Rodrigo R. Muhoz, MD, William D. Tap, MD

The characterization of molecular abnormalities implicated in the tumorigenesis of sarcomas is being increasingly applied to the classification, prognostication, and in particular situations, management of these diseases. Although the treatment for the majority of patients still relies on conventional cytotoxic agents, the elucidation of underlying genetic aberrations and mechanisms of these diseases are gradually translating into therapeutic progress. In this review, we address the emerging therapeutic strategies aimed at specific molecular targets that could potentially change the approach to patients with soft-tissue sarcomas.

Sarcomas represent a diverse group of neoplasms of mesenchymal origin that correspond to approximately 1.5% of all malignancies in adults.

Genetic aberrations in sarcomas occur as either simple karyotypic abnormalities, such as chromosomal translocations, amplifications, and deletions, or complex/unbalanced karyotypic changes that result from accumulated nonspecific gains and losses. Chromosomal translocations associated with gene fusions and subsequent transcriptional dysregulation account for the majority of the genetic hallmarks identified in sarcomas.

In addition to the formation of chimeric transcription factors involving oncogenes, translocations may prompt the activation of proteins with tyrosine kinase function or autocrine growth factors. Conversely, tumors without specific cytogenetic abnormalities are characterized by genome instability, which results in multiple and alternative genomic aberrations of unclear significance and a complex karyotype.

One of the major breakthroughs that emerged from a molecular-based approach is exemplified in gastrointestinal stromal tumors (GIST).

Despite this initial enthusiasm, targeting molecular alterations has been less fruitful in other histologies, in which genetic alterations rarely act as driver mutations. Nevertheless, the evolving molecular characterization and widespread use of next-generation sequencing techniques has unveiled a series of new potential targets through a better characterization on the molecular landscape. In this review, we will focus on novel potential molecular targets in non-GIST sarcomas, based on available clinically-relevant results.

Platelet-derived growth factor receptors (PDGFRs) are tyrosine-kinase receptors consisting of either &alpha;- or &beta;-chains forming 3 possible receptors: PDGFR-&alpha;&alpha;, PDGFR-&alpha;&beta;, and PDGFR-&beta;&beta;. The activation of PDGFR occurs upon interaction with platelet-derived growth factor (PDGF) ligands in the extracellular domain, which include 5 different isoforms.

After dimerization, each PDGFR partner phosphorylates tyrosine residues located on the cytosolic tails.

Across different subtypes of sarcomas, signaling through PDGF/PDGFR has been shown to promote progression through cell cycle and avoidance of apoptosis, and result in pro-angiogenic effects and modulation of the tumor stroma.

Activation of PDGFR results in downstream signaling through multiple pathways involving phosphatidylinositol-3-kinase (PI3K), phospholipase-C gamma (PLC&gamma;), Rous sarcoma oncogene (SRC) kinases and rat sarcoma oncogene (RAS)/mitogen-activated protein kinase (MAPK) proteins.

In normal cells, high expression of PDGFR&beta; is seen in fibroblasts, pericytes and smooth muscle, and PDGFR&alpha; in megakaryocytes, fibroblasts, myoblasts, pericytes, smooth muscle, and neurons.

In dermatofibrosarcoma protuberans (DFSP), inhibition of PDGF&beta;/PDGFR&beta; with tyrosine kinase inhibitors

, showed significant clinical activity, culminating with the approval of imatinib by the US Food and Drug Administration (FDA) for the treatment of patients with advanced/ metastatic disease.

The molecular hallmark of DFSP is the recurrent translocation t(17;22) (q22;q13), that results in the fusion of the collagen type I alpha 1 (COL1A1) promoter to PDGF, leading to constitutive activation of this pathway.

In a pooled analysis including 24 patients treated with imatinib, objective response rate (ORR) was 46% and median time to tumor progression (mTTP) was 1.7 years.

In addition, the PDGF/PDGFR pathway has been associated with the pathogenesis of different sarcomas including rhabdomyosarcoma, Kaposi&rsquo;s sarcoma, synovial sarcoma, chondrosarcoma, osteosarcoma, and Ewing family sarcoma. Elevated expression of PDGF&beta; and co-expression of PDGF&beta; and PDGFR&alpha; has been correlated with high histological grades and poor prognosis in sarcomas.

