mTOR Inhibitors in the Treatment of Breast Cancer
mTOR Inhibitors in the Treatment of Breast Cancer
The phosphatidylinositol 3-kinase/mammalian target of rapamycin (PI3K/mTOR) pathway is commonly dysregulated in breast cancer. In preclinical studies, hyperactivation of the PI3K pathway has been linked to resistance to both endocrine therapy and trastuzumab (Herceptin). Rapalogs, agents that primarily inhibit mTOR-raptor complex 1, have been studied in combination with endocrine therapy to overcome endocrine resistance. Trials of combination endocrine therapy and rapalogs in metastatic hormone receptor–positive breast cancer have demonstrated variable results. However, two independent trials have recently shown that combination everolimus (Afinitor) and tamoxifen or combination everolimus and exemestane (Aromasin) is more effective than either endocrine agent alone. These trials selected patients with cancer refractory to endocrine therapy, which may be important in sensitizing tumors to inhibition of this pathway. In human epidermal growth factor receptor 2 (HER2)-positive breast cancer, the early clinical data with combinations of PI3K/mTOR inhibitors and anti-HER2 therapies are encouraging. Efforts to identify clinical biomarkers of response or resistance to mTOR inhibitors are ongoing. This review will summarize results of preclinical and clinical studies as well as ongoing clinical trials with mTOR or dual PI3K/mTOR inhibitors.
The mTOR pathway
The phosphatidylinositol 3-kinase (PI3K)/Akt pathway, of which mammalian target of rapamycin (mTOR) protein is an important component, is commonly dysregulated in cancer. TOR protein, a highly conserved serine/ threonine protein kinase, was first identified in 1991 through yeast studies examining the mechanism of rapamycin. Complex regulatory mechanisms of the mTOR signaling pathway have been elucidated. These mechanisms have been important in the development of mTOR inhibitors for treatment of cancer and also in identifying predictors of response or resistance.
mTOR controls various cellular processes, including growth, survival, and autophagy. It receives input from upstream growth factor receptors, such as the PI3K pathway, and senses nutrient availability; its central role is in integrating these signals and altering cellular processes. Based on nutrient availability, the mTOR pathway can either promote cell growth or it can inhibit growth during the nutrient deprivation state. Autophagy, a catabolic process involved in eradicating damaged cellular material, is under the delicate control of the mTOR pathway. It is utilized as an adaptive rescue mechanism for starving cells to conserve energy and is highly dependent on nutrient availability.
The mTOR pathway also receives input from the adenosine monophosphate–activated protein kinase (AMPK) pathway. The AMPK pathway senses cellular energy and negatively regulates the mTOR pathway through the tuberin (TSC1)/hamartin (TSC2) complex. When energy stores are low, AMPK and TSC2 are activated, thereby inhibiting the mTOR pathway. An additional negative regulator of mTOR is phosphatase and tensin homolog (PTEN), which also tightly regulates the PI3K pathway. There are two distinct complexes of mTOR—mTORC1 and mTORC2—which have independent regulatory mechanisms and exert their cellular growth effects through different downstream targets. Activated mTOR-raptor complex 1 (mTORC1) results in enhanced protein synthesis and also inhibits PI3K signaling. Activated mTOR-rictor complex 2 (mTORC2) promotes cell survival.
PI3K/mTOR signaling in breast cancer
Activating PI3K mutations are frequent in human cancers and have been identified as oncogenic, making this pathway an attractive therapeutic target in cancer. These mutations can occur in any component of the PI3K pathway, resulting in its dysregulation; a number of mechanisms, including mutations, methylation, and loss of heterozygosity, may be involved. PIK3CA (p110 catalytic subunit alpha) mutations have been identified as a common occurrence in breast cancer, with a higher frequency in the estrogen receptor (ER)-positive and human epidermal growth factor receptor 2 (HER2)-positive subtypes than in triple-negative breast cancer (TNBC). Studies have confirmed that the PIK3CA gene is among the most highly mutated genes in breast cancer: mutations occur at a frequency of 27% to 36%.[8,9] One such study evaluated the mutational spectrum of PIK3CA by breast cancer subtype, determined by gene expression profiling.[10,11] The PIK3CA somatic mutation spectrum differed both by the frequency of mutation and by the type of PIK3CA mutation seen in each subtype. The luminal A subtype of breast cancer had the highest frequency of PIK3CA mutation (45%), and the basal subtype had the lowest (9%). These data are consistent with the results of prior studies, as luminal A and basal-like subtypes roughly correspond to ER-positive and triple-negative breast cancer by immunohistochemistry (IHC), respectively. Even though PIK3CA mutations are oncogenic, they are a good prognostic factor and are associated with improved survival. This is important to consider when assessing patient survival in trials in patients with PIK3CA mutations.
