Targeted Therapy for Advanced Prostate Cancer: Inhibition of the PI3K/Akt/mTOR Pathway

Abstract

A large number of novel therapeutics is currently undergoing clinical evaluation for the treatment of prostate cancer, and small molecule signal transduction inhibitors are a promising class of agents. These inhibitors have recently become a standard therapy in renal cell carcinoma and offer significant promise in prostate cancer. Through an understanding of the key pathways involved in prostate cancer progression, a rational drug design can be aimed at the molecules critical to cellular signaling. This may enable administration of selective therapies based on the expression of molecular targets, more appropriately individualizing treatment for prostate cancer patients.

One pathway with a prominent role in prostate cancer is the PI3K/Akt/mTOR pathway. Current estimates suggest that PI3K/Akt/mTOR signaling is upregulated in 30-50% of prostate cancers, often through loss of PTEN. Molecular changes in the PI3K/Akt/mTOR signaling pathway have been demonstrated to differentiate benign from malignant prostatic epithelium and are associated with increasing tumor stage, grade, and risk of biochemical recurrence. Multiple inhibitors of this pathway have been developed and are being assessed in the laboratory and in clinical trials, with much attention focusing on mTOR inhibition. Current clinical trials in prostate cancer are assessing efficacy of mTOR inhibitors in combination with multiple targeted or traditional chemotherapies, including bevacizumab, gefitinib, and docetaxel. Completion of these trials will provide substantial information regarding the importance of this pathway in prostate cancer and the clinical implications of its targeted inhibition. In this article we review the data surrounding PI3K/Akt/mTOR inhibition in prostate cancer and their clinical implications.

INTRODUCTION

Although the majority of the estimated 220,000 men diagnosed annually with prostate cancer do not die of this disease, the prognosis for men with advanced prostate cancer is very poor. Treatment for patients who experience recurrence of their disease after primary treatment or present with advanced disease typically involves androgen deprivation therapy. However, nearly all men on androgen deprivation therapy will progress within 18-24 months to castrate-resistant prostate cancer (CRPC), and no curative treatments currently exist for CRPC. As a result, prostate cancer remains the second leading cause of cancer-related deaths in men in the United States, with approximately 28,000 deaths each year [1]. New rational approaches to treatment of CRPC are needed, and signal transduction modulators offer significant promise. One signaling pathway with substantial therapeutic potential in prostate cancer is the PI3K/Akt/mTOR pathway. The role of this pathway in prostate cancer has been reviewed in the past [2-8], and it is a focus of constant and intensive investigations. The new findings regarding the biological rationale for inhibition of this pathway and the current status of PI3K/Akt/mTOR inhibitors in the treatment of prostate cancer are discussed in this review.

The current standard treatment for patients with CRPC derives from two large, prospective trials, TAX 327 [9] and SWOG 99-16 [10], which have placed docetaxel as the gold standard treatment for these men. In TAX 327 [9], over one-thousand men with CRPC were enrolled in a study of docetaxel versus mitoxantrone, which was considered standard therapy at that time [11, 12]. Men receiving docetaxel every 3 weeks had an increased median survival of 2.4 months over those receiving mitoxantrone (18.9 vs. 16.5 months, p=0.009). SWOG 99-16 [10] prospectively evaluated the effect of docetaxel plus estramustine compared with mitoxantrone and prednisone in men with metastatic CRPC. Patients receiving docetaxel had a significantly longer survival when compared with patients receiving mitoxantrone and prednisone (17.5 vs. 15.6 months, p=0.02), a greater prostate-specific antigen (PSA) decline (50% vs. 27% of patients with >50% reduction, p<0.001), and longer time to progression (6.3 vs. 3.2 months, p<0.001). While these two studies provided a clear standard of care for men with CRPC, they also demonstrated the need for new therapies capable of providing larger and more sustained benefits.

A number of drugs that rely on specific knowledge of cancer genetics and molecular pathways are emerging in prostate cancer. These molecular therapies include angio-genesis inhibitors, nucleotide-based targeted therapies, and small molecule signal transduction inhibitors. What these approaches have in common is a reliance on identification and inhibition of pathways critical to prostate cancer progression at the molecular level. In breast cancer, knowledge of molecular markers has already led to the development of effective, rational-based cancer therapies. A number of studies have demonstrated that targeting the receptor tyrosine kinase Her2/neu, which is overexpressed in approximately 30% of breast cancer, resulted in longer survival [13]. Patients with metastatic breast cancer overexpressing Her2/neu had a significant survival benefit when they received trastuzumab, a monoclonal antibody targeting Her2/neu, in addition to standard chemotherapy (25.1 vs. 20.3 month median survival, p=0.046) [13]. Trastuzumab has also demonstrated significant benefit as an adjuvant agent in women with surgically removed Her2/neu-positive breast cancer (HR 0.48 for recurrence, CI 0.39-0.59) [14]. In order to develop similar targeted therapies for prostate cancer, critical key molecules and signaling pathways need to be identified and the effects of inhibition investigated.

PI3K/Akt/mTOR PATHWAY

The PI3K/Akt/mTOR pathway is implicated in numerous cellular processes ranging from cell growth and survival to the promotion of angiogenesis. The basic outline of the PI3K/Akt/mTOR pathway is depicted in Fig. (1).

Fig. (1).

PIK3/Akt/mTOR Pathway.