Postulated mechanisms include the modulation of angiogenesis, regulation of stroma-derived fibroblasts and autocrine stimulation of cellular growth.

Overexpression of mRNA for PDGFR&alpha; and PDGF has been documented in sarcoma, highlighting the relevance of this autocrine loop.

In addition, preclinical data suggest that blocking PDGF/PDGFR results in a potentiation of response to chemotherapy.

Pazopanib, an oral, multi-targeted tyrosine kinase inhibitor also active against PDGFR, was shown to prolong progression-free survival (PFS) versus placebo in pretreated patients with advanced non-adipocytic soft tissue sarcomas (median PFS, 4.6 months vs 1.6 months; hazard ratio [HR], 0.31; P <.0001) in a phase III trial, and is currently approved by the FDA for clinical use.

Similarly, regorafenib, a distinct multikinase inhibitor, resulted in improved PFS in previously treated patients with leio- myosarcoma (4.0m vs 1.9m; HR 0.49; p=0.017) and other soft-tissue sarcomas (4.6m vs 1.0m; HR 0.38; p=0.002) in a randomized, placebo controlled, phase II study. In line with these observations, two partial responses were seen in a series of 31 patients with PDGFRb-expressing solitary fibrous tumors (SFT) treated with sunitinib, also a tyrosine-kinase inhibitor. Of note, recent studies revealed a 12q13 intrachromossomal fusion in SFT, resulting in the NAB2:STAT6 fusion protein and subsequent transcriptional activation of several genes and expression of receptor tyrosine kinases.

More recently, approaches focusing on more specific inhibition of PDGFR&alpha; with olaratumab led to promising results. Olaratumab (IMC-3G3) is a recombinant IgG1-type monoclonal antibody that binds with high affinity to PDGFR&alpha; and prevents PDGF from binding to the receptor; olaratumab also blocks downstream signaling through Akt and mitogen-activated protein kinase (MAPK).

The results of a phase 1b/2 study evaluating the efficacy of doxorubicin with or without olaratumab have been recently presented. In the randomized phase II portion of the study including 133 patients, the addition of olaratumab resulted in a dramatic gain in overall survival (OS) (median OS 25m vs 14.7m; HR 0.441; p=0.0004), despite only a marginal improvement in PFS (median PFS 6.6m vs 4.1m; HR 0.672; p=0.0615). Noteworthy, adverse events leading to treatment discontinuation were less frequent in the combination arm (13% versus 22%). A pivotal phase III trial addressing this combination in comparison to doxorubicin and placebo is currently ongoing (NCT02451943).

Potential Targets for Inhibition: KDR, PTPRB, and PLCG1 in Angiosarcomas

Angiosarcomas (ASs) are malignancies of vascular origin associated with a significant risk of metastases and poor prognosis.

In addition to its vascular origin, plentiful evidence indicates the activation of pro-angiogenic pathways in the development of ASs, including overexpression of vascular endothelial growth factor (VEGF) and vascular endothelial growth factor receptors (VEG- FRs).

Bevacizumab, a recombinant monoclonal antibody targeting VEGF-A, was studied in a single arm, phase II trial in patients with tumors of vascular origin (AS, N=23; epithelioid hemangioendothelioma, N=7). Four objective responses occurred (2 patients with AS and 2 patients with epithelioid hemangioendothelioma), resulting in an ORR of 17%; 15 patients (50%) (N=15) achieved stable disease and median PFS was 12.4 weeks.

In the recently presented ANGIOTAX-PLUS study, patients with AS were randomized to receive weekly paclitaxel with or without bevacizumab.

There were no statistically significant differences in ORR (40% vs 50%), median PFS (6.8 months vs 6.9 months), or median OS (19.5 months vs 15.9 months) between treatment arms.32 Sorafenib, a multi-targeted tyrosine kinase inhibitor with anti-angiogenic properties, was investigated in a prospective study encompassing different histologies.

