The Oncologist, Vol. 12, No. 8, 1007-1018, August 2007; doi:10.1634/theoncologist.12-8-1007 © 2007 AlphaMed Press
The Biology Behind mTOR Inhibition in SarcomaMolecular Oncology Section, Pediatric Oncology Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA Key Words. mTOR • Rapamycin • Sarcoma • Clinical trial Correspondence: Lee J. Helman, M.D., Molecular Oncology Section, Pediatric Oncology Branch, Building 10, Room CRC-1W-3816, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892-1928, USA. Telephone:301-496-4257; Fax:301-451-7010; e-mail: helmanl{at}nih.gov Received May 4, 2007; accepted for publication May 8, 2007.
Access and take the CME test online and receive 1 AMA PRA Category 1 CreditTM at CME.TheOncologist.com
Dysregulation of the mammalian target of rapamycin (mTOR) pathway has been found in many human tumors and implicated in the promotion of cancer cell growth and survival. Hence, the mTOR pathway is considered an important target for anticancer drug development. Currently, the mTOR inhibitor rapamycin and its derivatives CCI-779, RAD001, and AP23573 are being evaluated in cancer clinical trials. To date, clinical results have shown good tolerability of treatment with mTOR inhibitors in most reports and varying effectiveness of mTOR inhibitors in a variety of tumors in a subset of patients. For the targeted treatment of sarcomas, AP23573 has shown promising clinical efficacy and low toxicity profiles in patients. Further studies should define the optimal dose/schedule, patient selection, and combination strategies with other biological agents, especially those targeting signaling pathways crucial for cell survival. Disclosure of potential conflicts of interest is found at the end of this article.
Sarcomas are rare malignancies of mesenchymal origin that are generally categorized as soft tissue or bone sarcomas. Sarcomas are more common in pediatric cancers (15%–20%), whereas only 1% of adult cancers are sarcomas. Sarcomas have distinctive biological characteristics including a high incidence of aggressive local behavior and a predilection for metastasis. Several sarcomas, such as alveolar rhabdomyosarcoma (RMS), Ewing's sarcoma, and synovial sarcoma, tend to occur in younger patients and are characterized by tumor-specific chromosomal translocations that are being incorporated as diagnostic criteria, whereas other sarcomas, such as leiomyosarcoma and malignant fibrous histiocytoma, occur more frequently in older adults and are characterized by chaotic karyotypes accompanied by frequent chromosome copy number changes [1]. Most sarcomas have abnormalities in the retinoblastoma, p53, and/or specific growth-factor signaling pathways. The insulin-like growth factor 1 receptor (IGF-1R) pathway is the most commonly activated pathway in a variety of sarcomas. Activating mutations in growth factor receptors and/or autocrine/paracrine mechanisms lead to activation of the phosphatidylinositol 3' kinase (PI3K)–Akt–mammalian target of rapamycin (mTOR) pathway and of the mitogen-activated protein kinase pathway [1]. During the past few decades, the development of new therapeutic regimens for sarcomas has resulted in longer survival for patients with bone and soft tissue sarcomas. The overall survival rate for sarcomas averages 60%–70%, with differences depending on the sarcoma type, histotype, grade of malignancy, and stage of disease [2–5]. Patients with detectable metastases not responding to therapy or with disease relapse have a significantly poorer prognosis. Thus, developing new means of assessing prognosis, overcoming resistance of tumor cells to anticancer therapy, and preventing or treating metastasis are important goals in the treatment of patients with sarcomas. Increasing knowledge of the signal transduction pathways for growth factors has led to speculation that they could offer novel targets for cancer therapy. The topic of this review is the mTOR signaling pathway, which is aberrantly activated in many human cancers and has a central role in the regulation of cancer cell growth by control of the initiation of mRNA translation into protein. We discuss the potential targeting of this pathway together with the preclinical and clinical data on sarcomas.
