Mary-Ellen Taplin, Lank Center for Genitourinary Oncology, Dana-Farber Cancer Institute, 450 Brookline Avenue, Boston, MA 02215, USA. e-mail: email@example.com
Study Type – Therapy (cohort)
Level of Evidence 2a
What's known on the subject? and What does the study add?
Despite expanding treatment options for castration-resistant prostate cancer (CRPC), therapies with long response duration remain intangible due to prostate cancer cells’ natural ability to develop iterative resistance. Androgen receptor (AR) signaling has been shown to play a critical role in CRPC and its expression is regulated by the PI3K-Akt pathway. Thus inhibition of AR signalling and PI3K-Akt-mTOR (a downstream mediator of the PI3K-Akt pathway) pathway is a logical combination in CRPC and we report a phase II trial of RAD001 and bicalutamide.
Our study is the first clinical trial report of an AR inhibitor of PI3K-Akt-mTOR. The AR pathway and the PI3K-Akt-mTOR pathway are two of the most relevant growth pathway for CRPC. Despite low efficacy results from our trial there will be significant interest in the field for these data (dose, schedule, response, toxicity, trial design) as newer generations of both AR inhibitors and PI3K-Akt-mTOR inhibitors are in development and likely will be tested in combination in CRPC.
• To determine best overall response and duration of response of RAD001, a selective inhibitor of mammalian target of rapamycin, in combination with bicalutamide in castration-resistant prostate cancer (CRPC).
• To characterize the toxicity profile of RAD001 in combination with bicalutamide in patients with CRPC.
PATIENTS AND METHODS
• A phase II study was conducted to explore the efficacy and tolerability of RAD001 (10 mg daily) in combination with bicalutamide (50 mg daily) in men with progressive CRPC.
• The primary endpoint was a composite of prostate-specific antigen (PSA) level and measurable disease response by standard criteria.
• This single-stage trial with a sample size of 38 eligible patients provided 90% power to differentiate a response rate of ≥40% from a response rate of ≤20%, as expected for bicalutamide alone (α= 0.10, power = 0.90).
• In total, 36 men were enrolled, with a median (range) age of 68 (60–72) years and median (range) baseline PSA level of 22.2 (8.4–121.3) ng/mL, and 89% had metastatic disease.
• There were 31 (86%) patients had previously used bicalutamide for a median duration of 7.4 months.
• There were two patients with a confirmed PSA level decline ≥50%.
• The median (interquartile range) time to progression was 8.7 (7.9–15.9) weeks.
• The most common toxicity was grade 1/2 mucositis, which was observed in 20 (56%) patients.
• The combination of RAD001 and bicalutamide in men with CRPC was well tolerated but had low activity and failed to achieve the primary endpoint of improved response compared to the results previously achieved for bicalutamide alone in this population.
The treatment options for castration-resistant prostate cancer (CRPC) are expanding; however, therapies with long response durations remain elusive as a result of the inherent ability of prostate cancer cells to develop iterative resistance. Primary androgen-deprivation therapy and strategies for secondary androgen deprivation in CRPC remain the mainstays of non-chemotherapy treatment for metastatic prostate cancer.
Despite castrate levels of testosterone in CRPC, androgen receptor (AR) signalling has been shown to play an ongoing and critical role in cell growth through AR amplification, mutation, development of splice variants, alteration in AR co-stimulatory molecules and intracrine synthesis of androgens [1–4]. Bicalutamide (CasodexTM; AstraZeneca, London, UK), a competitive AR antagonist, is a Food and Drug Administration approved non-steroidal antiandrogen. Bicalutamide is frequently the first ‘secondary hormonal therapy’ added to luteinizing hormone-releasing hormone agonist/antagonist therapy to treat CRPC . The response rate of bicalutamide (150–200 mg) in CRPC is ≈20% with median response duration of 3–4 months [6–8]. The lack of efficacy of bicalutamide and other antiandrogens in CRPC is probably multifactorial, including weak antagonist binding affinity to AR, up-regulation of AR co-stimulatory pathways and increased intratumoural androgen levels as a result of intracrine androgen, as well as partial agonist properties of antiandrogens when AR is overexpressed [1,9–12]. We hypothesized that strategies inducing more potent AR inhibition by simultaneously targeting receptor ligands or co-stimulatory pathways such as the phosphoinositide 3-kinase (PI3K)-Akt-mammalian target of rapamycin (mTOR) pathway may be critical for longer response durations.
