Sustained Response to the Mitogen-Activated Extracellular Kinase Inhibitor Trametinib in a Spindle Cell Sarcoma Harboring a QKI-RAF1 Gene Fusion.

Introduction

The advent of molecular pathology tools such as next-generation sequencing (NGS) has helped refine the classification of sarcomas and may help find new therapeutic targets.[1] Recently, a new group of soft-tissue sarcomas was described as harboring a distinct immunohistochemistry profile of S100 and CD34 coexpression and recurrent gene fusions involving RAF1, BRAF, or NTRK1/2.[2] In this paper, we report a case of recently described spindle cell tumor displaying a QKI-RAF1 gene fusion resistant to standard cytotoxic chemotherapy. The use of the mitogen-activated extracellular kinase (MEK) 1 and 2 inhibitor trametinib yielded clinical benefit and an absence of progression for 10 months.

Case Presentation

The patient, a 27-year-old woman without significant medical history, sought medical attention in May 2019 in a hospital outside Canada for a spinal cord compression syndrome with severe pain and paraparesis. A laminectomy encompassing T12-L3 vertebrae was performed with an initial improvement in symptoms. She presented again 1 month later in June with cauda equina syndrome with recurrent severe pain and sphincter dysfunction. The magnetic resonance imaging (MRI) showed conus medullaris compression by a dural lesion and she underwent a second T12-L3 laminectomy. The pathology reports from surgery showed intradural high-grade round to spindle cell sarcoma suggestive of Ewing sarcoma.

At that point, the patient transferred her care to the McGill University Health Centre (MUHC) located in Montreal, Canada. A fluorodeoxyglucose (FDG) positron emission tomography-computed tomography (PET-CT) performed in July 2019 showed a metabolically active spinal lesion at the L3-L4 level without distant metastases. In the spine MRI, the presence of an extensive intradural and extramedullary mass occupying the spinal canal completely from T12-L4 and encasing the conus medullaris was noted. Given her initial diagnosis of Ewing sarcoma, while transfer of her tumor blocks and pathologic review of the case were pending, she was treated with eight cycles of alternating vincristine-doxorubicin-cyclophosphamide and ifosfamide-etoposide plus definitive radiation therapy to the thoracic and lumbar spine. She completed this treatment in February 2020. The patient improved neurologically with a decrease in pain and the ability to walk with a walker. An FDG PET-CT done in March 2020 showed no residual metabolically active lesion, although the mass had remained stable in size. The decision to stop chemotherapy was because of the progressively worsening hematologic toxicity of the regimen and rising uncertainty about the accurateness of the diagnosis of Ewing sarcoma, given the lack of radiologic response.

At the same time, the tumor specimens were reviewed at the MUHC and another institution. Given the absence of EWSR1 and SS18 gene rearrangement, the diagnoses of Ewing sarcoma and synovial sarcoma, respectively, were ruled out. On Archer NGS performed in an outside institution, a QKI-RAF1 gene fusion was identified, and a diagnosis of the newly described spindle cell tumor defined by S100 and CD34 coexpression with recurrent gene fusion was made.

The patient re-presented in April 2020, < 3 months after the end of her treatment, with recurrent cauda equina. The FDG PET-CT and MRI showed a hypermetabolic intradural relapse at T11-L1 causing conus medullaris compression. The patient was treated with temozolomide-irinotecan as second-line chemotherapy. This treatment was poorly tolerated, and the tumor size increased on MRI after one cycle.

In June 2020, after informed consent for participation to a special-access program, the patient was prescribed single-agent trametinib, a highly selective MEK1/2 reversible antagonist. This treatment was chosen to block the oncogenic effect of the RAF1 gene fusion. After the start of trametinib, her pain was significantly lower. Furthermore, there was a sustained complete metabolic response on FDG PET-CT.

Throughout her illness, the patient had multiple antibiotic courses for recurrent urosepsis secondary to compression of the urinary tract by tumor. She ultimately died of septic shock from multidrug-resistant bacteria in May 2021. No relapse from her sarcoma was observed during the 10 months of trametinib treatment.

Consent for Publication

During her illness, the patient provided verbal consent to publication of this case report.

Discussion

Using NGS, RAF1 gene fusion was recently described in eight soft-tissue sarcomas with various fusion partners including PDZRN3, SLMAP, and TMF1.[3] The tumors were found in different organs and usually displayed low-grade features. Only one of the four cases for which follow-up is available developed distant metastases. Our case is the first RAF1 gene fusion sarcoma to be intradural and fused with QKI. Additionally, it was high grade and behaved aggressively, relapsing quickly after definitive chemotherapy and radiation therapy, but did not metastasize. Importantly, it is the first sarcoma case where an RAF1 gene fusion was successfully targeted by an MEK1/2 inhibitor, resulting in a prolonged clinical benefit and metabolic response on imaging. Successful targeting of RAF1 gene fusions by MEK1/2 inhibitors has been described in melanoma and anaplastic pleomorphic xanthoastrocytoma.[4] RAF1 indeed promotes tumor growth through its interaction with MEK1/2 and the activation of the mitogen-activated protein kinase (MAPK) pathway.

RAF1 is a recently identified oncogene. RAF1 is part of the BRAF gene family and codes for the C-RAF protein. Like BRAF, RAF1 activates the MAPK pathway via its interaction with the MEK 1 and 2 to promote tumor growth and invasion.[5] Point mutations of RAF1 in tumor are rarer than BRAF because of its lower basal activity.[6] Nevertheless, multiple oncogenic gene fusions involving RAF1 have recently been described in melanoma, acinar pancreatic carcinoma, brain tumors, and sarcoma.[7-9] Thus, unlike BRAF where somatic mutations are frequent oncogenic drivers, gene rearrangement is thought to be the main event leading to the transformation of RAF1 into an oncogene. Moreover, identification of RAF1 rearrangement is essential because there is reported response to MEK inhibitors. For example, patients with melanoma characterized by wild-type BRAF but harboring an RAF1 gene fusion responded to trametinib and cobimetinib.[10-13] There is also a report of an anaplastic pleomorphic xanthoastrocytoma harboring an ATG7-RAF1 gene fusion that responded to the MEK inhibitor cobimetinib.[4]

The fusion partner of RAF1 in our case is QKI. The product of QKI is an RNA-binding protein implicated in glial development and myelinization. Recently, it has been identified in oncogene fusions in pediatric brain tumors with MYB and RAF1 genes.[14,15] Furthermore, the QKI-RAF1 fusion gene was shown to be a driver oncogene in cell lines by activating the MAPK and phosphoinositide-3 kinase/mammalian target of rapamycin (PI3K/mTOR) pathways.[15] The use of MEK1/2 reversible antagonist trametinib slowed tumor proliferation. The addition of everolimus, an mTOR inhibitor, to trametinib suppresses tumor growth more effectively and could be considered in case of resistance to single-agent trametinib.[14] In our patient, the use of single-agent trametinib was sufficient to halt cancer proliferation.

In conclusion, we describe here the first case of a QKI-RAF1 gene fusion in a high-grade soft-tissue sarcoma of intradural location. With secondary resistance to conventional multiagent chemotherapy, the use of the MEK1/2 inhibitor trametinib induced a prolonged metabolic response and clinical benefit. Unfortunately, access to sequencing methods and targeted drugs can be challenging, particularly in rare tumors where good evidence is scarce. Our case illustrates the benefit of NGS in diagnosis accuracy and identification of a therapeutic target, resulting in effective treatment for our patient.