A number of receptor tyrosine kinases can activate phosphatydidyl inositol-3-OH kinase (PI3K) at the cell membrane, initiating the signaling cascade. These receptor tyrosine kinases include the epidermal growth factor receptor (EGFR), insulin-like growth factor receptor (IGFR), and platelet derived growth factor receptor (PDGFR). Once activated, PI3K phosphorylates phosphatidylinositol-4,5-diphosphate (PIP2), leading to accumulation of phosphatidylinositol-3,4,5-triphosphate (PIP3) [15]. This lipid second messenger recruits Akt (also known as protein kinase B) and phosphoinositide dependent protein kinase 1 (PDK1) to the cell membrane, where Akt is phosphorylated by PDK1 [16]. Phosphorylated Akt regulates cellular processes by phosphorylation of a number of substrates, including checkpoint kinase 1 (Chk1), murine double minute (MDM2), BclxL/Bcl-2 associated death promoter (BAD), the forkhead box O (FOXO) family of transcription factors, and tuberous sclerosis complex 2 (TSC2) [17]. Most evidence to date, however, points to another Akt substrate, the mammalian target of rapamycin (mTOR), as having the most significant role in tumorigenesis. mTOR is a serine/threonine kinase that plays critical roles in the regulation of cell growth, survival, division, and motility [18]. mTOR acts through two separate complexes—mTORC1 and mTORC2. Each consists of mTOR bound to LST8 and either raptor (regulatory associated protein of mTOR) forming mTORC1 or rictor (rapamycin-insensitive companion of mTOR) forming mTORC2. When activated, mTORC1 increases mRNA translation by phosphorylation of the downstream molecules p70-S6K (S6K) and 4E binding protein 1 (4E-BP1). S6K phosphorylates the S6 component of the 40S ribosomal subunit, increasing translation of mRNA [19]. 4E-BP1 phosphorylation also leads to activation of translation, but by preventing association of 4E-BP-1 with the eukaryotic initiation factor 4F (eIF4F) complex [20]. mTORC2 functions are less understood. Recently, mTORC2 has been shown to be involved in phosphorylating Akt at Ser⁴⁷³ which, along with phosphorylation at Thr³⁰⁸ by PDK1, is required for full activation of Akt [21, 22].

The primary negative regulator of the PI3K/Akt/mTOR pathway is the tumor suppressor phosphatase and tensin homolog deleted on chromosome 10 (PTEN) [23]. This phosphatase acts on the D3 phosphorylated position of PIP3, promoting formation of PIP2 and directly opposing the action of PI3K [24, 25]. PTEN has a crucial role in controlling cell size, organ size, and proliferation, and loss of PTEN is the most common cause of overactivation of the PI3K pathway in human cancers [26].

PI3K/Akt/mTOR ACTIVITY IN PROSTATE CANCER

Alterations in the PI3K/Akt/mTOR pathway have been detected in prostatic tissues in multiple studies, suggesting that this pathway plays a prominent role in the development and progression of prostate cancer. It is estimated that upregulation occurs in 30-50% of prostate cancers, and aberrant signaling of the molecules in this pathway have also been detected in prostate cancer cell lines and xenografts.

PTEN is a negative regulator of activity of the PI3K/Akt/mTOR pathway. PTEN deletions and mutations that result in expression of inactive protein lead to increased activity of the PI3K/Akt/mTOR pathway. Mutations in the PTEN tumor suppressor are common events in prostate cancer, with studies showing loss of heterozygosity at the PTEN locus in up to 60% of prostate cancer samples [27-30]. Decreased expression of PTEN has been found in 85% of primary tumors relative to normal tissues from the same patients, and PTEN expression was also reduced in cancer relative to prostatic intraepithelial neoplasia (PIN) [31]. Alterations in PTEN expression are associated with a number of clinico-pathologic variables in prostate cancer. Loss of PTEN expression correlated with Gleason score and pathologic stage of primary tumors [30, 32] and increased the incidence of development of lymph node metastases [33]. Moreover, when combined with detection of phospho-Akt, PTEN status of the primary tumor was a better predictor of PSA recurrence than phospho-Akt alone (AUC 0.890) [34]. Importantly, 90% of the patients with PTEN-negative primary tumors with high levels of phospho-Akt experienced a biochemical recurrence, while 88% of PTEN-positive tumors with low phospho-Akt did not recur within the study period.

In vitro and preclinical studies have also shown that inactivation of PTEN leads to constitutively activated Akt and mTOR, as well as deregulation of cell size and cell growth [35]. A number of commonly employed prostate cancer-cell lines, including PC-3, LNCaP, and C4-2, are PTEN-negative or express inactive PTEN. Mice heterozygous for PTEN develop PIN with 100% incidence. PTEN homozygous knockouts die in utero, while mice with prostate-specific deletion of PTEN develop invasive prostate cancer [36].

Changes in expression and activation of Akt have also been reported in prostate cancer. Akt protein was detected in virtually every sample in a study of 56 prostatectomy specimens, with cancer cells having greater staining intensity and an increased percentage of positive-staining cells compared to non-neoplastic cells (p<0.001) [37]. Furthermore, phospho-Akt levels were also significantly greater in high-grade prostate tumors vs. low- or intermediate-grade tumors; phospho-Akt was detected in 14% of samples with Gleason score ≤6, 36% of samples with Gleason score 7, and 92% of samples with Gleason score ≥8 tumors (p<0.001) [38]. Levels of phospho-Akt were significantly increased in cancer cells relative to normal prostate epithelium and benign prostatic hyperplasia (45.8% vs. 8.4%) [38]. Phospho-Akt was found to be an independent predictor of biochemical recurrence (HR 3.44, CI 1.83-6.43) [39], and increased levels of phospho-Akt were detected in primary tumors of patients who eventually suffered PSA recurrence (p<0.001) while no correlation was found between Akt expression and biochemical recurrence [40]. Furthermore, increased levels of phospho-Akt were detected in CRPC tissues when compared with hormone-sensitive tissues and were associated with decreased disease-specific survival (HR 2.89, CI 1.43-5.8) [41]. Results of a study evaluating expression of Akt iso-forms with respect to prostate cancer recurrence showed that only high cytoplasmic Akt-1 combined with low nuclear Akt-1 independently predicted time to biochemical failure (HR 2.2, CI 1.12-3.99) [42].