Aberrant activation of the PI3K-Akt-mTOR pathway, a central regulator of cellular functions such as cell proliferation and cell growth, has been widely implicated in many cancers, including prostate cancer . Several levels of cross-talk between the AR and the mTOR pathway have been reported [14–18]. AR expression is regulated by PI3K/Akt signalling . Loss or functional alteration of the tumour suppressor gene Pten, which encodes a phosphatase that antagonizes PI3K activity, occurs in more than 20% of primary prostate cancers and in up to 70% of advanced cancers, which results in hyper-activation of PI3K and Akt [20–23]. PTEN loss is frequent in CRPC and correlates with poor disease-specific mortality . The mTOR serine/threonine kinase is a downstream target of the PI3K-Akt pathway . In response to both nutrient availability and mitogenic growth factors, mTOR controls cell growth by activating the 70-kDa ribosomal S6 kinase (S6K1) and inhibiting the elongation-initiation factor 4E binding protein-1 (4E-BP1), which are two events that stimulate protein translation . Combined AR blockade and mTOR inhibition is an attractive therapeutic strategy in CRPC.
RAD001 (everolimus) is an orally administered rapamycin analogue, originally developed as an immunosuppressant, which has been shown to have antitumour and anti-angiogenic activities . RAD001 can directly inhibit tumour cell growth and proliferation, as well as inhibit vascular endothelial growth factor-mediated signalling that promotes tumour angiogenesis [27,28]. RAD001 has been approved by the US Food and Drug Administration as second-line treatment for metastatic RCC after the failure of vascular endothelial growth factor receptor-targeting tyrosine kinase inhibitors [29,30]. A series of studies has shown that mTOR inhibitors reverse PTEN-Akt-mTOR-mediated cancer growth and reduce tumour volumes in mouse models, including genetically engineered models of prostate neoplasia [31–33]. Zhang et al.  showed that the mTOR inhibitor, rapamycin, had additive antitumour effects when it was given with antiandrogen compared to antiandrogen alone in a Pten-deficient mouse model .
We hypothesized that the combination of the antiandrogen, bicalutamide, and RAD001 would have additive and clinically significant effects as a result of the inhibition of both AR and mTOR in CRPC. We evaluated the efficacy and tolerability of RAD001 in combination with bicalutamide in a multi-institutional single-arm phase II trial in men with CRPC.
PATIENTS AND METHODS
PATIENTS AND ELIGIBILITY
This open-label, single-arm phase II trial was initiated at Dana-Farber Cancer Institute (DFCI). Participating institutions included the DFCI, Beth Israel Deaconess Medical Center, Cape Cod Healthcare and New Hampshire Oncology Hematology. The present study was reviewed and monitored by the DFCI Institutional Review Board and DFCI data and safety monitoring committee.
Eligible patients had histologically documented adenocarcinoma of the prostate and CRPC defined by one of the three criteria: (i) two rising PSA levels over nadir on castration therapy (testosterone ≤50 ng/dL); (ii) progression when on castration therapy as defined using Response Evaluation Criteria in Solid Tumors ; or (iii) two or more new lesions on bone scan when on castration therapy. Metastatic disease was not required. Additional criteria included an Eastern Cooperative Oncology Group performance status in the range 0–2; adequate liver, bone marrow and gastrointestinal function; and <24 months of previous bicalutamide therapy. Previous chemotherapy was allowed if in the adjuvant or neoadjuvant setting >12 months before enrollment and one previous chemotherapy regimen was allowed for CRPC. Exclusion criteria included previous treatment with an mTOR inhibitor, any investigational agent within 4 weeks preceding registration, and chronic treatment with systemic steroids or another immunotherapy.
