Off-label use of targeted therapies in osteosarcomas: data from the French registry OUTC’S (Observatoire de l’Utilisation des Thérapies Ciblées dans les Sarcomes)

High-grade osteosarcoma is the most common malignant bone tumor in adolescents and young adults [1]. Multimodal treatment including chemotherapy and radical surgery increased the Progression-Free-Survival (PFS) from 10 to 65 % [2]. However, we still observe 30 % of relapse, mainly with metastatic stage, with less than 20 % long-term survival for these patients [3].

The role of chemotherapy in recurrent osteosarcomas is not fully established [4]. There is no standard regimen recommended for second-line treatment [1, 5]. Except for muramyl tripeptide (L-MTP-PE) which demonstrated an improvement of median time to relapse from 4,5 months to 9 months in a phase II trial [6], recently tested drugs (etoposide, carboplatine, gemcitabine, high dose chemotherapy [7], ecteinascidin [8], samarium [9]) failed to improve long-term survival of these patients [10, 11].

Several biological pathways are implicated in bone sarcomas and represent a potential interesting approach for the treatment of such tumors with targeted therapies (TTs) : sustaining proliferative signal (IGFR, SHH/GLI, PDGFR, c-KIT), evading cell growth suppressors (p53, RB, CDK), resisting to cell death (ERK activation, proapoptotic molecule inhibition, antiapoptotic molecule activation Bcl2, syndecan-2), enabling replicative immortality, increasing angiogenesis (VEGFR, IGFR, PDGFR, HIF1α) and activating invasion and metastasis, genome instability (p53, GADD45), evading immune destruction (IFN), or interacting with the bone microenvironment (RANK/RANKL/OPG) [12]. Unfortunately, the rarity of these pathologies and the specificity of the pediatric population don’t hold pharma industries nor governments to delineate phase III trials and prove the benefit of such compounds for refractory osteosarcomas.

In 2008, the GSF-GETO established a National Observatory for The off-label Use Of Targeted Therapies in Sarcomas (OUTC’S) as a resource for the research into the use of TTs in routine practice. All medical data regarding the use of off-label TTs in sarcomas was collected in a prospective way to analyze activity and toxicity of TTs in these tumors [13]. This report aims to describe the utilization, efficacy, and safety data on osteosarcoma patients registered in OUTC’S in order to identify TTs which warrant further investigations in this pathology.

This study reported a 45.5 % disease control rate with TTs used off-label in refractory relapsed osteosarcomas with a good tolerance profile. In a multivariate analysis, PFS seemed superior for patients receiving sirolimus compared to other TTs.

Many molecular abnormalities are identified in osteosarcomas giving the cancer cells some particular characteristics: proliferative signals (PDGFR, IGFR, c-KIT), resistance to retroaction signals (p53, RB), resistance to cell death (ERK, Bcl-2), angiogenesis (VEGFR, PDGFR), resistance to immune destruction (IFN) [12]. Potential TTs could either inhibit growth factor signaling pathways, or enhance apoptosis, or inhibit the metastatic process, or modulate the antitumor immune response, or modulate the bone microenvironment to increase local control of the primary tumor, limit metastatic spread, and finally improve patient survival [17].

mTOR is an intracellular protein, playing a major role in protein synthesis and influencing the cell growth, differentiation and apoptosis: this pathway is unregulated in many cancers, leading to the permanent activation, often under the influence of IGF1R. mTOR also plays a role in angiogenesis by controlling the production of HIF (Hypoxia Inducible Factor) [18]. Preclinical studies demonstrated that sirolimus, the main mTOR inhibitor, blocks the ezrin pathway implicated in the metastatic migration of osteosarcomas [19]. In 2012, a phase II study reported a clinical benefit in 28.8 % of patients treated with ridaforolimus for a metastatic or inoperable sarcoma with an increased PFS compared to untreated patients [20]. Another phase II study testing the association of sirolimus and cyclophosphamide in soft tissue and bone sarcomas, highlighted a synergic effect of the two drugs, leading to an increased PFS with a good tolerance [21]. A double blind phase III maintenance trial comparing ridaforolimus and placebo in advanced sarcoma after stabilization or response with chemotherapy, enrolled 50 bone sarcoma patients showing a longer PFS and a 28 % reduction in the risk of death or progression with the maintenance strategy [22]. This data constituted the rational for using mTor inhibitors in refractory osteosarcomas, first in adults and recently in pediatric population (Table 5). Data provided by OUTC’S registry confirmed the value of this agent in osteosarcomas especially combined with conventional chemotherapy to prolong survival and time to progression in this particularly dismal prognosis group.

