By Keila Torres, MD, PhD; Quansheng Zhu, MD, PhD and Dina Lev, MD 

Abstract

Unclassified pleomorphic sarcoma / malignant fibrous histiocytoma (UPS/MFH) represent a major unresolved clinical problem for which novel and markedly more effective therapeutic approaches are crucially needed. The molecular diversity and intricacy of these genetically complex malignancies as well as the lack of appropriate bioresources such as human specimens, cell lines, and animal models have impeded incisive insights and progress. Molecular targeted therapy has recently emerged as an effective anti-cancer treatment approach. Given the heterogeneity of UPS/MFH, identifying common molecular deregulations amenable to therapeutic targeting would be of major importance. Several lines of evidence suggest a potential role for PI3K/AKT/mTOR deregulation in UPS/MFH progression, thereby rendering this pathway an attractive potential therapeutic target. Under the aegis of this project we aimed to further elucidate the functional contribution of PI3K/AKT/mTOR signaling to UPS/MFH biology and to test the anti-tumor effects of therapies targeting this pathway in the pre-clinical setting.

An Introduction to Malignant Fibrous Histiocytoma / Unclassified Pleomorphic Sarcoma and the PI3K/AKT pathway

Malignant fibrous histiocytoma (MFH) was coined in the 1960s by Kauffman and Stout as a term describing a group of lesions presumably derived from mixed histiocytic/fibroblastic lineage.¹ This diagnostic category has been questioned and in the most recent World Health Organization² classification system, it is now termed undifferentiated pleomorphic sarcoma (UPS). UPS/MFH diagnosis is currently reserved for a group of pleomorphic sarcomas which, by current technology, show no definable line of differentiation.³ Several morphologic, cytogenetic, immunohistochemical, and genome-wide expression profiling studies suggest that UPS/MFH is not a distinct entity, but instead, is a cadre of heterogeneous sarcomas that share gene expression patterns with other high-grade complex karyotype sarcomas.⁴ UPS/MFH local control utilizes surgery combined with radiotherapy. However, 30-50% local recurrence rates (depending on tumor site) remain problematic and are particularly ominous in loci where salvage radical surgery or re-irradiation may not be feasible.⁵ UPS/MFH dissemination to the lungs is controlled in about 35% of patients, account for 80% of disease-specific deaths, and is a major determinant of the five-year 50% overall survival rate in UPS/MFH, a percentage that has been stagnant for decades.^6-8 Even UPS/MFH initially responsive to chemotherapy frequently develops chemoresistance during therapy or upon recurrence, pointing to the crucial need for novel effective therapeutic approaches.

In light of UPS/MFH genetic complexity, identifying shared molecular aberrations amenable to therapeutic targeting is critical. Deregulated PI3K/AKT/mTOR signaling has been implicated in tumor progression and metastasis in multiple epithelial origin malignancies, including: brain, prostate, breast, lung, liver, gastric, colon, ovarian, and endometrial cancers.⁹ While not extensively explored, evidence suggests AKT pathway involvement in complex karyotype soft tissue sarcoma (STS) development and progression. Hernando et al reported increased expression of activated AKT in a large panel of human leiomyosarcoma, MFH, and dedifferentiated liposarcoma¹⁰; using a conditional PTEN knockout mouse model, they demonstrated a critical role for the AKT pathway in smooth muscle transformation and leiomyosarcoma development. Tomita et al identified a correlation between phospho-AKT (pAKT) expression in human STS specimens, subsequent tumor recurrence, and patient survival.¹¹ Recently, we showed that pAKT is highly expressed in a cohort of human diverse histological subtype STS (although UPS/MFH preclinical models were not included in this initial report), where it serves as a convergence point for many upstream deregulations.¹² Furthermore, we have demonstrated that AKT inhibition significantly impacts STS cell proliferation, cell cycle inhibition and apoptosis in vitro as well as STS growth in vivo. These initial insights highlight the potential importance of the PI3K/AKT pathway in genetically complex STS and formed the basis for the hypothesis (described below) underlying the objectives of the current study.

Purpose and Aims

Based on encouraging preliminary insights we hypothesized that AKT activation is a potential common deregulation in the heterogeneous UPS/MFH tumor cohort -- driving tumor growth and metastasis. Tumor progression is a complex multistep process requiring tumor cell proliferation, survival, migration and invasion. To that end, we proposed to evaluate the impact of PI3K/AKT/mTOR blockade on these biological properties of UPS/MFH. Two aims were set:

• Aim 1: To determine the effect of PI3K/AKT/mTOR inhibition on UPS/MFH cell growth, survival, migration, and invasion.
• Aim 2: To examine the effect of PI3K/AKT/mTOR inhibition on MFH/UPS growth in vivo.

