor with EWSR1 rearrangement and amplification and possible EWSR1-NFATC2 gene fusion (n = 1), CIC-rearranged sarcomas or undifferentiated sarcoma, most consistent with CIC-rearranged sarcoma (n = 7), sarcoma with BCOR-alteration or undifferentiated sarcoma, consistent with BCOR-associated sarcoma (n = 3), neuroblastoma (n = 2), unclassifiable neoplasm with neuroblastic differentiation (n = 1), malignant rhabdoid tumor (n = 2), lymphoblastic lymphoma (n = 1), clear cell sarcoma of the gastrointestinal tract (n = 1), small cell carcinoma (n = 1), sclerosing rhabdomyosarcoma (n = 1), desmoplastic small round cell tumor (n = 1), malignant peripheral sheath nerve tumor (n = 1), poorly-differentiated synovial sarcoma (n = 1), Possible gastrointestinal stromal tumor/GIST with predominant round cells (n = 1) and possible SMARCA4-deficient-sarcoma (n = 1). NKX2.2, ETV4 and BCOR immunoreactivity was observed in all ES, CIC-rearranged sarcomas and sarcomas with BCOR alteration, respectively. CIC-rearrangement by FISH was observed in many of the CIC-rearranged sarcomas. Conclusion Our analysis of 41 Ewing-like tumors confirms that there may be a significant pathological and IHC overlap among Ewing-like tumors, with prognostic and therapeutic impacts. Additional IHC (NKX2.2, ETV4 and BCOR) and molecular studies including FUS, CIC or BCOR analysis may support the final diagnosis when FISH or RT-PCR fail to detect EWSR1-rearrangements. Any molecular findings should always be interpreted in relation to the specific clinical and pathological context.