mTOR, also referred to as sirolimus effector protein, FK506-binding protein (FKBP12)-rapamycin associated protein, rapamycin and FKBP12 target, rapamycin-associated protein, or rapamycin target, is a 289-kD serine/threonine kinase orthologue of target of rapamycin 1 (TOR1) and TOR2 in Saccharamyces cerevisiae [6–10]. mTOR lies at the interface of two major signaling pathways, one initiated by PI3K and the other by an energy-sensing pathway through serine threonine kinase 11 (also called LKB1) (Fig. 1). Growth factor stimulation primarily regulates mTOR signaling through PI3K/Akt. PI3K, which is a heterodimer consisting of the p85 regulatory and p110 catalytic subunits, is a major signaling component downstream of growth factor receptor tyrosine kinases (RTKs) [11]. Specific phosphorylated tyrosine residues of the RTK or associated adaptor proteins, such as those of the insulin receptor substrate (IRS) family, interact with the Src-homology 2 domain of p85 and recruit the enzyme to the membrane. PI3K phosphorylates phosphatidylinositol-4,5-bis-phosphate (PIP2) and thereby converts PIP2 to PIP3 at the cell membrane [12]. Subsequently, PIP3 recruits Akt and 3-phosphoinositide-dependent protein kinase 1 (PDK1) to the plasma membrane through direct interaction with their pleckstrin homology domains, resulting in partial activation of Akt through phosphorylation at threonine 308 in its activation loop by PDK1 [13, 14]. Full activation of Akt requires its additional phosphorylation at serine 473 in the C terminus [15]. Phosphatase and tensin homologue deleted on chromosome ten (PTEN) dephosphorylates PIP3 back to PIP2 and thus shuts off PI3K signaling. Recently, important progress was made in identification of the tumor suppressors tuberous sclerosis complex 1 (TSC1) and TSC2 and their function in the regulation of mTOR signaling, which provides an evolutionarily conserved link between the growth factor-regulated PI3K–Akt and the nutrient-sensitive mTOR pathways [16]. The mechanism by which Akt activates mTOR appears to be through direct phosphorylation and inhibition of TSC2 [17, 18]. TSC1 and TSC2 form a heterodimer that inhibits the activity of Ras homologue enriched in brain (Rheb), a small GTPase required for mTOR activation [19]. In a parallel pathway, the energy-sensing pathway (i.e., through amino acids and ATP) is linked to mTOR signaling through LKB1, a tumor suppressor inactivated in Peutz-Jeghers syndrome. LKB1 activates AMP-activated kinase, which in turn activates TSC1/2, leading to mTOR inhibition [20].
Recent investigations have found that mTOR exists in two distinct complexes, mTOR complex 1 (mTORC1) and mTORC2, with each complex containing at least three distinctive proteins [21–23]. mTORC1 contains mTOR, regulatory-associated protein of mTOR (raptor), and G protein β subunit-like (Gβ L, also known as mLST8), whereas mTORC2 consists of mTOR, Gβ L, and rictor. The functions of these complexes are strikingly distinct, in that the raptor–mTOR complex is rapamycin sensitive, whereas the rictor–mTOR complex is rapamycin resistant [24]. Furthermore, it has been shown that the rictor–mTOR complex is actually the elusive PDK2, which phosphorylates Akt at the Ser473 position, leading to full activation of Akt and organization of the cytoskeletal structure [24]. However, the upstream signals regulating the rictor–mTOR complex remain poorly understood. Activation of the raptor–mTOR complex regulates protein synthesis, cell growth, and proliferation, a process whereby cells accumulate mass and increase in cell size and number, via at least two characterized downstream targets, eukaryotic initiation factor 4E (elF-4E) binding protein-1 (4E-BP1) and protein S6 kinase 1 (S6K1) [25]. 4E-BP1 acts as a translational repressor, binding and inhibiting elF-4E, the mRNA cap-binding protein. Activation of mTOR signaling causes 4E-BP1 to became hyperphosphorylated and to dissociate from elF-4E. Free elF-4E then binds to the scaffold protein elF-4G, the ATP-dependent RNA helicase elF-4A, and elF-4B and drives the translation of 5' cap mRNAs, including several oncogenic proteins such as c-Myc, hypoxia-inducible factor 1 (HIF-1 ), vascular endothelial growth factor (VEGF), and cyclin D1 [26, 27]. mTOR-dependent phosphorylation of S6K1 on Thr389 leads to the phosphorylation of multiple downstream targets relevant to mRNA translation, including ribosomal protein S6, elF-4B, and eukaryotic elongation factor (eEF) 2 protein kinase [27–29].