Levels of mTOR and cytoplasmic phospho-mTOR were greater in prostate cancer tissue vs. normal prostatic epithelium, with mTOR levels in cancer cells twice that of benign tissue [31]. Phospho-mTOR was detected at low levels in the cytoplasm and at moderate to high levels along the membrane in normal prostatic epithelium, while in cancer cells strong immunoreactivity of phospho-mTOR was detected both at the membrane and in the cytoplasm. Comparisons of levels of signaling molecules downstream of mTOR, such as 4E-BP1 and S6, also showed higher levels in prostate cancer vs. normal cells [31]. Further evidence surrounding the activity of mTOR in prostate cancer is indirect and comes from the use of mTOR inhibitors, which will be discussed below.

PI3K/Akt/mTOR PATHWAY INHIBITION IN PROSTATE CANCER

PI3K Inhibition

Multiple small molecule inhibitors of the PI3K/Akt/ mTOR pathway have been investigated in both in vitro and in vivo models of prostate cancer (see Table 1). The most studied PI3K inhibitors to date are LY294002 and wortmannin. LY294002 is a potent inhibitor and a competitive antagonist of PI3K. LY294002 treatment resulted in cell-cycle arrest of LNCaP cells and sensitized these cells to radiation [43]; decreased the invasive properties of LNCaP, PC-3, and DU 145 cells [44]; and inhibited angiogenesis in PC-3 cells via decreased levels of HIF1-α and VEGF [45]. LY294002 also lowered levels of phospho-Akt in PC-3 and LNCaP cells [46]. However, in addition to PI3K inhibition, LY294002 inhibits DNA-dependent protein kinase, ataxia teleangectasia mutated, estrogen receptor, mTOR, and even voltage gated K⁺ channels [47-50]. Therefore, some of the effects of LY294002 may not be directly related to its ability to inhibit PI3K. Wortmannin is a fungicide that was originally isolated from soil and is an irreversible inhibitor of PI3K [51]. Treatment with wortmannin decreased levels of phospho-Akt in PC-3 and LNCaP cells [52, 53], and induced apoptosis and radiosensitized DU 145 cells [54, 55]. Wortmannin, similar to LY294002, is not specific to PI3K and inhibits multiple other signaling molecules [48]. Unfortunately, in vivo use of both, LY294002 and wortmannin, have met with significant negative side effects [56]. Nonetheless, in vivo LY294002 decreased phosphorylation of eIF4F and translation of downstream proteins in the prostates of transgenic mice constitutively expressing an active catalytic subunit of PI3K [57].

Table 1.

Pre-Clinical Studies of Inhibition of PI3K/Akt/mTOR Pathway in Prostate Cancer

PI3K Inhibitors	Other Drugs Tested	Experimental Model	Results	Reference
LY294002		p110 transgenic	Decreased pAkt by 47%; significant decreases in eIF4G, Mst1 and RanBP2	[57]
Curcumin		TRAMP	Reduced formation of PIN and adenocarcinoma	[58]
Curcumin	TRAIL	LNCaP s.c.	Decreased tumor growth and cell proliferation, increased apoptosis	[59]
Curcumin	Gemcitabine Radiation tx.	PC-3 s.c.	Curcumin inhibited growth; additional benefits seen when used in combination	[61]
Akt Inhibitors
Deguelin		PC-3 s.c.	38 % reduction in tumor volume at 15 days	[80]
GSK690693		LNCaP s.c.	Significant inhibition of tumor growth by ~50%	[81]
Celecoxib	Atorvastatin	PC-3 s.c.	Inhibited formation of tumors used in combination	[74]
Celecoxib	Exisulind	Wister-Unilever Rats	Decreased incidence of PIN and adenocarcinoma	[72]
Celecoxib	green tea polyphenol	CWR22Rnu1 s.c.	57% growth inhibition with celecoxib alone	[75]
Celecoxib		TRAMP	Reduced formation of PIN and adenocarcinoma	[73]
Celecoxib		PC-3 s.c.	Reduced tumor volumes by 52% at highest dose	[76]
Celecoxib DMC		PC-3 s.c.	No inhibition with celecoxib – DMC with significant tumor growth inhibition	[77]
Genistein		Orthotopic PC-3	Reduced lung metastasis between 40-60%	[69]
Genistein		TRAMP	Reduced development of poorly differentiated CaP	[68]
Genistein	Docetaxel	SCID/hu PC-3	Inhibition of tumor growth with genistein alone - benefits in combination	[70]
Genistein	Radiation tx.	Orthotopic PC-3	Genistein – 30% tumor volume reduction; 84% reduction in combination; genistein alone increased metastasis	[65]
Genistein		LNCaP s.c.	Reduced tumor volume and incidence	[67]
mTOR Inhibitors
RAD-001	Docetaxel Zoledronic acid	C4-2 intra-tibial	Significant decreases in tumor growth with addition of drugs in combination	[100]
RAD-001		Akt1 transgenic	Reversed high grade PIN lesions	[94]
Rapamycin	IRS-1 ASO	PC-3 s.c.	Significant growth inhibition by rapamycin; additive effect with IRS-1 ASO; IHC - 20% decreased proliferative index by rapamycin	[97]
Rapamycin	HDAC inhibitor	PC-3 s.c	53% tumor volume reduction alone; 80% reduction in combination	[98]
CCI-779	Docetaxel	PC-3 & DU 145 s.c.	Inhibited growth – significant decrease in Ki-67 index	[91]
CCI-779	Doxorubicin	PC-3 s.c.	Reduced tumor growth by 40% alone and by 69% in combination with doxorubicin on doxorubicin resistant cells	[35]

s.c. subcutaneous.

