Osteosarcoma is the most common pediatric bone tumor, accounting for approximately 3% of all malignancies in children (1). Multi-agent chemotherapy and surgical resection of the tumor and the metastases are the cornerstones of the current treatment guidelines. With this multidisciplinary approach, up to 70% of the patients with localized disease can be long term survivors. The prognosis is poorer for patients with primary metastatic, nonresectable, or relapsed disease, and the overall survival in these cases was reported to be around 30%. Metastases at the time of the diagnosis, axial location, and chondroblastic histological subtype can indicate worse prognosis. Despite the advances in chemotherapy and surgical methods, the overall survival rates have not improved in the past few years (2). In order to develop new targeted therapies, better understanding of the molecular mechanisms of osteosarcoma is needed.

Research of cancer metabolism is an area that has gained attention in recent decades. In order to maintain the high proliferation rate, malignant cells need to rewire their metabolic pathways to gain more energy for growth and invasion. These changes consist of upregulated glycolysis (even in the presence of oxygen) also known as the “Warburg-effect,” increased glutaminolysis and lactate synthesis, alterations in lipid metabolism, and tricarboxylic acid cycle/oxidative phosphorylation. As a result of these alterations, several alternative metabolic pathways can be activated in tumor cells (3).

The role of phosphoinositide-3-kinase (PI3K)/protein-kinase B (AKT)/mammalian target of rapamycin (mTOR) pathway in tumorigenesis has been investigated thoroughly in the past years. The serin/threonine kinase mTOR is a master regulator of cell growth and proliferation, it promotes anabolic processes (protein, nucleotide, fatty acid synthesis) and inhibits catabolism. mTOR kinase is found in two different complexes, mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2). Each complex contains different proteins as they have distinctive roles and they also differ in inhibitor sensitivity. mTORC1 contains mTOR, regulatory‐associated protein of mTOR (raptor), and G protein β subunit‐like (Gβ L, also known as mLST8), and is sensitive to rapamycin, whereas mTORC2 consists of mTOR, Gβ L, and rapamycin-insensitive companion of mammalian target of rapamycin (rictor), and could be insensitive or less sensitive to rapamycin. The activation of mTORC1 leads to increased protein synthesis, cell growth and proliferation, whereas mTORC2 regulates cytoskeleton reorganization and cell survival (4, 5). Due to mutations of the regulatory proteins/enzymes of the cellular signaling network, constitutive activation of this pathway was described in a variety of tumors. Rapamycin analogues (rapalogs) have been investigated as therapeutic agents in several human malignancies, and successfully used in certain diseases such as renal cell carcinoma and mantle cell lymphoma (6).

The mTOR pathway is critical in tumor survival and progression, and it is potentially activated in osteosarcoma, as it has been already reported, although the mechanisms of this activation are poorly understood. Some upstream (PI3K) and downstream (S6K1, 4EBP1, and eIF4E) effectors of the pathway are overexpressed in osteosarcoma, leading to alterations of cellular transformation and poorer clinical prognosis (7, 8). Inhibition of mTORC1/2 or knockdown of mTOR or rictor prevents proliferation in osteosarcoma cell lines, and the combination of mTOR kinase inhibitor PP242 with cisplatin leads to synergistic antitumor activity (9). Combination with cisplatin or bevacizumab can enhance the effectivity of temsirolimus in osteosarcoma xenografts (10), and rapamycin with spautin-1 (specific and potent autophagy inhibitor-1) was found to be effective in apoptosis induction, and can be a possible treatment option (11). Despite encouraging preclinical data, mTOR inhibitor monotherapies and combined therapies have shown poor antitumor effect in the treatment of osteosarcoma (12, 13).

Better understanding of the different tumor types’ metabolic adaptation changes is crucial in order to improve personalized therapies. Our aim was to characterize the mTOR profile and certain metabolic alterations in pediatric osteosarcoma to determine the interactions between the mTOR pathway and metabolic pathways.

Osteosarcoma is an aggressive bone sarcoma, affecting mainly children and adolescents. In the 1960s the treatment of this disease was surgery alone, resulting in poor survival rates of 11%. The initiation of adjuvant chemotherapy in the 1970s increased 5-year survival rates up to 70%, but no significant improvements have been made since (2, 17). The survival of relapsed patients, or cases with primary progression remains poor. In order to improve outcome, it is crucial to investigate molecular mechanisms behind osteosarcoma tumorigenesis and progression. Our aim was to characterize the mTOR and metabolic profile of osteosarcoma cases, and to determine potential targets for personalized therapies.

Numerous preclinical data suggest the importance of mTOR activity in osteosarcoma cell proliferation and survival. Investigations with NVP-BEZ235, a PI3K-mTOR dual inhibitor suppressed osteosarcoma cell growth in vitro and in vivo (18). Similar results were found with the novel mTORC1/C2 dual inhibitor RES-529, which potently inhibits human osteosarcoma cell growth (19). mTOR inhibitors alone, or in combination have been also explored in phase I and II trials (20). Everolimus, mTORC1 inhibitor, combined with sorafenib induced partial response in 5%, and disease control in 63% of the studied 38 osteosarcoma patients (21). Sirolimus, mTORC1 inhibitor, was combined with gemcitabine in a recent phase II trial, resulting in low response rate (6%), but 42% of the patients had stable disease at 4 months of treatment (22). As the data suggest, inhibition of the mTOR pathway has been expected as a promising therapeutic approach, but mTOR inhibitors have not proved their efficacy in osteosarcoma treatment, only in combination with different cytotoxic agents.