The mTOR signaling pathway is abnormally activated in many human tumors that have multiple alterations both upstream and downstream of mTOR, leading to pathway activation (Fig. 1). In the following sections, we discuss the mechanisms and cell-intrinsic processes, including both mutation and amplification of oncogenes and loss of tumor suppressor genes, that result in dysregulation of mTOR signaling in human cancer as specified in Table 1; some specific data on sarcomas are mentioned where possible.
The mTOR signaling pathway can be activated via exogenous oncogenes, including overexpressed or mutated RTKs such as IGF-1R, members of the fibroblast growth factor receptor (FGFR) family, and members of the epidermal growth factor receptor (EGFR)/ErbB family, which are frequently activated in malignances [30–33]. Considering sarcomas, overexpression and activation of IGF-1R, FGFR, and EGFR have been found in different types of sarcomas [34–38]. The PI3K–Akt pathway is dysregulated by a variety of mechanisms in a large fraction of human tumors. For example, mutations of PI3K, as well as amplification and or overexpression of PI3K and Akt have been reported in various tumors [39, 40]. Aberrant activation of the TSC–Rheb–mTOR axis through loss of tumor suppressor protein function has been linked to the pathobiology of some tumor predisposition syndromes, such as Cowden's syndrome (PTEN), Peutz-Jeghers syndrome (LKB1), and tuberous sclerosis (TSC1/2) [41]. In addition to genetic syndrome–associated mutation, loss of PTEN function via gene mutation, deletion, or promoter methylation has been found in many cancers [42]. A critical outcome of PTEN inactivation is constitutive activation of PI3K/Akt, resulting in the activation of the mTOR pathway. Furthermore, tumors associated with PTEN inactivation are particularly susceptible to the therapeutic effects of mTOR inhibitors [43, 44]. Loss of LKB1 protein leads to hyperactivation of mTOR signaling [45], which is similar to the effect of loss of PTEN on the growth factor–mediated pathway. Numerous downstream targets of mTOR are known to be activated and overexpressed in many cancers and are involved in the transformation process as well in drug resistance. The overexpression and/or amplification of S6K1 or elF-4E has been associated with oncogenesis [46]. A link between translation initiation and oncogenesis first became evident in 1990, when overexpression of elF-4E was shown to cause cellular transformation and tumorigenesis in rodent fibroblasts and NIH3T3 cells [47]. Subsequently, elF-4E overexpression has been found in a large number of cancers and its overexpression has routinely been linked to progression of those cancers, including tumor recurrence, metastasis, and poor prognosis [26, 48]. In contrast to elF-4E, the levels of 4E-BP1 expression correlate inversely with tumor progression [49]. Overexpression of 4E-BP1 has been shown to reduce the elF-4E overexpression–mediated tumorigenic effect [50]. Furthermore, overexpression of wild-type 4E-BP1 resulted in a decrease in tumor size [50]. Moreover, expression of 4E-BP1 displayed a proapoptotic effect [51, 52]. Recently, our proteomic network analysis of microdissected human RMS specimens, which were obtained from the Children's Oncology Group Intergroup Rhabdomyosarcoma Study IV, D9502, and D9803, with a 12-year follow-up, demonstrated that high levels of Akt Ser473, 4E-BP1 Thr37/46, elF4G Ser1108, and S6K1 Thr389 (Fig. 2A) and a low level of 4E-BP1 expression (Fig. 2B) in stage III RMS patients were associated with poor overall survival and poor disease-free survival [53]. These data also revealed a statistically significant association between patient survival and status of Akt–mTOR pathway activation (Fig. 2A and B). Taken together, the PI3K–Akt–mTOR signaling pathway is intimately implicated in cancer development and progression, inasmuch as many of its components are mutated or overexpressed in human cancer. These alternations, both upstream and downstream of mTOR, lead to dysregulation of the mTOR pathway.