(HR 3.44, CI 1.83-6.43) [39], and increased levels of phospho-Akt were detected in primary tumors of patients who eventually suffered PSA recurrence (p<0.001) while no correlation was found between Akt expression and biochemical recurrence [40]. Furthermore, increased levels of phospho-Akt were detected in CRPC tissues when compared with hormone-sensitive tissues and were associated with decreased disease-specific survival (HR 2.89, CI 1.43-5.8) [41]. Results of a study evaluating expression of Akt iso-forms with respect to prostate cancer recurrence showed that only high cytoplasmic Akt-1 combined with low nuclear Akt-1 independently predicted time to biochemical failure (HR 2.2, CI 1.12-3.99) [42].

Levels of mTOR and cytoplasmic phospho-mTOR were greater in prostate cancer tissue vs. normal prostatic epithelium, with mTOR levels in cancer cells twice that of benign tissue [31]. Phospho-mTOR was detected at low levels in the cytoplasm and at moderate to high levels along the membrane in normal prostatic epithelium, while in cancer cells strong immunoreactivity of phospho-mTOR was detected both at the membrane and in the cytoplasm. Comparisons of levels of signaling molecules downstream of mTOR, such as 4E-BP1 and S6, also showed higher levels in prostate cancer vs. normal cells [31]. Further evidence surrounding the activity of mTOR in prostate cancer is indirect and comes from the use of mTOR inhibitors, which will be discussed below.

PI3K/Akt/mTOR PATHWAY INHIBITION IN PROSTATE CANCER

PI3K Inhibition

Multiple small molecule inhibitors of the PI3K/Akt/ mTOR pathway have been investigated in both in vitro and in vivo models of prostate cancer (see Table 1). The most studied PI3K inhibitors to date are LY294002 and wortmannin. LY294002 is a potent inhibitor and a competitive antagonist of PI3K. LY294002 treatment resulted in cell-cycle arrest of LNCaP cells and sensitized these cells to radiation [43]; decreased the invasive properties of LNCaP, PC-3, and DU 145 cells [44]; and inhibited angiogenesis in PC-3 cells via decreased levels of HIF1- and VEGF [45]. LY294002 also lowered levels of phospho-Akt in PC-3 and LNCaP cells [46]. However, in addition to PI3K inhibition, LY294002 inhibits DNA-dependent protein kinase, ataxia teleangectasia mutated, estrogen receptor, mTOR, and even voltage gated K⁺ channels [47-50]. Therefore, some of the effects of LY294002 may not be directly related to its ability to inhibit PI3K. Wortmannin is a fungicide that was originally isolated from soil and is an irreversible inhibitor of PI3K [51]. Treatment with wortmannin decreased levels of phospho-Akt in PC-3 and LNCaP cells [52, 53], and induced apoptosis and radiosensitized DU 145 cells [54, 55]. Wortmannin, similar to LY294002, is not specific to PI3K and inhibits multiple other signaling molecules [48]. Unfortunately, in vivo use of both, LY294002 and wortmannin, have met with significant negative side effects [56]. Nonetheless, in vivo LY294002 decreased phosphorylation of eIF4F and translation of downstream proteins in the prostates of transgenic mice constitutively expressing an active catalytic subunit of PI3K [57].

Recently, curcumin has been shown to inhibit PI3K activity, in addition to other mechanisms of action. Curcumin treatment induced apoptosis of LNCaP, PC-3, and DU 145 cells, inhibited growth of LNCaP and PC-3 xenografts, and inhibited PIN formation in TRAMP mice [58-61]. Because of the critical nature of PI3K in cancer and limited availability of specific inhibitors, there are several new PI3K inhibitors in development that may have improved selectivity against PI3K; however, these have not yet been evaluated in prostate cancer [62].

Akt Inhibition

Multiple Akt inhibitors have been investigated in prostate cancer. Perifosine is an alkylphospholipid that decreases Akt phosphorylation and upregulates expression of the tumor suppressor p21 [63]. Perifosine inhibited growth and induced cell-cycle arrest of PC-3 cells, and it also induced differentiation of PC-3 and LNCaP cells through activation of the GSK-3β pathway [64]. Although there are no published pre-clinical studies investigating perifosine activity against prostate cancer, perifosine has gone on to clinical trials in prostate cancer.

Genistein, an isoflavinoid found in soy, is another compound with Akt inhibitory activity. In vitro studies have showed that genistein inhibits activity of Akt and causes significant growth inhibition and apoptosis of prostate cancer cells [65, 66]. However, genistein also inhibits mTOR, AR, and a wide range of other targets (i.e. tyrosine kinases, topoisomerases, and telomerases). In vivo, genistein has shown significant potential [67, 68], sensitizing prostate tumors to radiation [65] and decreasing the incidence of lung metastasis in an orthotopic model using PC-3 cells [69]. In an experimental model of bone metastasis, genistein in combination with docetaxel further inhibited growth over either agent alone [70]. Interestingly, in one report, genistein increased the size of metastatic lymph nodes in a PC-3 orthotopic model [65].