Additional PI3K pathway alterations in breast cancer include Akt and PTEN mutations, or loss of PTEN protein.[7,13] Activation of the PI3K pathway in breast cancer can occur via a PI3K pathway component aberration or through activation of another crosstalk pathway. Beyond identifying PI3K pathway mutations for understanding breast cancer biology, there are important considerations when this information is used for patient selection for treatment. The specific PI3K mutations and the altered components of the PI3K pathway may both impact treatment response.
Development of mTOR inhibitors in cancer
Rapamycin, a macrolide, was first isolated from a soil sample on Easter Island (Rapa Nui) in 1975, and was shown to have antifungal properties. It was initially used clinically as an immune suppressant to prevent allograft rejection in renal transplant patients. Sirolimus (Rapamune), a rapamycin analog (rapalog), has been shown to inhibit the growth of cancer cell lines and xenografts from different tumor subtypes.[15,16] The first generation of mTOR inhibitors target mTORC1, but they do not bind to mTORC2, which is mostly considered to be rapamycin-insensitive. However, there are limited data that rapamycin reduces mTORC2 levels and inhibits Akt activation. Targeting only mTORC1 with rapalogs leads to increased signaling through upstream receptor tyrosine kinases and increased Akt activation, which promotes cell survival. It has been speculated that rapalogs have had limited clinical activity in cancer due to this mechanism, as well as activation of parallel signaling pathways. This limitation of rapalogs has fueled development of alternate methods of targeting the PI3K signaling pathway, with either adenosine triphosphate (ATP)-competitive mTOR inhibitors that target both mTORC1 and mTORC2, or by using dual PI3K/mTOR inhibitors. Several mTORC1 inhibitors are in clinical trials for various tumor subtypes, including everolimus (Afinitor), temsirolimus (Torisel), and ridaforolimus (AP23573). Temsirolimus was the first rapalog approved by the US Food and Drug Administration (FDA); it was approved in 2007 for the treatment of advanced renal cell cancer.
In breast cancer, the majority of studies have exploited the use of mTORC1 inhibitors in ER-positive or HER2-positive breast cancers, primarily to reverse treatment resistance. The focus of this review will be these preclinical and clinical studies by breast cancer subtype. We will also discuss ongoing breast cancer clinical studies using ATP-competitive mTOR inhibitors, which target mTORC1/mTORC2, and dual PI3K/mTOR inhibitors.
Hormone Receptor–Positive Breast Cancer
Preclinical studies, using hormone receptor (HR)-positive cell lines, have demonstrated activation of the PI3K/mTOR pathway after long-term estrogen deprivation.[18,19] Based on these studies, it appeared that estrogen-deprived cells relied heavily on the PI3K signaling pathway, making this an important mechanism of acquired endocrine resistance. This suggested that priming of the PI3K pathway with anti-hormonal treatment might be important in sensitizing these cells to PI3K/mTOR inhibitors. A natural next step was to use combination therapy, simultaneously targeting both the ER and PI3K pathways. Early combination studies showed that rapalogs were synergistic with anti-estrogens, including tamoxifen and letrozole (Femara); blocking both pathways not only enhanced antitumor activity but also reversed endocrine therapy resistance related to PI3K signaling.[20-22] Moreover, high Akt activity has also been shown to contribute to resistance to endocrine therapy, and this also can be reversed by rapalogs.[20,22]
Metastatic setting. Almost all patients with HR-positive breast cancer treated with endocrine therapy develop tumor resistance to treatment. Preclinical data, as described earlier, implicate the PI3K/mTOR pathway in acquired resistance to endocrine therapy, and synergistic preclinical anti-tumor activity has been seen with the combination of rapalogs and anti-estrogens. Based on this biological rationale, clinical trials have combined mTORC1 inhibitors and endocrine therapy in HR-positive breast cancer. Initial studies with temsirolimus and everolimus as single agents in the metastatic setting demonstrated response rates of 9% to 12%.[24,25] Another study with temsirolimus alone was limited to HR-positive or HER2-positive metastatic breast cancer, to enrich it for PIK3CA mutations. Clinical activity was again limited. Primary tumors from this study were analyzed for PIK3CA mutations and PTEN expression by IHC, but no association was seen with clinical response. A limitation of this study is that the PIK3CA mutation status of primary tumors was analyzed, as opposed to the metastatic site, which can be discordant.
The next approach was to combine anti-estrogens and mTORC1 inhibitors in clinical trials. A randomized phase II study of HR-positive metastatic breast cancer tested combination letrozole and temsirolimus vs letrozole alone and found that patients who received combination therapy had superior median progression-free survival (PFS) (13.2 vs 11.6 months). However, the clinical benefit rate (CBR) and the objective response rate (ORR) for patients in the combination arm were not significantly different from the rates in patients who received letrozole alone. Given these somewhat encouraging results, a large randomized phase III trial (N = 1112) was conducted in postmenopausal women with metastatic disease, with letrozole either alone or in combination with temsirolimus as first-line endocrine therapy. The trial was terminated early due to lack of benefit. It has been speculated that this trial failed since it limited the use of mTOR inhibition in combination with endocrine therapy to the first-line metastatic setting. Given lack of prior hormonal therapy exposure, the tumors might not have been dependent on the PI3K/mTOR pathway, thereby remaining insensitive to mTOR pathway inhibition. This highlights the need for identification and selection of patients, whose tumors are dependent on PI3K pathway activation.