Results

One of the major challenges hampering comprehensive "bed-side to bench" translational UPS/MFH research is the scarcity of bioresources/experimental models. Consequently, much effort was devoted into assembling needed resources, including a UPS/MFH TMA, human and murine cell strains, and mouse models (see information highlighted below).

Immunohistochemical staining of our UPS/MFH TMA demonstrated that pAKT is commonly expressed in these malignancies.¹³ Most importantly, univariable and multivariable analyses identified pAKT expression level as prognostic for unfavorable disease specific survival (DSS) independent of other clinical, pathological (i.e. grade), and molecular factors (i.e. Ki67).¹⁴

Western blot analysis confirmed increased pAKT expression in human UPS/MFH cell strains as compared to normal human fibroblasts (NHF; Figure 1A). Similarly, increased pAKT expression levels and activation of the AKT downstream mTOR signaling pathway were noted in the murine UPS/MFH cell lines (Figure 1B). Of potential relevance, high throughput Sequenom-based genotyping identified point mutations in N-Ras (Q61R) and PIK3CA (E542K) in the UPS060 cell strain and a PIK3R1 point mutation (M326I) in the UPS485 cell strain. Sanger sequencing further confirmed the presence of the above mutations in the UPS060 cell line and original tumor (Figure 1C).

PI-103, a novel dual PI3K/mTOR small molecule inhibitor,¹⁵ was used to determine the impact of the AKT/mTOR pathway on UPS/MFH tumorigenic properties. Western blot analyses confirmed a PI-103 dose-dependent inhibition of AKT phosphorylation and mTOR activation (evaluated via the phosphorylation status of mTOR downstream effectors 70S6 kinase and 4EBP1) (Figure 2A). MTS assays demonstrated a marked PI-103 dose and time dependent-induced decrease in UPS/MFH growth (Figure 2B). PI staining/FACS analysis identified G1 cell cycle blockade in UPS/MFH cells treated with PI-103 (Figure 2C). However, no significant sub-G1 fraction was noted in response to treatment. Finally, PI-103 significantly inhibited UPS/MFH cell migration and invasion (Figure 2D).

Finally, the effect of PI-103 on the growth of human UPS/MFH xenografts (UPS060) in SCID mice was evaluated. Three PI-103 doses were selected for testing (20, 30 and 50mg/kg) based on previous reports.^{16, 17} The experiment was terminated when control tumors reached ~10mm at size due to tumor ulcerations. A statistically significant (p<. 0.01) delay in tumor growth was noted with all three PI-103 therapeutic doses as compared to control (Figure 3A); no difference was noted amongst the treated groups. However, the effect was found to be cytostatic, at best, and not cytotoxic. PI-103 therapeutic experiments in the UPS/MFH GEM model described above are currently ongoing. Prior to using PI-103 we elected to first test the effects of the mTOR inhibitor, rapamycin, in this model. As shown in Figure 3B, mTOR blockade alone did not impact the growth of UPS/MFH in these mice.

Summary and Future Directions

There is a crucial need for more efficacious therapeutic strategies to improve the outcome of patients suffering from UPS/MFH. Studies reported here demonstrate a potential contribution of deregulated AKT/mTOR signaling to the tumorigenic phenotype of these devastating malignancies. It is of note however, that while studies in cell culture depicted a marked anti-UPS/MFH effect, only a cytostatic delay in tumor growth was noted in response to PI-103 in vivo. These findings have to be further validated and the use of additional inhibitors is currently being planned in order to attain affirmative conclusions. If confirmed, future studies to identify potential compounds that can be utilized in combination with PI3K/mTOR inhibitors should be devised. Such preclinical investigations will hopefully have the potential for significant impact on the management and outcome of patients suffering from UPS/MFH.

Acknowledgement: We would like to thank Dr. David Kirsch for sharing the UPS/MFH GEM model with us and for continuously providing advice regarding the use of this model for therapeutic experiments.

V8N5 ESUN Copyright © 2011 Liddy Shriver Sarcoma Initiative.

