Myxoid liposarcomas treated with neoadjuvant doxorubicin and ifosfamide and pre-operative radiation consistently develop extensive treatment effect that can include adipocytic maturation as demonstrated on MRI imaging and confirmed histologically at resection. In our series, the most common treatment change was extensive hyalinization (16/22, 72%) with decreased vasculature as seen in other studies. In addition, approximately one-third of our patients also developed prominent adipocytic maturation, comprising more than 50% of the treated tumor. Two tumors had extensive adipocytic differentiation giving a misleading appearance of a well-differentiated liposarcoma or even a benign lipoma. Fortunately, most tumors also harbored areas of hyalinization, characteristic of treated myxoid liposarcoma. They also lacked the fibrous bands with scattered pleomorphic cells seen in most well-differentiated liposarcomas. Molecular testing, either for rearrangements involving CHOP/DDIT3 by fluorescence in-situ hybridization assay or reverse polymerase chain reaction for the specific fusion transcripts, can also be helpful in confirming the diagnosis of myxoid liposarcoma[14, 30]. In our study, rearrangement was still identified in tumors with prominent fatty differentiation by fluorescence in-situ hybridization assay. We chose this modality since it allows one to determine the cellular component of the tissue showing the re-arrangement. Well-differentiated liposarcomas would lack this rearrangement and would instead demonstrate amplification of the 12q13-15 region.
The phenomenon of adipocytic maturation is comparable to myxoid liposarcomas in patients treated with trabectedin and the thioglitazone family of PPARγ agonists. In 2007, Grosso and colleagues described the development of “mature lipoblast-featuring cells” and “clear adipocytic differentiation” in myxoid liposarcomas treated with trabectedin. Gronchi and colleagues reported maturation to lipoblasts, decreased cellularity and vascularity in the tumors of 16 of 23 patients who received neoadjuvant trabectedin for advanced localized myxoid liposarcomas. Forni and colleagues saw similar findings in myxoid liposarcoma cell lines and two patients treated with trabectedin. Similarly, Demetri and colleagues also saw adipocytic differentiation as early as six weeks in tumors from three patients treated with troglitazone, a PPARγ agonist. These changes coincided with an increased expression in mRNA associated with adipocytic differentiation including AP2, adipsin and PPARγ. Mild increase in adipose density signaling was also seen in these tumors on MRI imaging. In our series, extensive treatment changes, predominantly hyalinization with decreased cellularity and vascularity, was seen in the tumors of our patients treated with doxorubicin and ifosfamide with pre-operative radiation therapy. However, adipocytic maturation was also seen in our tumors, histologically indistinguishable to what has been reported previously reported by others and also seen by us in myxoid liposarcoma tumors treated with trabectedin (WLW, AJL, PT). In most cases, MRI imaging correlated with evolution of treatment changes seen in the resection specimen. Decreased vascularity by contrast media enhanced imaging was seen in the majority of cases with extensive hyalinization and subsequent decrease of the thin-walled vessels in these areas. In addition, all four cases that had an increase of fat (>25%) by MRI also showed prominent adipocytic maturation on histological resection.
Myxoid liposarcoma can rarely have areas with mature-appearing adipocytes as well. Evans reported 29 cases of myxoid liposarcoma, some of which had areas that were reported to resemble mature, non-neoplastic adult adipose tissue. In a more recent study by Fritchie and colleagues, 10 of 46 (22%) myxoid liposarcomas had a mature fat appearing lipomatous component that composed 5-79% of the tumor. Full information regarding treatment was not provided in either study. Some degree of stromal hyalinization was also observed in seven cases in the Fritchie study, six of which were known to be untreated. However, this was only focal, composing only 1-9% of these tumors, in contrast to our treated cases which had extensive hyalinization. It is possible that some changes could have been present before treatment and not sampled in the pre-treatment biopsy, emphasizing the importance of comparing pre- and post-treatment imaging studies. One case with histological evidence of adipocytic maturation revealed no change in adipocytic content by MRI and adipocytic maturation was noted in the pre-treatment biopsy. However, in our experience (WLW, AJL) extensive fatty maturation and hyalinization are uncommon in untreated myxoid liposarcomas. The remaining pre-treatment biopsies lacked these changes and developed increased fat on MRI studies between pre- and post treatment tumors, which corresponded histologically to increased fatty maturation, increased hyalinization and decrease vascularity on the resection specimen.
The significance of adipocytic maturation is unclear. No patients with extensive adipocytic maturation died of disease, but patients did develop metastases regardless of adipocytic maturation. However, our series is small and additional larger studies are necessary. The mechanism of fatty maturation when exposed to doxorubicin and ifosfamide and radiation therapy is also not known. Doxorubicin is an anthracycline and proposed to have many mechanisms of actions among which include intercalcating DNA, formation of free radicals, inhibition of topoisomerase II, prohibiting DNA and RNA unwinding, and induction of cell death. Ifosfamide is a DNA alkylating agent which prohibits DNA replication and results in cell death. Interestingly, doxorubicin at low concentrations has been reported to induce differentiation in leukemia and breast carcinoma cell lines[34, 36, 37]. In breast cell lines, differentiation corresponded to the reduction in c-myc expression and growth arrest in response to doxorubicin induced DNA damage. To our knowledge, ifosfamide has not been reported to induce differentiation. Perhaps a similar non-specific response to DNA damage with doxorubicin can occur in myxoid liposarcomas. Other sarcomas have also been reported to differentiate in response to chemotherapy and radiotherapy including rhabdomyosarcoma. Our study would support the notion that these changes can be a general response to effective treatment in myxoid liposarcoma and may not necessarily be specific to a treatment modality such as trabectedin. In this context, we should be cautious in interpreting potential insights into the unknown mechanism of action of trabectedin from this phenomenon.
The fusion transcript characteristic in the vast majority of myxoid liposarcomas is composed of FUS, which encodes for a RNA/DNA binding domain, and CHOP/DDIT3, which encodes for a transcription factor which dimerizes with CAAT/enhancer binding proteins (C/EBP) and negatively regulates genes involved in adipocytic differentiation including PPARγ. Multiple fusion variants and types exist; however, none have been associated with increased adipogenesis. Trabectedin binds to the minor groove of DNA and is theorized to sterically hinder the binding of the FUS-CHOP/DDIT3 fusion protein (probably mediated primarily by the CHOP/DDIT3-donated DNA binding domain) and thus allow adipocytic differentiation to continue by relieving repression. It could also effect the post transcriptional modification of the fusion transcript. In addition, other mutations are known to exist including activation of the ERK/MAPK, PI-3-kinase/Akt pathways, alterations of p53, IGF1R and IGF2; some of these have been associated with round cell change and poor prognosis.[16, 22] None are known to correlate with adipocytic differentiation.
In summary, we report the histological and radiological findings seen in myxoid liposarcomas in patients treated with doxorubicin and ifosfamide and pre-operative radiation therapy. The vast majority of these tumors demonstrated significant treatment effect in the form of hyalinization; however, adipocytic maturation, once thought to be exclusive to tumors treated with trabectedin and PPARγ agonists, were also seen in some of our cases. Pathologists and radiologists should be aware of this as it may cause diagnostic confusion if prior history is not available and in comparing histological treatment changes between trials.