Protocol treatment was RAD001 (10 mg orally, once daily) and bicalutamide (50 mg orally, once daily). Patients on luteinizing hormone-releasing hormone agonists or antagonists continued on gonadal suppression. Dose modifications for toxicities were specified. Patients were evaluated every 4 weeks with history, physical examination and laboratory tests, including complete blood count and measurement of PSA levels. Toxicity was graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.0 (http://ctep.cancer.gov). Radiographic chest imaging (X-ray or CT) was carried out every 8 weeks to evaluate bicalutamide-induced pneumonitis. Radiographic restaging (CT or MRI for measurable disease; bone scan for bone metastases) was completed after 8 weeks, and then every 12 weeks. For responses, confirmatory scans were obtained at least 4 weeks after the initial documentation. Treatment continued until either disease progression or unacceptable toxicity.
The primary endpoint of the present study was the best overall response, taking into consideration measurable disease, bone metastases and PSA level. We used a modified version of the PCAWG2 criteria . The secondary endpoint was toxicity. The PSA response was defined as a decline ≥50% from baseline, as confirmed by a second measurement at least 4 weeks later. The baseline PSA level was measured no more than 2 weeks before the start of therapy. PSA progression was defined as an increase in PSA level ≥25% over baseline or nadir PSA, whichever was lowest, as confirmed by a second PSA level measurement at least 2 weeks later, provided that the increase was ≥5 ng/mL, and the PSA level was not declining. In patients who had achieved ≥50% decline in PSA level, PSA progression was defined as an increase in PSA level ≥50% above nadir, as confirmed by a second PSA level measurement at least 2 weeks later, provided that the increase was ≥5 ng/mL or back to pretreatment baseline, whichever was lowest. For patients with measurable disease, response and progression were evaluated according to Response Evaluation Criteria in Solid Tumors criteria. Progressive non-measurable disease was defined as a worsening bone scan, as indicated by the appearance of two or more new lesions, the appearance of new non-bony metastases or a requirement for radiation therapy. Patients remained on therapy for 8 weeks before assessing PSA levels and scanning for progression. Time to progression (TTP) was defined from the date of treatment initiation until the date that PSA progression criteria were first met, or the date of measurable or non-measurable disease progression. For patients who did not meet the criteria for progression, TTP was censored at the date of the last PSA level measurement without evidence of progression. Among patients who achieved a response, the duration of the response was defined from the date the response criteria were met until the date of PSA or disease progression, or was censored on the date of the last PSA measurement without evidence of progression. A patient was considered to have a favourable outcome if the PSA level declined in the absence of measurable disease and without the appearance of new bone lesions, or, if a response in measurable disease was observed, without an increase in PSA level or the appearance of new bone lesions. Patients with stable disease lasting ≥6 months were also considered to have a favourable outcome. All other treated patients, including unevaluable patients, were included in the denominator as non-responders. Safety was evaluated among all treated patients using the Common Terminology Criteria for Adverse Events.
Patient and clinical characteristics were summarized as numbers and percentages for categorical variables, and median and interquartile ranges for continuous variables. The response rate was summarized as a number and percentage with 90% CI using the binomial exact method. A Kaplan–Meier estimate was used to summarize median TTP. Any association between TTP and previous use of bicalutamide was assessed using the log-rank test.
In total, 36 patients were enrolled. Patient and disease characteristics are shown in Table 1. The median (interquartile range) age and median (interquartile range) baseline PSA level at enrollment were 68 (60–72) years and 22.2 (8.4–121.3) ng/mL, respectively. There were 33 (89%) patients who had metastatic disease at enrollment. A total of 31 (86%) patients had received at least one secondary hormonal therapy before treatment with the bicalutamide/RAD001 regimen. Previous bicalutamide exposure is discussed below.
There were six (16.7%) patients who had previously received chemotherapy. Of these six patients, three received chemotherapy for CRPC, two received it with local therapy and one received it in conjunction with primary hormone therapy. The chemotherapy agents that they received included docetaxel, estramustine and etoposide.