Celecoxib, a well-known cyclooxygenase-2 (COX-2) inhibitor, also inhibits phosphorylation of Akt in prostate cancer cells [71], and may have chemopreventive properties as demonstrated by its ability to reduce formation of PIN and adenocarcinoma in murine and rat models of prostate cancer [72-76]. However, because of the multiple activities of celecoxib, the observed effects cannot be solely attributed to inhibition of Akt. To further delineate anti-Akt activities of celecoxib, a structural analogue of celecoxib, dimethylcelecoxib (DMC), which does not posses COX-2 inhibitory activity has been developed. Administration of DMC to animals bearing PC-3 tumors resulted in inhibition of Akt phosphorylation but also in inhibition of PDK1 and was accompanied by decreased tumor growth [77].

There are two other recent additions to the family of Akt inhibitors, deguelin and GSK690693. Deguelin, a rotenoid, has been shown to inhibit Akt activation in vitro [78, 79] and PC-3 tumor growth in vivo [80]. However, as with other naturally occurring compounds, deguelin appears to have activity against multiple other molecules (i.e., heat-shock protein (HSP 90); nuclear factor kappa B (NFκB)) [79-82]. GSK690693, a compound that competes for ATP-binding sites on Akt, inhibited proliferation of PC-3 and DU 145 cells in vitro and caused inhibition of LNCaP tumor growth in vivo [81].

mTOR Inhibition

mTOR inhibitors have met with the most success among the inhibitors of the PI3K/Akt/mTOR pathway in treating solid tumors and have also received the most attention in the treatment of prostate cancer. The most extensively studied mTOR inhibitors are rapamycin and its analogues, including RAD-001 (everolimus), CCI-779 (temsirolimus), and AP23573 (deforolimus). A number of studies have demonstrated efficacy of these inhibitors against prostate cancer cell lines and xenografts.

Rapamycin

Rapamycin is a macrolide antibiotic with immunosuppressive and anticancer activities. It was originally found in the soil on Easter Island (known by natives as Rapa Nui) and was eventually isolated from the bacteria Streptomyces hygroscopicus [82]. The precise mechanisms of mTOR inhibition by rapamycin are not fully understood; however, it is known that rapamycin associates with FK506 binding protein 12 (FKBP12), and this complex then binds mTOR, resulting in inhibition of mTOR kinase activity. Chronic exposure to rapamycin also decreases levels of phosphorylated Akt [83]. In contrast, short-term treatment with rapamycin increases levels of phospho-Akt, potentially representing activation of the Akt survival pathway, a means for rapamycin resistance [84].

Studies of mTOR inhibition have increased our understanding of the roles of mTOR and its function in several cellular pathways important for prostate cancer development and progression. Rapamycin treatment decreased levels of the phosphorylated substrates of mTOR (i.e., activated S6K) and led to cell-cycle arrest in PC-3 and DU 145 cells [84-87]. Rapamycin also decreased levels of p4E-BP1, leading to an increase in bound (and inactive) eIF4E [84, 86, 87]. Several studies have focused on the changes in gene expression that occur after treatment with rapamycin: increased expression of bone morphogenetic protein 4 (BMP4), suppression of follistatin and a resultant increase in Smad activity have been detected in LNCaP and PC-3 cells treated with rapamycin, implicating the effects on transforming growth factor beta (TGF[notdef]) signaling [88]. Rapamycin has also decreased HIF-1α expression in PC-3 cells despite placement in hypoxic environments, with further decreases observed when rapamycin was used in combination with histone deacytelase inhibitors [89].

There are also emerging data suggesting that mechanism of rapamycin action may be cell-specific. From the two mTOR complexes (mTORC1 and mTORC2) only mTORC1 has been thought to be sensitive to rapamycin. mTORC2 has been considered resistant to mTOR inhibition. However, new evidence suggests that mTORC2 is inhibited by prolonged exposure to rapamycin, but only in certain cells. Interestingly, suppression of mTORC2 was demonstrated in PC-3 prostate cancer [22, 90]. Thus, while mTORC1 is inhibited by rapamycin in all cells, mTORC2 inhibition is likely cell-dependent. This differential effect lends some insight into why mTOR inhibition may be an effective therapy for some tumors and not others, and the identification of the molecular characteristics associated with mTORC2 susceptibility to rapamycin remains an important goal. This could further inform the use of mTOR inhibitors in future clinical trials.

Rapamycin Analogues

Although limited, there are reports on in vitro and pre-clinical investigations demonstrating the efficacy of the rapamycin analogs CCI-779 and RAD-001 in the treatment of prostate cancer. CCI-779 inhibited growth of PC-3 and DU 145 cells in a dose-dependent manner in vitro, and in vivo, reduced tumor volumes in PC-3 and DU 145 xenografts [91]. RAD-001 treatment resulted in decreased proliferation of prostate cancer cells in vitro [92, 93] and reversed PIN lesions in vivo through HIF-1α-dependent pathways in Akt-1 transgenic mice [94]. There are no published reports on the rapamycin analog AP23573 (deforolimus) in prostate cancer at present. Nonetheless, all three analogs, along with rapamycin, are currently under investigation in clinical trials for treatment of prostate cancer (Table 2).

Table 2.