By Quansheng Zhu, MD, PhD and Dina Lev, MD 

Introduction to Unclassified Pleomorphic Sarcomas / Malignant Fibrous Histiocytoma

Unclassified pleomorphic sarcomas (UPS) represent a major diagnostic and most importantly a therapeutic challenge. The history of this soft tissue sarcoma subset dates to the 1960’s when it was first defined as malignant fibrous histocytoma (MFH). Thirty years later this terminology was refuted demonstrating that MFH truly represent a cluster of poorly differentiated malignancies of heterogeneous origin (both epithelial and mesenchymal) sharing a similar morphological pattern; a large proportion of tumors originally diagnosed as MFH could be re-classified using electron microscopy evaluation, immunohistochemical analysis, and molecular genetics. With this highly clinically relevant paradigm shift, sarcoma caregivers are still confronted by a subset of soft tissue sarcomas exhibiting pleomorphic morphology that can not be further classified. To that end the 2002 WHO classification recognized the existence of an undifferentiated category of pleomorphic sarcoma, which was termed UPS/MFH.¹ Clinically, UPS/MFH exhibit peak incidence in the 6th and 7th decades of life, show no gender predilection, and occur mainly in the deep soft tissues of the extremities and trunk.² Therapy for localized disease consists of surgery ± radiotherapy. However, 30-50% local recurrence rates (depending on tumor site) remain problematic and are particularly ominous in loci where salvage radical surgery may not be feasible. Systemic failure, mainly lung metastasis, is the major determinant of poor patient outcome and current chemotherapies are of only minimal impact.^3-4 Thus, there is a crucial need for novel anti-UPS/MFH effective therapeutic approaches.

The AKT Pathway as a Novel Therapeutic Target

Molecularly-targeted therapy has emerged as the new anti-cancer paradigm with the hope of more selectively impacting cancer cells rather than normal cells, thus potentially minimizing treatment-related morbidities. This approach has led to recent advances in several diseases, including GIST,⁵ an STS subtype. Effective targeted therapy requires an accessible and functional target. To utilize targeted therapy for UPS/MFH, enhanced knowledge of potential targets and their role in UPS/MFH progression and metastasis are needed. In light of UPS/MFH genetic complexity and molecular heterogeneity, identifying a common molecular deregulation amenable for therapeutic targeting is highly important. AKT kinase, known also as protein kinase B (PKB), a serine-threonine kinase discovered originally as the cellular homolog of the v-AKT oncogene,⁶ is a convergence point for several extracellular and other upstream signals (Figure 1).

AKT activation functions as a master switch to generate a plethora of intracellular signals and intracellular responses. AKT kinase is activated by phosphorylation on two critical residues: threonine 308 (T308) within the activation loop, and serine 473 (S473) at the C-terminal portion of the protein. AKT activation is mediated by PI3 kinase which in turn is activated by several cell surface receptors and cognate molecules,^7-8 while negative regulation of PI3K-dependent AKT activation occurs via tumor suppressor genes such as PTEN and SHIP phosphatases.⁹ Once activated, AKT induces multiple downstream pathways, promoting proliferation and increased cell survival. Multiple mechanisms may contribute to cancer-associated AKT hyperactivation, including amplifications, mutations and/or aberrant overexpression and activation of upstream tyrosine kinase receptors.

Considerable evidence supports a key AKT role in human cancer. For example, AKT activation has been demonstrated in brain, prostate, breast, lung, liver, gastric, colon, ovarian, and endometrial cancers and has been implicated in cancer progression and metastasis.⁹ While not extensively explored in soft tissue sarcoma (STS), recent evidence suggests AKT pathway involvement in complex karyotype STS development and progression.¹⁰ Hernando et al reported increased expression of activated AKT in a large panel of human leiomyosarcoma, UPS/MFH, and dedifferentiated liposarcoma ; using a conditional PTEN knockout mouse model, a critical role for the AKT pathway in smooth muscle transformation and leiomyosarcoma development was shown. Tomita et al identified a correlation between phospho-AKT (pAKT) expression in human STS specimens, subsequent tumor recurrence, and patient survival.¹¹

Recently we showed that pAKT is highly expressed in a diverse cohort of human STS histological subtypes (although UPS/MFH were not included), where it serves as a convergence point for many upstream deregulations. Furthermore, we have demonstrated that AKT inhibition significantly impacts STS cell proliferation, cell cycle inhibition and apoptosis in vitro as well as STS growth in vivo.¹² Together, these results suggest a possible role for AKT activation in STS. Further studies in UPS/MFH will hopefully further illuminate the function of AKT and most importantly, the possible yield of AKT inhibitors as a novel therapy for this STS histological subset.