The PSA response rate (confirmed PSA level decline ≥50% from baseline) was 6% (two of 36 subjects; 90% CI, 1–16%). A PSA level decline of any amount, ≥30% and ≥50%, without a second confirmatory PSA level decline, was observed in 16 (44%), six (17%) and four (11%) patients, respectively (Fig. 1) Stable disease and progressive disease were observed in nine (25%) and 20 (36%) patients, respectively. However, all cases with stable disease lasted <6 months. There were seven patients who were unevaluable for a radiographic response as a result of treatment cessation. No patient experienced a radiographic complete or partial response to treatment. Median (range) TTP was 8.7 (0–43.6) weeks (Fig. 2). The most common type of progression was PSA only progression, which was observed in 14 (54%) patients. There were four (15%) patients who had measurable disease progression without PSA progression. The reasons for stopping the treatment are summarized in Table 2.
Table 2. Reasons for stopping treatment
Other (not specified)
ASSOCIATION BETWEEN TTP AND PREVIOUS USE OF BICALUTAMIDE
Most patients had a previous use of bicalutamide in any setting (n= 31; 86%). The median (range) duration of previous use of bicalutamide was 7.4 (5.4–14.7) months. A total of 25 (69%) patients previously received bicalutamide as secondary hormonal therapy. The median (range) duration of time on previous bicalutamide as secondary hormonal therapy was 7.3 (0.16–34.3) months. We evaluated the association between TTP and previous use of bicalutamide, although there was no significant association in any setting (P= 0.28), nor previous use of bicalutamide as secondary hormonal therapy (P= 0.57).
Toxicities were reported among all patients (Table 3). The most common toxicity was grade 1/2 mucositis, which was observed in 20 (56%) patients, followed by rash (47%), fatigue (44%) and diarrhoea (25%). The most common grade 3 toxicity was hyperglycaemia, which was reported in three (8%) patients. There were no reported grade 4 toxicities.
Table 3. Toxicity profile (CTAE, version 3)
No grade 4 toxicities were observed. CPK, creatine phosphokinase; CTAE, Common Terminology Criteria for Adverse Events; LFT, liver function test.
A multi-institutional, single-arm phase II study was conducted to assess the efficacy and tolerability of RAD001 in combination with bicalutamide in 36 men with CRPC. PSA level declines of ≥50% (with a confirmation PSA level decline at least 4 weeks later) were observed in only two patients, with a median TTP of 8.7 weeks. Stable disease by PSA level was observed in nine (25%) patients; however, stable disease did not exceed 6 months. The response to combination RAD001 and bicalutamide was lower than the reported responses to bicalutamide alone in bicalutamide-naïve CRPC patients. Because most (86%) patients treated in the present study had previous treatment with (and progression on) bicalutamide, it can be concluded that RAD001 did not overcome bicalutamide resistance.
RAD001 is a hydroxy-ethyl ether derivative of rapamycin. RAD001 forms a complex with FK506 binding protein-12. This complex binds and allosterically inhibits mTOR when it is coupled with raptor (regulatory-associated protein of mTOR) and mLST8 to form a rapamycin-sensitive complex (mTORC1) . mTORC1 regulates cell growth in response to nutrients and can be hyper-activated by oncogenic PI3K signalling to promote cancer cell growth [38,39]. mTOR also forms a second protein complex: mTORC2. mTORC2 regulates cytoskeletal dynamics and Akt activation and appears to be insensitive to rapamycin and RAD001 in many circumstances [40,41]. The mTORC2 complex phosphorylates Akt on Ser473 and positively regulates Akt phosphorylation on Thr308 by PDK1 . Negative regulators of mTOR activity include PTEN and tuberous sclerosis complexes 1 (hamartin) and 2 (tuberin).