Among a subset of 37 evaluable AS patients, ORR was 14% and median PFS was 3.8 months, with the majority of patients with AS having stable disease as best response.

Similarly, sorafenib resulted in response rates of 0% and 14.6% and transient tumor control in phase II trials reported by von Mehren et al35 and Ray-Coquard et al34, respectively. Recent studies applying whole genome, whole exome, and targeted sequencing techniques provided further insight into potential therapeutic targets in ASs.

Using gene expression profiling, Antonescu et al showed upregulation of vascular-specific receptor tyrosine kinases, including kinase insert domain receptor (KDR) and tyrosine kinase with immunoglobulin-like and EGF-like domains 1 (TIE1) and FLT1. Of note, KDR gene mutations were associated with high expression of KDR protein and occurred in 10% of patients with ASs (all arising from the breast).

In an analysis of 26 primary and 29 secondary ASs, Styring et al36 demonstrated significant deregulation of 103 genes. Once again, secondary ASs showed a genetic signature distinct from primary ASs, with upregulation of MYC, KIT, and RET and downregulation of CDKN2C.36 Behjati et al37 reported at least 1 driver mutation in angiogenesis signaling genes in 15 of 39 ASs (38%), the most common being MYC. Of note, these driver mutations were not mutually exclusive.37 PTPRB truncating mutations were found in 26% (10 out of 39) of AS samples and exposure of PTPRB-silenced cell cultures to sunitinib or vatalanib resulted in suppression of angiogenesis. In addition, activating mutations of PLCG1, a signal transducer of tyrosine kinases involved in cell growth, signaling, and maintenance of membrane phospholipids, were documented in 9% of the cases.

In line with these observations, a retrospective series investigating the effects of sorafenib in cohorts limited to patients with RT-associated AS of the breast found MYC and FLT4 coamplification in 2 out of 3 patients with objective responses (in total, 9 patients were assessable for response; ORR = 33%).

In summary, although objective responses are limited to a subset of patients, tyrosine kinase inhibitors and agents targeting angiogenesis are active and able to induce disease stabilization in unselected AS patients, representing a reasonable alternative beyond conventional cytotoxic chemotherapy. Findings from molecularly based studies provide conceptual models for fur- ther investigation of anti-angiogenic therapies in AS.

ALK in Inflammatory Myofibroblastic Tumors

Inflammatory myofibroblastic tumors (IMTs) are tumors composed of spindle cell proliferation mixed with a collagenous stroma in an inflammatory background of granulocytes, lymphocytes, and plasma cells.

In almost 50% of the cases, IMT harbors chromosomal aberrations that result in rearrangements involving the anaplastic lymphoma kinase (ALK) locus on chromosome 2p23 or increased copy numbers without rearrangement. The product of this genetic variation is a receptor tyrosine kinase (RTK) that is constitu-tively activated, leading to cell growth and tumorigenesis.

Crizotinib is an oral tyrosine kinase inhibitor targeting ALK, MET, and ROS1 that was initially approved for the treatment of patients with nonsmall-cell lung cancer. Recently, Butrynski et al reported a partial response in a patient with ALK-rearranged IMT treated with crizotinib, suggesting that this agent might also be effective in treating patients with ALK-rearranged IMT.41 Given these promising results and based on a strong rationale, a phase II trial is ongoing (NCT01524926).

MDM2 and CDK4 in Well-Differentiated/ Dedifferentiated Liposarcoma

Liposarcomas are soft tissue sarcomas of adipocytic origin that comprise different histologic subtypes, including well-differentiated liposarcoma (WDLPS) and dedifferenti- ated liposarcoma (DDLPS).

WDLPS and DDLPS are marked by amplifications in- volving chromosome 12, which occur in more than 80% to 90% of the cases.

These 12q gains result in co-amplification of cyclin-dependent kinase 4 (CDK4) and MDM2 oncogenes. Several agents targeting MDM2 and CDK4 are currently in development, although the success rate with inhibitors used as single agents has been limited.

CDK4 is a cyclin-dependent kinase that regulates the G1S phase transition in the cell cycle and promotes tumorigenesis through inhibition of the retinoblastoma (Rb) protein family.