As a result of this association of mTOR activation with human cancer progression, the mTOR pathway is considered an important target for cancer drug development. A significant effort has been made in recent years to define and develop specific inhibition of this pathway. Rapamycin (sirolimus) is a natural product isolated from a strain of Streptomyces hygroscopicus found in the soil of the Vai Atore region of Easter Island (Rapa Nui), for which rapamycin was named. Rapamycin is a macrolide antibiotic initially discovered as an antifungal agent in the mid-1970s [54]. Rapamycin was approved by the U.S. Food and Drug Administration as an immunosuppressant for use after renal transplantation in the 1990s. The potential for rapamycin to act as an antitumor agent was evaluated by the National Cancer Institute screening program, which demonstrated potent inhibitory activity against several human cancers in vitro and in vivo. In essence, rapamycin forms a complex with the intracellular immunophilin FKBP12 and the resultant complex binds to the FK-rapamycin binding domain of mTOR, in turn leading to inhibition of the function of mTOR in mTORC1 and the mTORC1-mediated signal pathway, thereby preventing phosphorylation of S6K1 and 4E-BP1. However, rapamycin's poor aqueous solubility and chemical stability limited its clinical development as an anticancer agent, which led to efforts directed at synthesized analogues with more favorable pharmaceutical characteristics. The most notable of the rapamycin analogues currently in preclinical and clinical development as anticancer agents include CCI-779 (temsirolimus), RAD001 (everolimus), and AP23573 (Table 2). Rapamycin and its analogues in clinical development are highly specific inhibitors of mTOR, and differ only slightly in chemical structure (Table 2) with superior chemical stability and pharmaceutical properties.
Data from preclinical studies have shown that rapamycin and its analogues inhibit cell growth in a wide variety of human cancer cell lines derived from RMS, osteosarcoma, neuroblastoma, breast cancer, glioblastoma, small-cell lung cancer, Ewing sarcoma, pancreatic cancer, leukemia, and prostate cancer, and tumor models in vivo [56]. Furthermore, recent studies have demonstrated that rapamycins have antiangiogenic effects by inhibition of HIF-1 translation as well as by intercepting VEGF/VEGFR. Rapamycin has been shown to inhibit the expression of HIF-1 and VEGF in vitro [56, 57] and to inhibit metastatic tumor growth and angiogenesis in in vivo mouse models by reducing translational production of VEGF, which leads to the inhibition of endothelial cell proliferation and survival [58]. Our recent study demonstrated that CCI-779 inhibited human RMS xenograft growth and its antitumor activity was associated with an antiangiogenic mechanism linked to targeting mTOR–HIF-1 –VEGF signaling [59]. Moreover, treatment with rapamycin and CCI-779 led to a significant inhibition of ezrin-mediated metastasis and prolongation of survival in a murine osteosarcoma model [60]. These results suggest that the effect of rapamycin and its analogues on antiangiogenesis is important for inhibition of both primary and metastatic tumor growth.
Three rapamycin analogues, CCI-779, RAD001, and AP23573, are currently being evaluated in clinical trials for cancer therapy (Table 3). The phase I trials evaluating the pharmacokinetics and biological effects of CCI-779 were tested using both daily for 5 days (0.75–19.1 mg/m2 per day) every other week and once weekly (7.5–220 mg/m2 per week) i.v. treatment in patients with different solid tumors such as renal, breast, and lung cancers [61]. Overall, the main dose-limiting, mild, and reversible toxicities of CCI-779 treatment were mucositis and skin reaction for once-weekly treatment [62] or grade 3 toxicities including hypocalcemia, vomiting, thrombocytopenia, and elevated hepatic transaminases for daily treatment [61]. Overall, during the phase I trials, major tumor responses were observed in patients with lung, renal, and breast carcinomas and neuroendocrine tumors, whereas minor responses (MRs) and stable disease (SD) were observed in soft tissue sarcoma and uterine and cervical carcinoma. Of note, the maximum-tolerated dose was not reached (signifying a high therapeutic index) and no immunosuppressive effects were observed. Thus, results from phase I studies of CCI-779 suggest that the agent is well tolerated and has antitumor activity.