Clinical Trials of mTOR Inhibitors in Prostate Cancer

mTOR Inhibitor	Other Drugs Tested	Institution/Sponsor	Patients	Outcomes	Phase	Status
RAD-001a	Docetaxel Bevacizumab	Cedars-Sinai Medical Center, Novartis Pharmaceuticals, Genentech	Metastatic CRPC	Survival/response, safety	I/II	Enrolling
RAD-001b	Docetaxel	Dana-Farber Cancer Institute, Novartis Pharmaceuticals, Massachusetts General Hospital, Beth Israel Deaconess Medical Center, Oregon Health and Science University	Metastatic CRPC	Survival/response, safety	I/II	Enrolling
RAD-001c		Duke University	Metastatic CRPC	PSA response, changes in mTOR pathway (tissue bx)	II	Enrolling
RAD-001d	Bicalutamide	Dana-Farber Cancer Institute, Beth Israel Deaconess Medial Center, Novartis	CRPC	Survival/response, safety	II	Enrolling
RAD-001e	Gefitinib	Memorial Sloan-Kettering Cancer Center, NCI	Glioblastoma multiforme or metastatic CRPC	Survival/response, safety	I/II	Ongoing, not enrolling
AP23573f		Ariad Pharmaceuticals	Taxane-resistant CRPC	Survival/response, safety, molecular markers	II	Ongoing, not enrolling
CCI-779g		Jonsson Comprehensive Cancer Center, NCI	High risk newly dx'd CaP undergoing RP	mTOR pathway markers	II	Ongoing, not enrolling
CCI-779h		Wyeth Pharmaceuticals	Newly dx'd CaP undergoing RP	mTOR pathway markers	II	Completed
RAD-001i	Androgen deprivation	Sheba Medical Center	Intermediate or high-risk CaP undergoing RP or RadRx	mTOR pathway markers, time to biochemical failure	II	Not yet open
Rapamycinj		Sidney Kimmel Comprehensive Cancer Center, NCI	Intermediate or high-risk CaP undergoing RP	Safety, mTOR pathway markers, PSA response	I/II	Enrolling

^aGross, M.E. NCT00574769.

^bTaplin, M.E. NCT00459186.

^cGeorge, D.J. NCT00629525.

^dTaplin, ME. NCT00630344.

^eScher, H.I.; Rosen, N.; Abrey, L.E. NCT00085566.

^fBedrosian, C. NCT00110188.

^gSawyers, C. NCT00071968.

^hNCT00235794.

ⁱNCT00657982.

^jCarducci, M.A. NCT00311623.

In attempts to find improved efficacy, much focus has been placed on finding therapies for advanced prostate cancer with synergistic or additive effects. A key challenge with the use of mTOR and other signal transduction inhibitors is the overlap of signaling pathways, enabling cells to bypass the targeted molecule when exposed to these inhibitors. Resistance to signal transduction inhibitors likely arises from either mutations of key factors in the pathway that allow the signaling cascade to proceed or through upregulation of alternative pathways which enable cell growth and survival via different mechanisms [95]. Therefore, a large number of studies have focused on mTOR inhibition as part of a combination regimen rather than as monotherapy (see Table 1).

A combination of rapamycin and receptor tyrosine kinase inhibitors decreased survival of LNCaP and CWR22Rv1 cells in vitro [93], and a combination of rapamycin and D-glucosamine (an inhibitor of p70-S6K) increased growth inhibition of DU 145 cells [96]. Rapamycin, in combination with an insulin receptor substrate (IRS-1) antisense oligodeoxinucleotide exhibited a more pronounced inhibition of PC-3 tumor growth than treatment with the IRS-1 antisense alone [97]. Growth inhibition of PC-3 and C4-2 tumors was increased with the combination of rapamycin and histone deactylase inhibitors over either agent alone [98]. CCI-779 reversed doxorubicin resistance of PC-3 and DU 145 tumors [35] and had additive effects when used in combination with docetaxel [91]. RAD-001 used in combination with an epidermal growth factor receptor inhibitor (gefitinib) and a novel anti-androgen, VN/124-1, had additive inhibitory effects on growth of LNCaP cells in vitro [99]. RAD-001 also sensitized prostate cancer cells to radiation [92]. We have recently shown that RAD-001 in combination with zoledronic acid and docetaxel more effectively inhibited growth of prostate tumor cells in the bone environment over any of these agents alone [100].

INTERACTION BETWEEN PI3K/Akt/mTOR PATHWAY AND ANDROGEN RECEPTOR

Detailed investigations have shown that crosstalk exists between the PI3K/Akt/mTOR and AR signaling pathways. AR is a key modulator of growth and development of the prostate and of prostate cancer progression [101-104]. AR-mediated transcription is tightly controlled and mechanisms of regulation of AR-transcriptional activity include association with transcriptional cofactors as well as phosphorylation and acetylation [105-110]. A better understanding of the molecular interactions and crosstalk between AR and other signaling pathways might have a dramatic positive impact on strategies to treat prostate cancer [111, 112].