Hypothesis and Specific Aims

Based on previously published data and our own preliminary studies we hypothesize that AKT activation is a potential common deregulation in UPS/MFH and that AKT inhibition can abrogate tumor and metastasis growth. To that end, we propose to evaluate the impact of AKT blockade on UPS/MFH pro-tumorigenic and pro-metastatic processes. The following two Specific Aims are suggested:

• Aim 1: To evaluate the effect of AKT inhibition on UPS/MFH cell growth, survival, migration, and invasion. Effects of targeting specific AKT isoforms will be tested separately.
• Aim 2: To examine the effect of AKT inhibition on UPS/MFH growth and metastasis in vivo.

UPS/MFH Bioresources and Study Design

The major obstacle for comprehensive "bed-side to bench" UPS/MFH translational studies is the relative lack of requisite bioresources; i.e., tumor specimens, cell lines and animal models. To that end we have established a clinically annotated human UPS/MFH tissue microarray (~180 primary UPS/MFH specimens) that can enable us to evaluate the correlation between marker expression and patient outcome.

A UPS/MFH primary cell culture bank has also been created and currently contains more than 15 fresh surgical specimen-derived UPS/MFH cell strains available for in vitro studies; all have been characterized cytogenetically and shown to grow in soft agar. These cells harbor complex karyotypes and several reproducibly result in tumor growth when injected into SCID mice.

As part of the current proposal gene expression profiling of cell lines/strains and original tumors will be conducted using the illumina platform to determine the applicability of their utilization as UPS/MFH models. Screening several of these cell strains we have recently confirmed high pAKT expression levels. Moreover, through collaboration (Dr David Kirsch, Duke University) we have recently acquired a temporally restricted STS mouse model¹³ where intramuscular injection of adenovirus expressing Cre recombinase (Ad-Cre) into the extremities of mice (mixed S4/SvJae and C57/Bl6) with conditional mutations of Kras and Trp53 (both are very common STS genetic aberrations) induces the development of high grade sarcoma resembling UPS/MFH in more than 90% of mice after a median time of 3 months. Once tumor is established it will grow to 1.5cm within 3-4 weeks; at this time 20% of mice will exhibit spontaneous microscopic lung metastasis. Furthermore, if more time is allowed for metastasis growth via resection of the primary tumor, upon local recurrence (occurring in 100% of mice within several weeks), ~50% of mice will exhibit lung metastasis which have been confirmed histologically to be sarcomas. This model optimally mirrors human UPS/MFH clinical behavior and is therefore a highly relevant system in which to assess novel therapies.

Utilizing these unique bioresources experiments will be conducted towards the completion of both study Aims. Specifically, AKT blockade will be induced using synthetic inhibitors as well as siRNA knockdown of each one of the different three AKT isoforms and effect on tumor cell growth, migration, invasion and survival which will be evaluated utilizing appropriate assays, respectively. Effect on AKT downstream target expression and phosphorylation will be assessed via western blotting and kinase assays. The expression of AKT downsteam targets in human samples and their correlation with patient clinical outcome will be evaluated using the UPS/MFH TMA. Furthermore, in vitro evaluations will be complimented by in vivo testing of the effect of AKT blockade on tumor and metastasis growth using the models described above.

Significance

There is a crucial need for more efficacious therapeutic strategies to improve the outcome of patients suffering from UPS/MFH. Extensive studies of the AKT signaling pathway in a variety of epithelial tumors suggests it to be an attractive molecular target for cancer therapy, leading to the development of several AKT specific inhibitors now ready for clinical trials. Based on limited previously published studies and our own preliminary data, it is possible that AKT-mediated intracellular signaling might be relevant in UPS/MFH progression. Studies proposed here will enable validation of these initial insights while also further evaluating the potential efficacy of AKT as an anti-UPS/MFH therapeutic target. Positive findings have the potential for significant impact on the management and outcome of patients suffering from UPS/MFH.

V7N2 ESUN Copyright © 2010 Liddy Shriver Sarcoma Initiative.

Grant Funding

The Liddy Shriver Sarcoma Initiative funded this $50,000 grant in April 2010. The study was made possible, in part, by a generous donation from Cliff and Arlene Blaker, and by donations made in memory of Ronald Rosenfeld.