Following prolonged disease stabilization in patients with WDLPS/DDLPS in a phase I trial,

our group reported the results of a phase II trial with the CDK4/6 inhibitor palbociclib (PD0332991), which demonstrated a PFS rate of 66% at 12 weeks. Among 29 evaluable patients with CDK4 amplification enrolled in the therapeutic cohort, the median PFS was 18 weeks.

Palbociclib was recently approved by the FDA for the treatment of patients with breast cancer. LEE011, a distinct orally bioavailable CDK4/6 inhibitor, was also shown to be tolerable and able to induce long-lasting disease stabilizations in a phase I clinical trial including multiple solid tumors. Stable disease sustained for more than 4 and 6 cycles occurred in 26% and 14% of the patients, respectively.

Additional studies to better characterize the efficacy of LEE011 are ongoing (NCT01237236, NCT02187783, NCT02343172).

The MDM2 gene encodes an E3 ubiquitin ligase that binds to tumor suppressor p53 and targets p53 for proteasomal degradation. Therefore, MDM2 amplification/over-expression lead to increased p53 inactivation and result in a pro-oncogenic effect. In a pilot study investigating the oral MDM2 inhibitor RO5045337 (also known as RG7112), both reactivation of the p53 pathway and decreased cell proliferation were demonstrated between pre- and post-treatment biopsies.49 Among 20 patients, there was 1 partial response and 14 patients achieved disease stabilization.

In another phase I dose escalation study that included a sarcoma extension cohort, RO5045337/RG7112 demonstrated manageable toxicities, biomarker activity, and objective responses in patients with liposarcomas.50 Nevertheless, the same drug in combination with doxorubicin (NCT01605526) resulted in high incidence of neutropenia and thrombocytopenia.

Additional agents targeting MDM2 are currently un- der clinical development. These include SAR405838, presently being investigated in early-phase clinical tri- als (NCT01636479, NCT01985191), and AMG-232, a MDM2 inhibitor derived from AM-8553 (NCT01723020, NCT02110355).

Microphthalmia Family Transcription Factors in Alveolar Soft Part Sarcomas and Clear Cell Sarcomas

The microphthalmia (MiT) family transcription factors include MITF, TFE3, TFEB, and TFEC, which are involved in regulation of cell growth, differentiation, and survival. Dysregulation of MiT family members results in aberrant activation of multiple genes, including c-Met. This leads to oncogenesis in different cancer types and is thought to be particularly relevant in alveolar soft part sarcoma (ASPS) and clear cell sarcoma (CCS).

The c-Met product is a receptor tyrosine kinase activated upon interation with hepatocyte growth fact (HGF) frequently expressed in mesenchymal cells and also linked to intracellular signaling through intracellular signaling MAPK and AKT path- ways,

providing a rationale for further study of MET and other TKI inhibitors in ASPS and CCS.

In ASPS, TFE is fused to the ASPL gene as a consequence of a typical unbalanced t(X;17)(p11;q25) translocation, resulting in the ASPL-TFE3 fusion protein. Although the function of ASPL needs to be further clarified, TFE3 acts as a transcription factor and the fusion transcripts lead to MET transcriptional upregulation through the cMet gene.

Also an MiT-associated tumor, CCS expresses different cytogenetic abnormalities. Typically, EWSR1-ATF1 fusion results from the t(12;22)(q13;q12) translocation

; in a smaller subset of patients, t(2;22)(q34;q12) translocation produces a EWSR1-CREB1 fusion-gene.

These abnormalities induce MITF and c-Met expression. In preclinical models, blockade of c-Met activity and HGF resulted in CCS cell growth arrest.

Disappointingly, a phase II study of the MET inhibitor tivatinib in MITF-associated tumors produced no objective responses in 27 ASPS patients. In the same trial, only 1 objective response occurred among 11 patients with CCS.

In both entities (ASPS and CCS), although responses were seen with multi-targeted tyrosine kinase inhibitors, the role of MET inhibitors remains to be further investigated and a clinical trial with cabozantinib is ongoing (NCT01755195).