Increasing evidence suggests that key factors of the PI3K/Akt/mTOR pathway may directly regulate the expression and transcriptional activity of AR [113, 114]. In particular, it has been demonstrated that AR phosphorylation and activation by Akt occurs predominantly at low androgen concentrations, suggesting a significant role of Akt in stimulating cell growth in the castrate state [115, 116]. Inhibition of the PI3K/Akt pathway with LY294002 decreased DHT-induced expression of AR in LNCaP cells, while expression of a dominant-negative Akt blocked AR expression [116]. Conversely, stimulation of LNCaP cells with DHT led to AR-mediated activation of mTOR independent of PI3K/Akt stimulation [117]. Recent data has also shown that androgen-dependent LNCaP cells respond weakly to mTOR inhibition in vitro, while growth of the castrate-resistant C4-2 cells is significantly reduced [118]. Reintroduction of PTEN in C4-2 cells increased their sensitivity to androgen ablation with bicalutimide [119]. Furthermore, increased levels of phospho-mTOR and phospho-Akt have been detected in high-passage LNCaP cells after treatment with an androgen inhibitor [99]. Interestingly, treatment with the mTOR inhibitors RAD-001 or rapamycin has resulted in increased AR-transcriptional activity in both high-passage/androgen-independent and low passage/androgen-dependent LNCaP cells [120, 121].

A recent report has demonstrated the clinical relevance of these in vitro results. A comparison of matched hormone-sensitive and hormone-resistant tissues from patients who progressed to CRPC revealed that upregulation of the PI3K/Akt pathway was associated with AR phosphorylation during transition from a hormone-sensitive to a hormone-refractory state [41]. Furthermore, increases in phospho-Akt and phospho-AR were each associated with decreased disease-specific survival. These results together suggest that, as clinical trials with inhibitors of the PI3K/Akt/mTOR pathway move forward, efficacy may be highly dependent upon patient populations in terms of exposure to hormonal therapies and resistance to castration.

CLINICAL USE OF PI3K/Akt/mTOR INHIBITORS IN PROSTATE CANCER

The results of in vitro and preclinical studies suggest that, due to adverse effects, current inhibitors of PI3K and Akt may have limited use in clinical practice. At present, the most promising inhibitors of PIK3/Akt/mTOR pathway for the treatment of prostate cancer are mTOR inhibitors, some of which are already in clinical use for other pathological conditions as well as for other malignancies.

Only two of the compounds that inhibit Akt activation, perifosine and celecoxib, have been investigated in the clinical setting. In a trial investigating the effects of perifosine in patients with metastatic CRPC (n=19), no complete or partial responses were detected and only four patients had a PSA stabilization for 12 weeks or more [122]. There was, however, a decrease in the detection of circulating tumor cells in 11/14 of these patients after treatment. These results may be significant since circulating tumor cells are considered evidence of disseminated disease [123], and decreases in circulating tumor cells have been shown to correlate with increased survival in patients with metastatic breast cancer [124]. Long term follow up is needed to determine whether these effects of perifosine will result in clinical improvements. In a phase II study in men with biochemically recurrent, hormone-sensitive prostate cancer, perifosine administration resulted in PSA decreases in 5/24 patients; however, no patients met the predefined criteria for a true response (a 50% or greater reduction in PSA) [125]. A phase II clinical trial investigating the use of celecoxib in patients with biochemically recurrent prostate cancer (n=40) after radiation or radical prostatectomy showed a significant inhibition of PSA doubling time [126]. Three months after treatment initiation, 90% of patients had a lower PSA doubling time with 11/40 experiencing a decrease in absolute PSA levels. However, a trial of celecoxib vs. placebo in a similar patient population (n=78) did not show any differences in PSA doubling time [127]. Celecoxib in combination with docetaxel and estramustine in CRPC patients resulted in a median survival of 19.2 months, relatively similar to TAX 327 and SWOG 99-16 [128]. Further trials, such as STAMPEDE, will help to determine the role of COX-2 inhibition in treating advanced prostate cancer and whether any anti-cancer activity is due to its Akt-inhibiting properties.

Clinical investigation of mTOR inhibitors in the oncologic setting is a relatively new, but promising area of investigation that started within the past decade. Rapamycin was initially developed as an immunosuppressive agent and was approved by the FDA in 1998 for this purpose [33]. The pharmacokinetics of this drug is well known, with excellent absorption after oral dosing and peak concentrations at approximately 1.5 hours after administration. The incidence of severe toxicity reactions has been rare, and only mild adverse effects including hyperlipidemia, thrombocytopenia, leukopenia, diarrhea, skin rash, pneumonitis, and electrolyte abnormalities have been reported. There are also quickly accumulating data regarding pharmacologic profiles of rapamycin analogs showing that these analogs are well-tolerated and exhibit minimal negative side effects [129-134].

Efficacy of mTOR inhibition was demonstrated in early phase clinical trials in a number of malignancies, and mTOR inhibitors are now in clinical development in endometrial cancer, breast cancer, glioblastoma, lymphoma, and sarcomas [135]. CCI-779 was investigated in a large phase III trial in advanced renal cell carcinoma, and median overall survival was significantly increased vs. IFN-α (10.9 months vs. 7.3 months, p=0.008) [136]. CCI-779 (temsirolimus) was subsequently approved by the FDA in 2007 for the treatment of advanced renal cell carcinoma.