Sorafenib inhibits B-raf, c-KIT, PDGFR, VEGFR and RET. In osteosarcoma, sorafenib inhibits the proliferation of tumor, angiogenesis (VEGF), invasion (MMP2), the emergence of pulmonary metastases (Erzin/β4-integrin/ PI3K) and induces apoptosis [23]. This drug has already been approved for renal and hepatocarcinoma treatment and has shown good responses in angiosarcomas [24]. Yet, the use of sorafenib in osteosarcomas is mainly based on a phase II study, conducted in 35 patients with progression despite standard treatment and reporting 5 PRs, a clinical benefit rate of 29 % and a four-month PFS of 46 % [25].

Sunitinib inhibits FLT3, c-KIT, PDGFR and VEGF. Efficacy was observed with in vivo models, mostly pediatric tumors, including Ewing sarcoma xenografts [26]. Clinical benefit is reported for 4 patients with sarcomas in phase I [27] and 34 in phase II studies [28].

Pazopanib is mainly steered against VEGFR and PDGFR. A phase II study reported 9 cases of PR and improvement of OS and PFS for 143 patients with progressive soft tissue sarcoma [29]. A randomized double blind phase III study of pazopanib versus placebo, showed improved OS and PFS for a metastatic soft tissue sarcoma after failure of chemotherapy treatment [30]. A randomized double-blinded phase II is currently open to evaluate regorafenib, a promising TKI [31] in advanced bone sarcomas [32]. Based on this literature, TKI have been used off-label in adult refractory sarcoma first, thereafter by pediatricians influenced by adult practices despite the paucity of pharmacological data in pediatric population.

We report in this study only one objective response after initiation of TT. It has been suggested that the evaluation of TTs efficacy could not be done by RECIST compared to conventional treatments because TKI are mainly cytostatic. Some cases of cystic tumors after treatment by TKI have been reported [33]. Indeed, a stable disease induced by a TT could be considered as a satisfying response and a significant clinical benefit given the poor prognosis of metastatic refractory sarcomas. In order to guide the objectives of clinical trials, the EORTC Sarcoma Group (European Organization for Research and Treatment of Cancer) defined that a second-line treatment could be considered active if it showed a 6-month PFS of 40 % and as inactive if it was below 20 % [33]. In our study, six-month PFS was 15 % (22 % with sirolimus, 0 % with TKI), but all patients included had very poor prognosis factors: inoperable tumor, high grade histology, treatment-line failures. Most published series about this population reported dismal prognosis, with short median survival especially after several relapses [11, 34]. In this cohort, the one-year OS of 24 % and median survival of 6.8 months could be a significant result. The difference observed in median PFS between sirolimus group and TKI group (2.3 versus 1.8 months) encourages investigating this drug in a clinical trial.

Given the number of different mechanisms involved in carcinogenesis and treatment failures, a molecular study of each tumor could guide the indications of TTs and compounds. Some mechanisms lead to the cell resistance to Sirolimus, in particular because only the complex MTORC1 is sensitive to Sirolimus, whereas MTORC2 is resistant [17]. The activation of MTORC2 leads to treatment failure. This mechanism can be blocked by the association with Sorafenib: in vitro and in vivo, the combination of the two drugs increases the anti-tumor, anti-angiogenic and anti-metastatic activity [35]. Despite this data, no combination of TKI with mTOR inhibitor was reported in OUTC’S: it could be worth exploring this strategy.

In this study, tumor control lasted more than 6 months for 5 patients. These patients had a median age of 17 at the TT initiation, which is below the median age of the whole group and compatible with data showing a better response to chemotherapy in children [36]. All these patients received sirolimus in association with cyclophosphamide. One patient was treated at first relapse and the others at second relapse, suggesting that efficiency of sirolimus could be optimized when used with minimal tumoral disease.

We must underscore that three patients received a maintenance treatment combining sirolimus-cyclophosphamide, after complete remission by surgery and chemotherapy. This strategy is developing in sarcomas, supported by studies suggesting that it could improve survival and decrease the risk of relapse in high-risk patients [22, 37] and must be confirmed in randomized clinical trial dedicated to maintenance therapy including PFS, OS and quality of life.

Observed data of toxicity are similar to what was already described in clinical trials [13]. No major toxic effect has been reported and only one patient had to stop TTs because of toxicity, showing that tolerance to TTs is acceptable, even in children.

The main limitation of this study is the small number of patients, due to the rarity of these tumors, which can reduce the statistical power, in particular for the comparison between TKI and sirolimus (since the CI of the hazard ratio approximates 1). The specificities of pediatric population make it difficult to launch clinical trials assessing efficacy of TTs in osteosarcomas. Registering patient in a national database like OUTC’S is an opportunity to obtain more information about safety and efficacy of drugs used off-label with a rational based on published data.

Targeted therapies could play a part in the treatment of refractory osteosarcomas or in maintenance for patients with a high risk of relapse. Tolerance is acceptable, even for patients under 18. This data suggests that sirolimus could have an interesting anti-tumor activity in osteosarcomas and deserves to be evaluated in a prospective trial, either alone or in combination with chemotherapy.