On the basis of the promising data seen in the phase I study, several phase II trials were conducted to evaluate the safety and efficacy of CCI-779 in advanced stage refractory renal cell carcinoma (RCC) [63], refractory metastatic breast cancer [64], refractory mantle cell lymphoma [65], and recurrent glioblastoma [66] (Table 3). Recently, a phase III randomized trial comparing CCI-779 with interferon- or a combination of both using time to progression as the primary endpoint was reported. Six hundred twenty-six patients with advanced and metastatic RCC treated with CCI-779 displayed a statistically significant longer median survival time (10.9 months) than patients receiving interferon- alone (7.3 months) [67]. The phase I study with orally administered RAD001 in patients with solid tumors has shown that weekly doses of 5–30 mg are generally well tolerated, with effects on tumors similar to those observed for CCI-779 (Table 3). Side effects included grade 1 or 2 fatigue, anorexia, rash, headache, mucositis, or hyperlipidemia [68]. RAD001 has also been tested in a phase II study in 25 patients with advanced and metastatic RCC. Response was observed in 33% of RAD001-treated patients [69]. Two phase I studies were conducted to establish the safety profile of and to determine the best dosing strategy for AP23573 in patients with refractory or advanced solid tumors, which were based on two dose-escalation trials, with AP23573 administered either daily for 5 days every 2 weeks or once weekly i.v. [70–73]. When administered weekly, side effects were generally mild to moderate and reversible, with the dose-limiting toxicity being oral mucositis at the 100-mg dose level. As with the weekly schedule, the dose-limiting toxicity with daily treatment was mucositis at the 28-mg dose (140 mg total dose). The recommended phase II dose is 18.75 mg daily for 5 days. Both dosing strategies provided therapeutic AP23573 concentrations to patients. As noted, of 27 patients evaluable for response, 11% had a partial response (PR) while 59% had either a PR, an MR, or SD. With respect to tumor type, 100% of the sarcoma patients and 100% of the RCC patients experienced PR, MR, or SD [71]. The results of these phase I studies therefore support further evaluation of AP23573 in phase II trials, particularly in patients with sarcoma.
A phase II study of AP23573 administered i.v. daily for 5 days every 2 weeks in patients with advanced sarcoma was recently reported [74]. In that study, 193 patients were evaluable for clinical benefit response assessment, defined as tumor response by standard criteria or SD for
Although preliminary results in relapsed sarcoma patients are encouraging, objective response rates with rapamycin derivatives have been low in most tumor types. Strategies to improve response rates could include: (a) identification of the patients who are most likely to benefit from mTOR inhibitors (matching the inhibitor to the patient), (b) identification of the tumor types that respond to mTOR inhibitors, (c) identification of optimal surrogate markers for assessing the efficacy of mTOR inhibitors and for predicting sensitivity/resistance, and (d) development of combination regimens. The identification of patients most likely to respond to mTOR inhibitors may be enhanced using analysis of global gene and protein expression profiles, gene mutation analysis, and methylation patterns by DNA determination. Knowledge on the status of PTEN and PI3K–Akt–mTOR pathway activation might help in the selection of patients as well as tumor types that are most likely to respond to mTOR inhibitors. Malignancies that are principally driven by paracrine or autocrine stimulation of receptors (e.g., IGF-1R) that constitutively stimulate the PI3K–Akt–mTOR pathway or tumors with aberrations that activate PI3K–Akt-related elements (e.g., PTEN loss of function) may be especially sensitive to mTOR inhibitors. There is considerable evidence that aberrant stimulation of the PI3K–Akt pathway in cancer cells may increase the dependency of such tumors on mTOR functions and their sensitivity