A number of potential mechanisms may account for the limited clinical activity associated with combined RAD001 and bicalutamide observed in the present study. Pharmacodynamic analysis of both cancer cell lines and patient tumours showed that treatment with RAD001 induced insulin receptor substrate-1 expression and phosphorylation of Akt [42,43]. This Akt activation after mTORC1 inhibition by RAD001 is dependent on rictor, a component of the (predominantly) rapamycin-insensitive mTORC2 complex . Although RAD001-induced Akt activation may contribute to the low response rate observed in the present study, the true impact of this loss of feedback inhibition remains uncertain. For example, Breuleux et al.  observed no correlation between the activation of Akt after treatment with RAD001 and an antiproliferative effect in several human cancer cell lines, including PC3M. The loss of additional inhibitory feedback pathways may also contribute to the low observed response rate. Mitogen-activated protein kinase activation was seen in both patient tumour samples and mouse models of prostate cancer after treatment with RAD001 . We speculate that RAD001 activation of feedback loops via the PI3K-Akt and/or the mitogen-activated protein kinase pathway may have resulted in the low activity observed in the present study. In support of this hypothesis, a recent study found that a novel dual PI3K-mTOR inhibitor, PI-103, exhibited superior activity to RAD001 in prostate cancer cells, possibly as a result of impaired feedback activation of Akt induced by mTOR inhibition [40,42,45,46].
Although rapamycin analogues showed benefit to patients with advanced RCC refractory to sunitinib or sorafenib , the effect of inhibition of mTOR pathway in prostate cancer remains unclear. Our conclusions regarding the low response rate to RAD001/bicalutamide would have been strengthened by correlative pre- and post-RAD001-treatment tumour biopsies aiming to assess for changes in the phosphorylation of mTOR, Akt, S6K and 4EBP1, as well as PTEN status. PTEN loss leading to Akt activation is a frequent event in CRPC, and the antiproliferative effect of RAD001 has been shown to correlate with the level of basal phosphorylation of Akt at Ser473[24,44]. However, although some studies report a correlation between PTEN loss and sensitivity to mTOR inhibition [48,49], other studies do not .
The selected dose of bicalutamide used in the present study, as well as the high percentage of patients who had received previous bicalutamide, may have contributed to the low activity observed for the combination of RAD001 with bicalutamide. Although previous studies of patients with CRPC have shown an ≈20% response rate to bicalutamide [6–8], these trials all employed higher doses (150–200 mg) of bicalutamide. Furthermore, although some of these studies included patients who had received previous flutamide treatment, none included patients who had received previous bicalutamide. We studied the influence of the previous use of bicalutamide on the response to combination therapy because the previous use of bicalutamide was not excluded in the present study. Most (86%) patients previously received bicalutamide, and 69% received bicalutamide as secondary hormonal therapy. Although we found no significant association between TTP and previous use of bicalutamide in any setting, only five of 36 patients were bicalutamide-naïve. When AR is overexpressed or mutated in CRPC, bicalutamide may act as a partial agonist and promote cancer cell growth . Consequently, the use of novel AR signalling inhibitors that lack this property, such as MDV3100, could result in improved response rates when combined with PI3K/mTOR pathway inhibitors .
The combination of RAD001 plus bicalutamide was well tolerated. Previously reported adverse events associated with RAD001 include non-infectious pneumonitis, infections and stomatitis, as well as metabolic abnormalities such as hyperglycaemia and hypercholesterolaemia [30,52]. In the present study, the most common toxicity was grade 1/2 mucositis, which was observed in 19 (53%) patients. There were three (8%) patients who had grade 3 hyperglycaemia and three (8%) patients who had grade 1/2 hypercholesterolaemia. None of the three patients who developed grade 3 hyperglycaemia had a history of diabetes. There were three patients (8%) who developed grade 1/2 pneumonitis/pulmonary infiltrates during treatment with RAD001. All of the adverse events observed in the present study are either similar to or less than the previously reported toxicity in kidney cancer studies.
In conclusion, a phase II trial of RAD001 and bicalutamide showed that the combination regimen was well tolerated, although it showed very limited activity in men with CRPC. The low activity may be a result of RAD001 inhibition of negative feedback loops, resulting in Akt and mitogen-activated protein kinase activation that could counteract potential synergistic effects of the combination treatment. Insufficient AR inhibition or partial agonist activity by bicalutamide may also have played a role. This particular combination therapy will not be pursued further; however, the concept of AR inhibition combined with PI3K-Akt pathway inhibition remains valid. With the next generation of PI3K-Akt inhibitors and AR signalling inhibitors now available, rational combinations will be tested in CRPC.