Concurrent Antitumor and Bone-Protective Effects of Everolimus in Osteotropic Breast Cancer

Andrew J. Browne; Marie L. Kubasch; Andy Göbel; Peyman Hadji; David Chen; Martina Rauner; Friedrich Stölzel; Lorenz C. Hofbauer; and Tilman D. Rachner


Breast Cancer Res. 2017;19(92) 

In This Article


Most hormone-ablative breast cancer therapies decrease patients' bone density and integrity by increasing osteoclastogenesis and bone resorption.[2,20] This is especially the case for aromatase inhibitors, which are now standard of care for postmenopausal patients with ER-expressing breast cancer. In the BOLERO-2 trial, mTOR inhibition using everolimus exerted a clinical benefit to patients with ER-positive breast cancer progressing despite hormone therapy,[12] and exploratory assessment of markers of bone turnover from this trial revealed a lower increase in the group receiving everolimus in combination with aromatase inhibitors than aromatase inhibitor treatment alone. However, bone turnover reflects only one aspect of bone biology and does not necessarily correlate with bone density. Researchers in a number of preclinical studies have previously linked mTOR signaling to bone biology;[14] however, assessment of mTOR inhibition in bone biology in the context of osteotropic breast cancer has not been performed in detail.

In this study, low concentrations of everolimus in vitro (1, 10, and 100 nM), when used as a single treatment agent, were sufficient to potently inhibit mTOR signaling and cancer cell viability in vitro (Figure 1a and b). Continuing with the antitumor assessment of everolimus, we tested the everolimus dose of 1 mg/kg/day in mice subcutaneously inoculated with either osteotropic B16-F10 murine melanoma cells[21] or MDA-MB-231 human breast cancer cells.[22] These tumor-bearing mice demonstrated an effective therapeutic response to treatment with everolimus, with significant decreases in the final tumor burdens in both models. First, we used the B16-F10 model to confirm potential antitumor effects of everolimus because it is an aggressive osteotropic cell line with the propensity to form osteolytic bone metastases in vivo and can be inoculated into syngeneic hosts. Putting this into context, everolimus has previously been assessed in clinical trials of patients with metastatic melanoma, often combined with inhibitors of angiogenesis.[23,24] The therapeutic benefit of pairing everolimus in these combinations is unclear, however, with reports ranging from major responses and disease stabilization to no response at all. Obviously, there is room for improvement, and focusing on trial design and everolimus pharmacology in the future may help to elucidate the current controversies.[25] Here, we observed a potent growth-inhibitory effect of everolimus on subcutaneous B16-F10 tumors. This may be linked to possible activating mutations in mTOR,[26] and these mutation-specific influences should also be considered in the context of improving the outcomes of patients with melanoma by using everolimus treatment.

With the focus returning to breast cancer, we could show that 1 mg/kg was effective at inhibiting subcutaneous tumor growth of ER-negative MDA-MB-231 cells. Of interest, this dose is tenfold lower than what is currently advised for ER-positive HER2-negative breast cancer in the clinic. However, this inconsistency could likely be explained by the administration route (intraperitoneal injection versus oral), the chosen animal model, and the cell line used. Previously, when already established subcutaneous MDA-MB-231 tumors were treated with everolimus as a monotherapy at a dose of 10 mg/kg three times weekly, the results were rather disappointing. In this case, only a modest antitumor response could be demonstrated, which was not significant.[27] Observed differences in the potency of everolimus as an antitumor agent may result from differences in the protocol, most important being the time point when treatment was commenced. Despite this, the results presented here do support other studies in which everolimus was used as a sensitizing agent to radiotherapy or chemotherapeutics against the same cell line.[28,29,30] Dose, route, and frequency of administration would all appear to define the potency of everolimus as an antitumor agent in breast cancer. In this context, further investigations should expand on the mechanism of how everolimus inhibits the growth and tumorigenesis of breast cancer cells. As one example, it has previously been shown that mTOR inhibition by everolimus in aromatase inhibitor-resistant breast cancer results in a concurrent induction of autophagy and a downregulation of the ER.[11] Overall, our results confirm and endorse the potent antitumor effects of everolimus in melanoma and in ER-negative breast cancer, in immunocompetent and immunocompromised models, respectively.