It is proved that the mTOR pathway has a potential role in facilitating metastasis. In a murine model of osteosarcoma, rapamycin reduced tumor cell metastasis by blocking the mTORC1 pathway (23). Even though our results suggested low mTOR activity in most of the samples, we found significantly higher H-scores of rictor in primary metastatic diseases. Similar results were found in metastatic lung adenocarcinoma (24). Consistent with these findings, pmTOR activity turned out to be higher in poor responders. In addition, the only patient with mTORC2 activity had a primary progressive disease, showed no clinical response to chemotherapy, and passed away due to disease progression. We observed low expression of rictor in young patients. Younger age at time of diagnosis is associated with better survival (25, 26). We also found significant difference in pmTOR activity between the osteoblastic subtype and other histological subtype group, but due to the low number of non-osteoblastic cases, this data cannot be considered representative. Our results suggest the importance of the mTOR pathway in disease progression and metastasis progression.

Considering our results, we suggest to perform in situ mTOR activity profiling before administering mTOR inhibitor therapy for osteosarcoma patients. The most widespread mTOR pathway inhibitors are the mTORC1 inhibitors of the rapalog family. Our results suggest low mTORC1 activity in osteosarcoma, which could stand behind the failure of rapalog therapy in this malignancy.

The mTOR pathway plays a key role in the altered metabolic processes of malignant cells (27, 28). The metabolic profiles of osteosarcoma cells were investigated by characterizing enzymes that play a role in glycolysis (PFK and LDHA), oxidative phosphorylation (ATPB), pentose-phosphate pathway (G6PDH), glutaminolysis (GLS), fatty acid beta oxidation (FASN) and long chain fatty acid synthesis (CPT1A). Malignant cells generate huge amounts of energy for their survival by anaerobic glycolysis (29). By suppressing this process in osteosarcoma cells in vivo with the combination of glycolysis inhibitor 2-deoxyglucose (2DG) and adriamycin, a potent antitumor effect was observed (30). We also detected elevated LDHA levels in most of our samples, and moderate expression of ATPB. These findings suggest the hyperactivity of Warburg effect -related enzymes and the lower importance of oxidative phosphorylation in osteosarcoma tissue in vivo. Targeting the glycolytic process has been demonstrated to be effective for controlling growth and enhancing anticancer therapies (31, 32).

Enhanced pentose-phosphate pathway activity of numerous malignancies has also been reported (33, 34). In a recent study, osteosarcoma cell growth inhibition was observed by suppressing the level of G6PDH (35). Elevated G6PDH levels in our samples underline the importance of the pentose-phosphate pathway in the nucleotide biosynthesis of osteosarcoma cells.

The importance of glutaminolysis has also been widely investigated in malignant diseases (36). In metastatic cases, cells rely on glutamine as a nutrient for proliferation in vitro and in vivo. In a recent study, inhibition of glutamine anaplerosis with GLS-1 inhibitor CB-839 reduced the growth of primary osteosarcoma progression and metastasis progression in mouse (37). Our results also suggest a high glutamine demand in progressive osteosarcoma cases, as the H-score levels were significantly higher in those patients who died due to relapse or progression.

Some tumors are dependent on lipids for their progression and survival (38). The characterization of lipid metabolism was also performed in this study. According to our results, high expression of CPT1A suggests that fatty acid beta oxidation may play an important role in osteosarcoma metabolism. Pharmacological inhibition of CPT1A with etomoxir was mostly used in research, but some experiments also indicated promising anti-tumor results (39). Previous experimental evidence has suggested that fatty acid synthase (FASN) may be involved in cancer metastasis, and have been investigated as a therapeutic target (40, 41). The role of FASN in progression and metastasis development in osteosarcoma was demonstrated in previous studies (42). In our study, the expression of FASN was significantly higher in metastatic cases and in patients who died, than those without metastasis and who survived, showing similar results to previous studies (43). This result suggests that FASN might play a role in metastasis development in osteosarcoma.

We also investigated the correlations between metabolic markers and mTOR activity. We found positive correlation between pmTOR and GLS, PFK, and ATPB, suggesting that activated mTOR enhances the activity of certain metabolic pathways as a master regulator. Similar findings were described in pediatric rhabdomyosarcoma as well (44).

Despite the low number of cases available for our studies, our results emphasize the importance of metabolic alterations of osteosarcoma cells. Further investigation of metabolic plasticity may lead to better understanding of therapy resistance and new ways of targeted therapies. Our results indicate potential targets for therapy, but it is still unknown whether it would result in a clinically significant outcome.

In conclusion—based on our results—we suggest performing detailed mTOR profile characterization before administering mTOR inhibitor therapy for osteosarcoma patients. mTOR activation is linked to several metabolic pathways, such as enhanced glycolysis, oxidative phosphorylation, and glutamine demand. These findings may be an explanation of therapeutic failure in this malignancy. Overexpression of GLS and FASN may be a prognostic factor in survival.