Although UPS is one of the most common soft tissue sarcomas, its genetic features remain poorly defined. While recurring pathogenic variants of some genes such as RB1, TP53, and ATRX are detected in UPS, they are not specific.7 The discovery of PRDM10 in soft tissue tumors in 2015 offered a new era for investigation. In that study, PRDM10 fusion was detected in three tumors which despite significant pleomorphism, has remarkably low mitotic activity (less than 1 per 10 HPFs). 4 This finding suggests that these tumors might be low-grade soft tissue tumors in contrast to pleomorphic sarcomas.
Along with the detection of PRDM10 fusion, research on PRDM10 broadened to encompass low grade soft tissue tumors. However, immunohistochemical studies haven't reached a consensus on which staining pattern (nuclear and/or cytoplasmic) is significant for the PRDM10-rearranged tumors. Puls et al. considered only nuclear staining is significant, whereas Sugita et al. showed the importance of cytoplasmic staining as well in PRDM10-rearranged tumors.5, 6 Furthermore, PRDM10 IHC has been found positive not only in low-grade malignant or borderline tumors but also in many high-grade soft tissue tumors. 8
In our study, we considered both nuclear and cytoplasmic staining as positive. 5, 6 While all IHC positive cases were highly cellular and pleomorphic, one tumor with widespread PRDM10 positivity (90% of cells stained) stood out. This case (case no 1), divergent from the other four PRDM10 positive cases, exhibited exceptionally low mitotic activity (1-2 per 50 HPFs) and no tumor necrosis. It was also the only positive case for both PRDM10 (IHC&FISH) and CD34 (IHC). (Figure 2)
The other FISH positive case (case no 2) exhibited diffuse positivity with CD34, but had PRDM10 negativity (Figure 3) and it raised the possibility of SC34FT diagnosis.
Both of these cases (case no 1 and 2), one aged 37 and the other 43, are females, and both are located in the subcutis of the lower extremity. Histopathologically, these lesions are well-delineated and highly cellular, arranged in fascicles, typically composed of epithelioid or spindle cells with abundant eosinophilic cytoplasm, oval to round nuclei and prominent nucleoli. Some cells exhibit nuclear pseudoinclusions, while the others show cytoplasmic vacuolization resembling lipoblasts. No local recurrence or distant metastasis has been detected in a follow-up period of 21 and 27 months, respectively. Considering the clinical, morphological, immunohistochemical and molecular findings, we suggest that these two cases (case no 1 and 2) initially diagnosed as UPS may actually be SC34FT.
The detection of FISH positivity in cases 7 and 8, apart from SC34FT cases, is an interesting finding. These two cases are notably distinguished by being observed in older patients, having a worse prognosis, higher mitotic activity, and being CD34 negative immunohistochemically. Until now, there have been only two PRDM10 FISH studies 6, 9, and our study is the first to demonstrate PRDM10 gene rearrangement in a tumor group distinct from SC34FTs.
The lack of PRDM10 staining in one of these tumors (case no 2) suggests a low sensitivity of the PRDM10 IHC. Moreover, the intensity and extent of staining should be considered together. Studies examining PRDM10 IHC are insufficient in determining a precise threshold for the extent of staining. In our study, the case with high-intensity, widespread (90% of cells) nuclear staining was definitely considered positive through correlation with FISH. However, in the case with high intensity but only 20% of tumor cells stained, it was concluded that a higher threshold value might be necessary. Therefore, high intensity and a high extent of staining seem meaningful; however, a more accurate threshold is needed for the extent of tumor cell staining. We suggest that low-intensity or low proportion of stained cells should be further confirmed with molecular techniques, such as FISH.
The detection of PRDM10 gene rearrangement involves the use of break-apart probes targeting the breakpoint in the gene. In our study, a threshold value of 15% of tumor cells is acceptable. 5 Various studies have so far chosen a threshold value, ranging from 10% to 20% of tumor cells. 5, 6 The lack of standardization for the threshold value could pose an issue when there are borderline values.
In conclusion, soft tissue tumors that has PRDM10 gene rearrangement, superficial location, low mitotic activity, no necrosis and CD34 positivity exhibit distinct clinical and prognostic features, suggesting potential overlap with the tumor spectrum of SC34FT. The clinical, morphological, and immunohistochemical similarity of soft tissue tumors with PRDM10 gene rearrangement to SC34FTs, along with molecular studies, led to the consideration that these tumors might represent a subtype of PRDM10-rearranged soft tissue tumors. 9-11A combined approach using IHC, FISH, RNA sequencing, and RT-PCR methods is likely to yield the most reliable results. Lastly, it should be noted that PRDM10 gene rearrangement can also be observed in high-grade tumors, unlike SC34FTs.