to signal modulation by mTOR inhibition [81]. Recently, our study using proteomic network analysis of microdissected human stage III RMS specimens demonstrated that high phosphorylation levels of Akt Ser473, 4E-BP1 Thr37/46, elF-4G Ser1108, and S6K1 Thr389 in tumors were associated with poor overall survival and poor disease-free survival (Fig. 2). This association of poor outcome with Akt–mTOR activation could identify a subset of patients who might benefit from mTOR inhibitory therapy [53]. The prediction of tumor types that respond to mTOR inhibitors remains an important issue. So far, the most promising results with mTOR inhibitors from clinical trials have been obtained in RCC, mantle cell lymphoma, and sarcoma (Table 3). However, the sensitivity or resistance of a specific tumor to mTOR inhibitors cannot be predicted solely on the basis of histology. Thus, elucidating the active status of the PI3K–Akt–mTOR pathway may be helpful in identifying subgroups within a specific histology potentially sensitive to mTOR inhibitors. For example, loss of PTEN function or activation of Akt, S6K1, or 4E-BP1 have been found to be associated with objective responses, whereas low levels of Akt and S6K1 phosphorylation were associated with resistance to mTOR inhibitors [43]. S6K1, S6 (the physiological downstream target of S6K1), and 4E-BP1 are considered as surrogate markers for assessing the efficacy of mTOR inhibitors. Measurement of the phosphorylation status of S6K1/S6 or 4E-BP1 in the skin or in peripheral blood mononuclear cells, and/or in the tumor itself, may help to monitor the biological effects of mTOR inhibitors and narrow down the biologically active doses in patients. However, inhibition of S6K1 and 4E-BP1 is not sufficient to guarantee sensitivity to mTOR inhibitors. Thus, an important strategy is the development of predictive biomarkers, to allow identification of those patients, and tumors, most likely to respond to mTOR inhibitors. In addition, optimal assessments of mTOR inhibition in the tumor and surrogate tissue may require additional information that is relevant to target inhibition in tumor cell growth, apoptosis, autophagy, and angiogenesis. FDG-PET was successfully used to monitor the glucose uptake of tumors following early treatment with AP23573. It is possible that FDG-PET could serve as an early pharmacodynamic marker. Another important issue is that cancer cells often develop resistance to mTOR inhibition. Recent studies have shown that rapamycin induces feedback activation of Akt in human non-small cell lung cancer and breast cancer cell lines [82, 83]. In our study, we found that treatment with rapamycin or CCI-779 in RMS cell lines as well as RMS xenografts resulted in activation of Akt [84]. Thus, mTOR inhibitor–induced feedback activation of Akt may contribute to resistance developed during treatment, thus attenuating potential antitumor activity. Furthermore, our data in vitro also demonstrated that pretreatment with IGF-1R antibody led to blocking rapamycin-induced feedback activation of Akt and enhancing the effect of rapamycin on the inhibition of RMS cell growth [84]. These results suggest that combining therapy with mTOR inhibitors and IGF/IGF-1R inhibitors may overcome the resistance of some tumors to mTOR inhibitors. A phase II study of RAD001 combined with depot octreotide, a growth hormone inhibitor that has been shown to inhibit IGF-1 production in solid tumors, in patients with advanced neuroendocrine carcinoma has been reported [85]. Of 27 patients, four (15%) achieved PRs and 19 (70%) had SD. Moreover, mTORC2 was identified recently as a hydrophobic motif kinase for Akt phosphorylation on Ser473, which is rapamycin resistant [23, 24]. Thus, combining therapy with mTOR inhibitors and Akt specific inhibitors to block IGF-1R or the IGF-1R/IRS-1–dependent feedback loop may be clinically undesirable.
Recently, limited studies from clinical trials evaluating mTOR inhibitors with interferon-
The authors indicate no potential conflicts of interest.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||