Both bone loss observed in hormone ablation and in the progression of breast cancer bone metastases are mediated mainly by an enhanced osteoclast activity. Previously, mTOR signaling has been shown to regulate osteoclastogenesis and osteoclast function,[17,31] and in the pathological setting, the activation of mTOR is responsible for joint destruction in arthritis.[32] In this study, 1 nM of everolimus exerted highly antiosteoclastic effects. Although osteoclasts differentiated from murine bone marrow-derived mononuclear cells appeared slightly more resistant to everolimus, they demonstrated a potent decrease in all parameters investigated at 10 and 100 nM of everolimus treatment (Figure 2c, d). Additionally, it could be shown that everolimus inhibits the ability of mature osteoclasts to resorb bone (Additional file 3: Figure S3). Interestingly, these results are in contrast to a previous study where rapamycin (an earlier-generation mTOR inhibitor) enhanced RANKL-induced osteoclastogenesis and functionality in RAW 264.7 cells.[33] This may be explained by different proficiency and mechanisms of the mTOR inhibitors rapamycin and everolimus.[34,35] For example, different downstream signaling effects with regard to inhibited phosphorylation S6 and AKT have been described.[36] Despite these considerations, a recent study supports our observations of everolimus as being an inhibitor of osteoclastogenesis. When investigating the effects of everolimus on the osteoclastogenesis of peripheral blood monocytes in a coculture model with the triple-negative breast cancer cell line SCP2, everolimus was effective at inhibiting osteoclastogenesis induced by SCP2-derived factors.[37]

Previous studies deleting key components of mTORC1 signaling in mice have shown pro-osteoblast effects of mTORC1 signaling,[38,39] with mTOR activation promoting later stages of osteoblast differentiation in particular.[40] In contrast, studies using the mTOR inhibitors rapamycin and BEZ235 demonstrated an augmented osteoblastogenesis in human embryonic stem cells and hMSC, respectively.[41,42] In the present study, the effects of everolimus on the differentiation of both mMSC and hMSC along the osteoblast lineage were assessed. In general, we have shown that osteoblasts were less sensitive to inhibitory effects by everolimus than osteoclasts, and some markers of osteoblastogenesis even increased when hMSC were differentiated at low doses of everolimus (Figure 3).

When translating these in vitro osteoclast and osteoblast observations to an in vivo setting using an OVX model to mimic hormone ablation therapy and to induce bone loss, a low-dose treatment regime of everolimus (1 mg/kg/day) had potent bone-protective effects as assessed by μCT analysis. Further assessment by bone histomorphometry revealed that these positive effects were achieved primarily by an inhibition of the osteoclast-mediated bone resorption (Figure 4b). Interestingly, whereas the numbers of osteoclasts were reduced in the bone, there was no evidence of increased anabolic activity. This is a common finding in classic antiresorptive drugs such as bisphosphonates.[43] At the effective doses, these drugs exert highly antiresorptive effects but also impair osteoblast action. However, because bone loss mediated by hormone deficiency is mediated mainly by increased osteoclastic bone resorption, the antiresorptive effects are sufficient to protect bone. In line with this assumption, we did not observe an effect of everolimus on bone when used in a nonchallenged setting. In the context of breast cancer, the mTOR signaling pathway has been shown to be involved in crosstalk with ERα signaling under estrogen stimulation.[44] With this in mind, in the absence of ER signaling, mTOR could hypothetically signal independently, promoting and enhancing the differentiation of osteoclasts.

Using the same dose that was required to suppress subcutaneous tumor growth and prevent bone loss in a hormone-deprived environment, everolimus was proven to be effective in preventing the establishment and progression of breast cancer bone metastases when mice were intracardially inoculated with the ER-negative, bone-seeking MDA-MB-231 cells (Figure 5a, Figure 5b and Figure 5c). In line with this finding, the trabecular bone of the femur was protected against tumor-induced osteolysis and bone loss (Figure 5e). Supporting this finding in another model of malignant bone disease prevention, treating MDA-MB-231 cells with everolimus before intratibial injection resulted in a reduction in the area of the subsequent osteolytic lesions.[45] Furthermore, at the first time point measured in the context of bone metastases arising from lung cancer, the bisphosphonate zoledronic acid was shown to be less effective than everolimus in preventing bone metastases, with 27.8% more mice developing bone lesions than in the everolimus-treated group.[46] In the intracardiac bone metastasis model, it was discovered that the number of osteoclasts in the femurs of everolimus-treated mice was significantly reduced (Additional file 5: Figure S5). This could support an additional antiosteoclastic role for mTOR in the context of cancer-induced osteolysis. However, it remains a challenge to separate the relative contributions to antitumor efficacy of everolimus as a bone-targeted agent that potentially simultaneously promotes both direct antitumor and indirect anti-bone-resorptive actions. For example, this reduction in the number of osteoclasts could be a result of an inhibition in the expression of the pro-osteoclastic factor interleukin-6 by the MDA-MB-231 cells.[45] Here, we postulate that the antibone metastatic effect of everolimus is a result of direct antitumor effects, supported by a decrease in the differentiation and activity of osteoclasts. However, further research is needed to dissect the molecular mechanisms of this intertwined effect.