De‐differentiated liposarcomas with solitary fibrous tumour‐like pattern and STAT6 nuclear expression: an important diagnostic pitfall

De-differentiated liposarcoma is an important mimicker of several soft tissue sarcomas, and its diagnosis should always be ruled out in case of large retroperitoneal masses. MDM2 amplification is key for diagnosis, and can be easily assessed by fluorescence in situ hybridisation (FISH).1 Solitary fibrous tumour (SFT) is a mesenchymal neoplasm that can arise in any anatomical site. Diagnosis is challenging, and this tumour is often referred to as having a ‘patternless pattern’. Common histological features include haphazard arrangement of spindled to ovoid cells within a collagenous stroma, variable cellularity and admixed branching or hyalinised vessels. Although most cases show good prognosis, 10–30% of cases behave aggressively. Infrequently, some cases show an abrupt transformation into non-distinctive high-grade sarcoma (also called de-differentiation).2 While STAT6 immunohistochemistry (specifically strong and diffuse nuclear expression) is a good surrogate of the typical NAB2::STAT6 fusion, it is not 100% specific for its diagnosis. An SFT-like pattern has been described in de-differentiated liposarcoma,3 which may lead to requesting STAT6 immunohistochemistry. However, STAT6 nuclear positivity has been demonstrated in a subset of de-differentiated liposarcomas,4-6 constituting an important (and still under-recognised) pitfall. Herein we highlight this diagnostic challenge by reporting two cases erroneously diagnosed as malignant SFT based on morphology and immunohistochemistry (including diffuse positivity for CD34 and STAT6), but later shown to display MDM2 amplification, qualifying as de-differentiated liposarcomas. We also briefly review the molecular mechanisms that explain the up-regulation of STAT6 in these tumours. The first patient (68-year-old female, Figure 1) presented with an 8-month history of abdominal discomfort, and imaging revealed a 37-cm right retroperitoneal mass encasing the kidney, adrenal gland and pancreas. Biopsy of the mass showed a high-grade spindle cell proliferation with non-specific features, focal necrosis and high mitotic count (four mitoses/HPF). There were foci of transitions to more bland areas suggestive of SFT, with thin-branched vessels (highlighted by CD34 and ERG) and abundant interstitial collagen deposition. Both areas showed strong nuclear expression of STAT6. Expression of SMA was focal, and calretinin, S100, HMB45, DOG1 and desmin were negative. While diagnoses such as ‘unusual’ de-differentiated liposarcoma and myopericytic neoplasms were considered, a final diagnosis of malignant SFT was favoured. The patient underwent resection, which showed a 37 × 22 × 15-cm lobulated and well-demarcated mass. On sectioning the lesion was multinodular, with tan firm areas and others more greyish and hard, with calcifications. The kidney and adrenal gland were surrounded, but uninvolved by the tumour. Histology recapitulated the aspects seen in the biopsy, with SFT-like areas showing transitions to areas of necrosis and of non-descriptive sarcoma. Areas of well-differentiated liposarcoma were observed adjacent to the kidney parenchyma (which was, however, uninvolved). FISH revealed MDM2 amplification (MDM2/CEN12 ratio of 10; cut-off ≤ 2). The patient had a local recurrence in the tumour bed 5 years later, but is alive with no evidence of disease at 8 years of follow-up. The second patient (76-year-old female, Figure 2) presented with an enlarged abdomen, with imaging revealing a 34-cm right retroperitoneal mass suspected to represent a renal cell carcinoma. Resection showed a 34 × 17.5 × 13.5-cm mass encasing and infiltrating the upper pole of the kidney and the adrenal gland. The mass was overall tan, but on sectioning dominantly showed whorled whitish/greyish nodules. Necrosis was evident. The tumour showed SFT-like areas and transition to high-grade sarcomatous areas with atypical/bizarre cells and tumour necrosis. It was diffusely positive for CD34 and STAT6, focally positive for SMA and negative for desmin, caldesmon, pancytokeratins, HMB45, CD99, S100 and CD117. No areas of well-differentiated liposarcoma were found. At this time, a diagnosis of malignant SFT was favoured. However, FISH revealed MDM2 amplification (MDM2/CEN12 ratio of 3.6, with one-third of nuclei with ratio > 10). The patient was lost to follow-up after 1 year. Both cases showed a dominant morphology and immunophenotype suggestive of SFT and histological characteristics indicating high-grade sarcomatous transformation. In such cases, distinction of SFT from de-differentiated liposarcoma can be challenging. Extensive sampling looking for well-differentiated liposarcoma areas (found in one of our cases) and detection of amplification of the 12q13-15 chromosomal region are critical [including MDM2 (12q15) and CDK4 (12q14.1)].3, 4 Studies have shown STAT6 immunoexpression in up to 11–14% of de-differentiated liposarcomas because of co-amplification of its encoding gene (not fusion-driven), given its position at 12q13, still part of the critical 12q13-15 amplified chromosomal region.4, 5 The pattern of STAT6 expression could be informative, as studies on de-differentiated liposarcomas report weak nuclear staining,4, 5 or strong nuclear staining accompanied by weaker cytoplasmic expression,6 a pattern recapitulated by our two cases. Conversely, weak or absent STAT6 expression has also been reported in SFT with de-differentiation. To conclude, de-differentiated liposarcoma can assume a wide array of morphologies and should always be ruled out in the setting of retroperitoneal masses. FISH for MDM2 should always be performed when facing the differential diagnosis between SFT and liposarcoma, especially when areas of more conventional liposarcoma are not present. Writing—drafting manuscript: JL. Methodology: JL, MM, BZS, SH. Writing—critical revision of draft: CMG. All authors read and approved the final manuscript. None. The authors declare that they have no conflicts of interest. All procedures were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. All data are provided in the case description. Data S1: See Supplementary Methods for immunohistochemistry details. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.