The PI3K-Akt-mTOR Pathway in PEComas, Myxoid/ Round Cell Liposarcomas, and Angiosarcomass The phosphotidylinositol 3-kinase (PI3K)-Akt-mammalian target of rapamycin (mTOR) is a key pathway in the control of cell proliferation, survival, and angiogenesis. Several pre-clinical and translational studies have addressed the role of the PI3KAkt- mTOR pathway and investigated the mecha- nism of action of mTOR inhibitors; evidences of activation of this pathway have been well demonstrated in sarcomas.

Ridaforolimus showed some degree of activity in phase I/II studies,

but failed to demonstrate a survival benefit when used as a maintenance therapy in a randomized phase III trial (

= .46).66 Different mTOR inhibitors, in- cluding temsirolimus, everolimus, and sirolimus, were also evaluated in small prospective trials.

Despite limited efficacy in unselected patients with sarcomas, a histology-driven approach could further clarify the role of mTOR inhibitors in the treatment of this group of malignancies. Perivascular epithelioid cell tumors (PEComas) are rare mesenchymal neoplasms that exhibit features of both smooth muscle and melanocytic origin

and are often characterized by the presence of inactivating mutations in tuberous sclerosis complex (TSC) genes TSC1 and TSC2, which are upstream inhibitors of the mTOR complex 1 (mTORC1).

Although data are limited due to the rarity of this disease, durable responses have been reported with sirolimus and everolimus. In a series of 5 patients with non-pulmonary PEComas treated with sirolimus or everolimus, 3 complete responses and 1 partial response occurred. Molecular studies identified TSC2 aberrations in 4 of these patients.

Barretina et al reported a relatively high prevalence of PIK3CA mutations (18%) in patients with myxoid/round-cell liposarcomas.

Similarly, Demicco et al showed that although activating PIK3CA mutations occurred in only 14% of the cases and more frequently in round cell than myx- oid tumors (33% vs 3%; P = .013), additional mechanism for PI3K/Akt activation were observed, including loss of PTEN and upregulation of IGF1R.

In angiosarcomas, activation of the PI3K/Akt/mTOR pathway occurs as a result of VEGF downstream signaling, involving mTORC1, mTORC2, and phosphorylation of p70 S6-kinase (S6K). In preclinical models, the growth of AS cell lines was supressed by celecoxib and rapamycin,75 and inhibition of mTORC1, mTORC2, and phosphorylation of Akt was demonstrated.

In addition to conventional mTOR inhibitors, there has been growing interest in the investigation of dual PI3K- mTOR inhibitors in these diseases, including BEZ235 (NCT01690871), VS5584 (NCT01991938), and DS7423 (NCT01364844).

The isocitrate dehydrogenase protein (IDH) family comprises 3 enzymes, IDH1, IDH2, and IDH3, which show different intracellular spatial distributions. These isoenzymes are involved in the oxidative decarboxylation of isocitrate. Mutations in genes encoding IDH1 and IDH2 result in proteins with aberrant enzymatic activity that produce (D)-2-hydroxyglutarate (2HG). Cells with mutant IDH1 and IDH2 show accumulation of 2HG and subsequent epigenetic modifications involved in tumorigenesis.

Initially identified in gliomas, IDH mutations are present in up to 50% to 60% of chondrosarcomas.

Preclinical data suggest that differentiation of mesenchymal progenitor cells can be disrupted by transfection with IDH2 mutant constructs in mice, and that exposure to the demethylating agent 5-azacytidine was able to restore a differentiatedphenotype.

Both IDH1 (AG120) and IDH2 (AG221) inhibitors are currently under development for patients with different malignancies associated with IDH mutations, including gliomas and chondrosarcomas.

Desmoid tumors (DT) (or aggressive fibromatosis) are uncommon mesenchymal neoplasms characterized by locoregional aggressiveness, despite the absence of capacity of metastatic dissemination. The molecular hallmark of DT is the deregulation of the Wnt signaling pathway, frequently associated with gain-of-function mutations involving genes enconding for b-catenin (CTNNB1). Recent studies have suggested that DT also frequently express Notch pathway components and the existence of crosstalk between Wnt and Notch pathways, with preclinical evidence providing the rationale for the clinical exploration of Notch signaling suppression through gamma-secretase inhibition.

