First year of two-year project for correlative research to define novel biomarkers of response in desmoid tumors treated with sorafenib.

Desmoid tumors (DT/DF) are clonal connective tissue malignancies of fibroblastic origin. DT/DF lack metastatic potential however, can cause significant morbidity, loss of function and pain through mass effects. Mortality from DT/DF results from local infiltration of vital structures. There is no standard of care for the treatment of DT/DF. Surgery is curative however associated with high recurrences. In unresectable patients, systemic therapies range from anti-estrogens to cytotoxic chemotherapies. DT/DF over express c-KIT and PDGFR and a Phase II study of imatinib showed a partial response (PR) of 6%. Sorafenib is a multi-target kinase inhibitor of b-Raf, PDGFR and VEGFR and available through an expanded access program. We treated an index patient with recurrent, unresectable DT/DF with sorafenib. A PR and clinical benefit prompted us to evaluate our experience of sorafenib in 26 patients. We noted a 25% PR by RECIST 1.1 and improvement of symptoms in 70% of patients. We noted a 30% decrease in the MRI T2 signal in 90% of patients with extra-abdominal disease. We hypothesize that sorafenib will lead to improved response rates, symptoms and progression free survival (PFS) in DT/DF patients. This is a multicenter, Phase III, randomized, double-blind, placebo-controlled trial of sorafenib in DT/DF patients. The primary endpoint is to evaluate overall response rates (ORR) for patients treated with sorafenib versus those treated with placebo. This study will enroll 126 patients over 2 years and will have a 90% power to detect a 22% increase in response rates (1-sided a=0.05). The secondary endpoints are 1) PFS and 2) pain palliation by the Brief Pain Inventory scale (BPI). Patients on placebo will cross over to sorafenib. Explorative studies include pre- and post-treatment biopsies in responding patients to determine biomarkers of response and mechanism of action of sorafenib. Decrease in MRI T2 signal may be an early marker of RECIST response and may be a more sensitive imaging marker for clinical benefit than RECIST response. We will conduct an explorative study a novel imaging biomarker by correlating changes in T2 signal to RECIST response and clinical benefit by pain palliation.

A major focus of my research program is to study the molecular pathology of desmoid tumors (DTs), with a long-term goal of developing improved treatments. DTs are locally invasive soft tissue tumors, which we found are caused by mutations resulting in the stabilization of the protein, ß-catenin. Stabilized ß-catenin binds to TCF transcription factors to regulate the expression of genes in a cell type specific manner. We developed mouse models of DT based on this knowledge and used these to study the role of target genes in the DT phenotype. Our ultimate goal is to develop an improved approach to DT therapy.
In this project we will continue our ongoing work funded by the DTRF in which we identified pharmacologic agents that target DT cell viability by screening compound libraries composed of agents which have a high potential to be rapidly translated into patient care. We identified agents that decreased cell viability in DT cultures but not in normal fibroblasts. Over the past two years, we tested several compounds in multiple DT cell lines, tested one in a mouse model of DT, and this work is in press in PLoS One, and found three others that have strong potential to be used as a therapeutic approach to DTs.
Two agents target platelet derived growth factor (PDGF) signaling, a pathway we also found can modulate ß-catenin activity. The other targets Notch signaling. We also developed a new mouse model in which we can regulate expression of conditional stabilized ß-catenin alleles, which can be used to rapidly screen signaling pathway interactions in-vivo. We hypothesize that compound screening will identify agents that can be developed into novel therapies to treat DTs, and rapidly be brought to patient care. In this next phase of our ongoing work, we propose to test this hypothesis by answering the following two questions:
How does Platelet Derived Growth Factor (PDGF) and Notch signaling modulation target DT cell viability? In the initial phase of our DTRF work, we identified three compounds that decreased cell viability in DT cultures but not normal fibroblasts: suntinib and mastinib, both of which block PDGF signaling; and DAPT, a gamma-secretase inhibitor which blocks canonical notch signalling. In our preliminary data, we found that PDGF signaling regulates ß-catenin activity. Cells in culture will be tested for how they alter DT cell behavior, by studying proliferation rate, apoptosis rate, and ß-catenin protein level. A new mouse model we developed will be used to test the interaction of signaling pathways identified in the compound screen.
Can PDGF and Notch pharmacologic blockade be used to treat DTs? Suntinib, Mastinib, and DAPT will be tested using a genetically modified mouse that develops DTs and using xenografted human DT tumors in immunodeficient mice. The mechanistic data generated in aim one will also be verified in-vivo in this aim.
The role of these agents in modulating different cell subpopulations in DTs will also be analyzed. In our previous work, we found that DTs contain a small cell population, which excludes Hoechst dye (called side population, or SP, cells). The SP cells have an enhanced ability to form tumors when implanted into immunodeficient mice, and as such act as tumor initiating cells (TICs), also called cancer stem cells. TICs may be resistant to anti-neoplastic drugs, and as such SP cells may be responsible for the resistance to treatments, and for tumor recurrence. We will also test theses compounds for differential effects on SP and non-SP cell populations.
This work will identify compounds that can rapidly be developed as a novel DT treatment. Such agents will ultimately be translated to clinical trials, resulting in an improved pharmacologic approach to DT management. In addition, the various agents likely target different aspects of controlling DT viability (e.g. proliferation -vs- apoptosis), and as such, this knowledge could be used to develop a rationale multi-drug approach to DT treatment.

Desmoid tumors (DTs) are a locally aggressive tumor type that can cause remarkable morbidity and even mortality in afflicted patients. While desmoid clinical phenotype has been well characterized, the underlying tumor molecular biology is poorly understood. In studies conducted within the premise of our previously funded DTRF seed grant we were able to establish a large desmoid tissue and cell strain repository and have confirmed that β-catenin mutations are highly prevalent in sporadic desmoid tumors. Furthermore, we showed that a specific mutation, 45F possibly significantly correlates with dismal outcome for patients with primary lesions. This initial observation merits further confirmation and in the recent grant period, through an international collaboration, we have been able to assemble a large cohort (>160) of clinically annotated primary DT samples; all were sequenced for β-catenin. Interim results are promising, confirming our initial observation.
We have also evaluated the effects of agents commonly used for the treatment of DT such as tamoxifen, celecoxib, and Gleevec on the growth of a large panel of desmoids primary cultures. Sensitive and resistant cell strains were identified; therapeutic response was independent of β-catenin mutation status or type or the expression level of the presumed drug targets. Finally, we have identified a potential DT gene expression signature enriched for genes regulated by TCF/LEF highlighting the role of β-catenin deregulation in these tumors. Midkine is one of the DT overexpressed targets; our studies further identified a correlation between midkine expression level and primary DT recurrence potential and a role for this protein in DT cell migration and invasion.
In the current, third year of the project, we aim to:
1. Finalize our studies evaluating the role of β-catenin mutations as DT molecular prognosticators.
2. Identify molecular signatures corresponding with sensitivity/resistance to commonly utilized anti-DT systemic agents.
3. Determine the role of the notch pathway in DT and its utility as a node of DT therapeutic vulnerability.
4. Unravel additional gene mutations operative in DT through whole exome sequencing of human samples (including matched specimens). We hope that these studies will result in meaningful observations that will enhance the management of patients burdened by this devastating disease.