The Tamoxifen Plus Everolimus (TAMRAD) study (N = 111) randomized patients with prior exposure to an aromatase inhibitor (AI) in the metastatic setting, to tamoxifen alone versus combination tamoxifen and everolimus. This study demonstrated improvement in CBR (42% vs 61%; P = .045), the primary endpoint, and in time to progression (TTP) (4.5 vs 8.6 months; hazard ratio [HR] = 0.54; 95% confidence interval [CI], 0.36–0.81; P = .002) favoring the combination treatment. This supports that prior endocrine therapy resulting in priming of the PI3K/mTOR pathway may allow for meaningful synergy through attempts to overcome acquired endocrine resistance. In an exploratory analysis, patients were stratified based on primary hormone resistance, defined as relapse during adjuvant AI therapy or progression within 6 months of AI treatment in the metastatic setting, or secondary hormone resistance, defined as late relapse or progression on an AI in the metastatic setting more than 6 months after treatment. A higher CBR (48% vs 74% [secondary]; 36% vs 46% [primary]) and increased TTP (5.5 vs 14.8 months; HR = 0.46; 95% CI, 0.26–0.83; P = .009 [secondary]; 3.8 vs 5.4 months; HR = 0.70; 95% CI, 0.40–1.21; P = nonsignificant [primary]) was predominantly observed in patients with secondary hormone resistance in the everolimus arm.
A phase III trial, Breast Cancer Trials of Oral Everolimus-2 (BOLERO-2), enrolled 724 patients with HR-positive advanced breast cancer to assess the efficacy of everolimus (at a dose of 10 mg per day) and exemestane (Aromasin), in patients with disease refractory to nonsteroidal AIs, including letrozole or anastrozole (Arimidex). An improved median PFS was observed by both local and central assessment with the combination of exemestane and everolimus (2.8 vs 6.9 months; HR = 0.43; 95% CI, 0.35–0.54; P < .001 [local]; 4.1 vs 10.6 months; HR = 0.36; 95% CI, 0.27–0.47; P < .001 [central]). Overall survival results have not been reported. A total of 23% of patients receiving everolimus had serious adverse events compared with 12% receiving placebo, resulting in everolimus discontinuation in 19% of the combination group vs 4% in the placebo group. The most common grade 3 or 4 events with combination therapy were stomatitis, anemia, hyperglycemia, dyspnea, fatigue, and pneumonitis. Notably, there were seven deaths attributed to adverse events (1%) in the everolimus arm; these were due to sepsis, pneumonia, tumor hemorrhage, cerebrovascular incident, renal failure, and suicide.
Forthcoming trials are assessing the role of mTORC1 and dual PI3K/mTOR inhibitors in combination with other anti-hormonal therapies and even chemotherapies, in various lines of metastatic disease (Table).
In summary, trials of combination endocrine therapy and mTORC1 inhibitors in metastatic HR-positive breast cancer have demonstrated variable results. Single-agent temsirolimus or everolimus has limited clinical activity in metastastic breast cancer. A large study that combined temsirolimus with letrozole vs letrozole alone in first-line hormonal therapy for metastatic disease found no benefit from the combination. Two trials have found that combination everolimus and tamoxifen (TAMRAD study) or combination everolimus and exemestane (BOLERO-2) is more effective than either endocrine agent alone. The variability among the reported studies may be related to patient selection, prior endocrine therapy exposure, and the specific drug combination being tested. It is noteworthy that both of the positive studies selected patients who were previously exposed to endocrine therapy in the metastatic setting. Thus, prior endocrine therapy exposure may be an important factor in priming the PI3K/mTOR pathway and thereby sensitizing the tumors to inhibition of this pathway.
Adjuvant/neoadjuvant setting. Everolimus was studied as a single agent in a neoadjuvant trial and was associated with a significant reduction in Ki67 after 14 days of therapy. A neoadjuvant, randomized study in postmenopausal women (N = 270) with ER-positive breast cancer compared letrozole and everolimus vs letrozole and placebo. There was an improved clinical response rate and decreased proliferation in the everolimus-plus-letrozole arm compared with letrozole alone. Response was seen in both wild-type and mutant PI3K tumors. In addition, a reduction in phospho-S6, a pharmacodynamic marker, was noted in post-treatment biopsies in the everolimus-containing arm, signifying that mTOR was being inhibited at the dose used.
In the adjuvant setting, a phase III randomized trial is evaluating the role of combining everolimus with standard adjuvant endocrine therapy, for women with high-risk breast cancer (Table).
Based upon the available data, there is no standard role for adjuvant or neoadjuvant use of mTOR or dual PI3K/mTOR inhibitors in combination with endocrine therapy or chemotherapy.