In a phase I trial, the oral gamma-secretase inhibitor PF-03084014 resulted in 5 partial responses among 7 patients with DT (objective response rate of 71.4% in this population), with evidence of down-modulation of NOTCH-related gene expression, and further work is needed to confirm these preliminary findings.

While conventional cytotoxic chemotherapy remains the predominant treatment modality for patients with advanced sarcomas, genomics may continue to pave the way for the development of personalized care. Despite evidence supporting a mech- anistic rationale in different types of sarcomas, targeting these molecular aberrations translated into meaningful therapeutic and clinical benefits only in a small subset of scenarios and the use of wide molecular analyses remains investigational and limited to select clinical trials. Nevertheless, some of these discoveries begin to shape the treatment decisions in specific histologies, as exemplified by the use of mTOR inhibitors in PEComa, imatinib in DFSP, pazopanib in pre-treated, non-adipocytic sarcomas, agents targeting angiogenesis in vascular sarcomas and, potentially MDM2 and CDK4 inhibitors in lipo- sarcomas.

Further exploration of novel targets in sarcomas demands the development of adequate cell lines and preclinical models, as well as practical tools for the selection of patients for early phase clinical trials, in order to potentiate the benefit of tailored approaches.

Ducimetiere F, Lurkin A, Ranchere-Vince D, et al. Incidence of sarcoma histotypes and molecular subtypes in a prospective epidemiological study with central pathology review and molecular testing. PLoS One. 2011;6(8):e20294.

Wibmer C, Leithner A, Zielonke N, et al. Increasing incidence rates of soft tissue sarcomas? a population-based epidemiologic study and literature review. Ann Oncol. 2010;21(5):1106-1111.

Siegel RL, Miller KD, Jemal A. Cancer statistics, 2015. CA Cancer J Clin. 2015;65:5 -29.

Fletcher C DM, Bridge JA, Hogendoorn P, et al. WHO classification of tumours of soft tissue and bone. Fourth Edition. WHO 2013;IARC WHO Classification of Tumours, No 5. ISBN-13 9789283224341

Taylor BS, Barretina J, Maki RG, et al. Advances in sarcoma genomics and new therapeutic targets. Nat Rev Cancer. 2011;11(8):541-557.

Frith AE, Hirbe AC, Van Tine BA. Novel pathways and molecular targets for the treatment of sarcoma. Curr Oncol Rep. 2013; 15:378-385.

Jain S, Xu R, Prieto VG, et al. Molecular classification of soft tissue sarcomas and its clinical applications. Int J Clin Exp Pathol. 2010;3(4):416- 428.

Miettinen M, Lasota J . Gastrointestinal stromal tumors-definition, clinical, histological, immunohistochemical, and molecular genetic features and differential diagnosis. Virchows Arch. 2001;438:1.

Hirota S, Isozaki K, Moriyama Y, et al. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science. 1998;279:577580.

Heinrich MC, Corless CL, Duensing A, et al. PDGFRA activating mutations in gastrointestinal stromal tumors. Science. 2003;299:708710.

Tap WE, Schwartz GK. That&rsquo;s the &ldquo;GIST&rdquo; of it: use of adjuvant imatinib after resection of a primary GI stromal tumor. J Clin Oncol. 2014;32:15431546.

Fredriksson L, Li H, Eriksson U. The PDGF family: four gene products form five dimeric isoforms. Cytokine Growth Factor Rev. 2004;15: 197204.

Articles

In this review, Rodrigo R. Muhoz, MD, William D. Tap, MD, and Sandra P. D'Angelo, MD address the emerging therapeutic strategies aimed at specific molecular targets that could potentially change the approach to patients with soft-tissue sarcomas.

Rodrigo R. Muhoz, MD, William D. Tap, MD | Authors

Articles

Novel Molecular Targets for Drug Development in Non-GIST Sarcomas