In prostate cancer, there are several ongoing phase I and II clinical trials with mTOR inhibitors (Table 2). Some of these trials are designed in the neoadjuvant and/or the adjuvant setting. These trials are focused on analysis of key factors in mTOR signaling and their alterations in response to mTOR inhibition. Certainly, the ability to assess the molecular response to therapy is one of the central advantages of cell signaling modifiers. Molecular stratification of patients to mTOR inhibitor therapy may help to identify those patients most likely to benefit from treatment while sparing those patients who are unlikely to respond. Neoadjuvant use of mTOR inhibition could also provide an opportunity to target tumor cells before they have accumulated the large number of mutations that typically arise with advanced disease. PTEN mutations and deletions within primary tumors have been associated with an increased risk of metastasis [33] and early targeting may prove to be beneficial in preventing metastatic spread. Additionally, it has been suggested that mTOR inhibitors could be used as chemopreventive agents in patients who have deleted or inactivated PTEN in benign prostate epithilium or PIN at the time of prostate needle biopsy [7]. Because the clinical trials testing mTOR inhibitors are ongoing or still accruing patients, there are only limited results available at this time. A preliminary analysis of a phase II clinical trial of neoadjuvant administration of RAD-001 in patients prior to radical prostatectomy has not only showed that the drug is well tolerated but also that it decreases the levels of activated mTOR substrates in the primary tumor.¹

A majority of the ongoing trials in prostate cancer are assessing mTOR inhibition in the setting of CRPC. One study in patients with metastatic CRPC is evaluating the cellular and molecular responses to RAD-001 by comparing pre- and post-treatment bone-derived tumor biopsies.² Results of this trial, similar to the neoadjuvant studies assessing phenotypic changes in the primary tumor, will provide important information regarding the efficacy of these therapies on a molecular level.

Because of the heterogeneity of prostate cancer [137, 138] and the ability of tumor cells to undergo cellular alterations that allow survival under changing conditions, most of the trials investigating mTOR inhibition in CRPC utilize combinations of drugs. The majority of these trials are designed to provide a horizontal blockade within the cancer cell. Horizontal blockade refers to the simultaneous inhibition of multiple different targets. For example, inhibiting a MAP kinase at the same time as mTOR may block one of the key pathways that overlaps with the PI3K/Akt/mTOR pathway. Another approach to horizontal blockade involves targeting different cell types, such as targeting endothelial cells with a VEGF inhibitor, pericytes with a PDGF inhibitor, and/or osteoblasts with an endothelin A inhibitor, while also targeting the tumor cell directly [139]. The second approach to combination therapy is to administer agents according to a vertical blockade rationale. A vertical blockade is designed to target multiple key factors within one specific pathway. For example, simultaneous inhibition of PI3K, Akt, and mTOR may be required to fully suppress activity of this pathway. Since upstream molecules in the mTOR pathway may be upregulated with administration of mTOR inhibitors—proposed as mechanism for mTOR inhibitor resistance [84, 95]—vertical blockade may prevent the shunting of upstream molecules down alternative signaling pathways. However, initial analysis of AP23573 used in combination with the epidermal growth factor inhibitor gefitinib in patients with advanced prostate cancer showed that only 5/29 patients had no disease progression at 12 weeks.³

CONCLUSIONS

Increasing molecular evidence from in vitro studies, prostate cancer animal models, and staining of human prostate tissues demonstrates that the PI3K/Akt/mTOR pathway plays a critical role in prostate cancer development and progression. Data generated using clinical samples have shown that increased expression of key factors in this pathway correlates with disease stage and lower survival rates. Inhibition of the PI3K/Akt/mTOR pathway in prostate cancer models has lent significant insight into the mechanisms behind the development of advanced prostate cancer. Unfortunately the in vitro studies have also demonstrated that present inhibitors of PI3K and Akt are not highly specific, and preclinical studies have shown that use of these compounds is associated with significant negative side effects. More promising, currently, are the inhibitors of mTOR. These have been shown to inhibit proliferation of prostate tumor cells and show high specificity for mTOR in vitro, and these inhibitors have inhibited tumor growth in the preclinical setting with minimal negative side effects.

The in vitro and preclinical results are encouraging, and multiple phase I and phase II clinical trials are underway to evaluate the efficacy of mTOR inhibitors in both the neoadjuvant setting and in advanced prostate cancer patients. Because of the ability of tumor cells to adapt to new conditions, mTOR inhibitors are also being investigated in combination with other drugs. Since the clinical trials in prostate cancer are in their early stages, it remains unclear what role mTOR inhibitors will have in the care of patients with prostate cancer. The results from these trials will help to determine the efficacy of mTOR inhibition in prostate cancer, further the mechanistic understanding of these therapies, and move prostate cancer treatment towards rational strategies based on tumor-specific markers. Given the tremendous heterogeneity of prostate cancers, particularly metastatic cancers, therapies based on protein expression profiles of individual tumors may provide the best chance for success. Further work will determine whether the growing knowledge of tumor biology can be translated into real clinical gains.

ABBREVIATIONS

AR

androgen receptor

BAD

Bcl-xL/Bcl-2 associated death promoter

BMP4

bone morphogenetic protein 4

Chk1

checkpoint kinase 1

CRPC

castration resistant prostate cancer

DMC

dimethyl-celecoxib

EGFR

epidermal growth factor receptor

eIF4F

eukaryotic initiation factor 4F

FKB12

FK506 binding protein

FOXO

forkhead box O

HIF-1α

hypoxia inducible transcription factor 1 alpha

HSP 90

heat shock protein 90

IGFR

insulin-like growth factor receptor

IRS-1

insulin receptor substrate

MAP

mitogen activated protein

MDM2

murine double minute

mTOR

mammalian target of rapamycin

PDGFR

platelet derived growth factor receptor

PDK1

phosphoinositide dependent protein kinase 1

PIN

prostatic intraepithelial neoplasia

PIP2

phosphatidylinositol-4,5-diphosphate

PIP3

phosphatidylinositol-3,4,5-triphosphate

PTEN

phosphatase and tensin homolog deleted on chromosome 10

PSA

prostate specific antigen

PI3K

phosphatydidyl inositol-3-OH kinase

S6K

p70-S6 kinase

TGFβ

transforming growth factor beta

TSC2

tuberous sclerosis complex 2

4E-BP1

4E binding protein 1