Novel cytotoxic and biological agents for prostate cancer: Where
`will the money be in 2005?
`
`John M. Strother a, Tomasz M. Beer a,*, Robert Dreicer b
`
`a Division of Hematology and Medical Oncology, Oregon Health and Science University, Mail Code CR-145,
`3181 SW Sam Jackson Park Road, Portland, OR 97239, USA
`b Cleveland Clinic Lerner College of Medicine, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195, USA
`
`Received 18 August 2004; received in revised form 7 February 2005; accepted 8 February 2005
`Available online 10 March 2005
`
`Abstract
`
`In 2004, docetaxel-based chemotherapy became the first treatment capable of extending life in androgen-independent prostate
`cancer. The era of therapeutic nihilism in this disease has thus been put to rest and a broad range of agents is being tested with
`the goal of improving on the successes of 2004. Lessons learned from other tumour types will need to be applied to prostate cancer
`in order to harness the bounty of available ideas. Target amplification or activating mutations and not merely the presence of a
`target are likely to be important to the success of targeted agents. Thus, the promise of the current crop of targeted agents is most
`likely to be realised when pursued in the context of well-credentialed targets and tested in highly translational clinical trials that are
`capable not only of assessing tumour response, but also of evaluating the status of the targeted pathway. The most promising agents
`in clinical development are reviewed.
`Ó 2005 Elsevier Ltd. All rights reserved.
`
`Keywords: Prostate cancer; Chemotherapy; Targeted therapy; Novel agents
`
`1. Introduction
`
`2. Current cytotoxic agents
`
`In 2004, docetaxel became the first drug to prolong sur-
`vival of patients with androgen-independent prostate
`cancer (AIPC) and the long-held view that AIPC is an
`untreatable disease was put to rest. The advance seen with
`docetaxel is critically important because it demonstrates
`that progress in this disease is possible and it establishes
`a clear standard of care upon which future studies will
`be built. A substantial number of investigational agents,
`belonging to a wide range of drug classes and targeting
`a broad range of cancer pathways are in clinical develop-
`ment. Thus, the future has never been brighter in treating
`this disease, which is the second-leading cancer killer of
`men in the United States of America (USA) and in 2000
`was the cause of more than 200,000 deaths worldwide.
`
`0959-8049/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
`
`The range of currently available cytotoxic agents has
`been summarised in detail by Bhandari, Petrylak and
`Hussain elsewhere in this volume. Active agents include
`mitoxantrone, the taxanes (paclitaxel and docetaxel,
`which seem to illustrate some schedule dependency)
`and estramustine [1–24]. The level of reported anti-can-
`cer efficacy has depended somewhat on the case selection
`characteristics, the nature of ÔresponseÕ (subjective or
`objective) and the use of surrogate markers of response.
`
`3. New cytotoxic agents
`
`3.1. SB-715992
`
`Mitotic kinesins, which play essential roles in the
`assembly and function of
`the mitotic spindle, are
`
`JANSSEN EXHIBIT 2032
`Mylan v. Janssen IPR2016-01332
`
`

`

`J.M. Strother et al. / European Journal of Cancer 41 (2005) 954–964
`
`955
`
`expressed preferentially in neoplastic cells. They repre-
`sent novel targets for cancer treatment [25]. Inhibition
`of the mitotic kinesin spindle protein (KSP) results in
`formation of monopolar spindles within single cells
`and the induction of apoptosis in pre-clinical studies.
`SB-715992 is the first KSP inhibitor to enter clinical tri-
`als and has a broad spectrum of activity in pre-clinical
`models of cancer, including models that are refractory
`to chemotherapy and prostate cancer models. This agent
`has completed phase I evaluation and is entering phase
`II trials. The Southwest Oncology Group (SWOG) will
`be evaluating this agent in patients who have progressed
`on docetaxel-containing therapy.
`
`3.2. Epothilones
`
`The epothilones are a new class of cytotoxic agents
`with anti-neoplastic activity in pre-clinical models of a
`range of tumours insensitive or resistant to paclitaxel;
`their mechanism of action is microtubular stabilisation
`resulting in mitotic arrest [26]. EPO906 is a novel epothi-
`lone that is not a substrate for multidrug-resistance pro-
`tein. A phase IIa trial of weekly EPO906 in 37 patients
`with previously-treated AIPC found that it was well tol-
`erated; the most common adverse events were gastroin-
`testinal. Twenty-two percent of patients responded by
`prostate-specific antigen (PSA) criteria and 4 out of 20
`patients with measurable disease responded [27]. Two
`recent phase II studies demonstrate that the epothilone
`analogue BMS-247550 has activity in chemotherapy-
`naı¨ve patients with metastatic AIPC. SWOG 0111
`showed 41% PSA and 30% measurable disease response
`rates in patients with chemotherapy-naı¨ve AIPC [28].
`Another multi-institutional phase II study reported
`56% PSA and 23% measurable disease response rates
`for BMS-247550 alone and 69% PSA and 44% measur-
`able disease response rates for the combination of BMS-
`247550 with estramustine [29]. The most frequent grade
`3 toxicities in both of these studies were fatigue and sen-
`sory neuropathy. These studies demonstrate activity for
`this agent, although its place in the treatment of AIPC is
`unclear in light of current docetaxel results. With the
`establishment of docetaxel-based chemotherapy as
`front-line therapy, suggestive evidence that there is some
`degree of non-cross-resistance with epothilones becomes
`increasingly relevant [30].
`
`3.3. Satraplatin
`
`Satraplatin is a novel oral platinum complex that has
`shown activity against AIPC in cisplatin-resistant human
`tumour lines and in phase I trials. Satraplatin plus pred-
`nisone was compared with prednisone alone in 50 chemo-
`therapy-naı¨ve AIPC patients. The satraplatin arm had a
`better progression-free survival (median 5.2 months
`versus 2.5 months, P = 0.023) and a higher frequency of
`
`PSA response (33.3% versus 8.7%; P = 0.046) [31]. A
`phase III trial of satraplatin plus prednisone versus
`prednisone as second-line chemotherapy is underway.
`
`3.4. Amonafide
`
`Amonafide is a synthetic imide anti-neoplastic agent
`with DNA intercalative properties that demonstrated
`significant activity in pre-clinical studies and some activ-
`ity in phase I trials, including at least one partial re-
`sponse in a patient with prostate cancer [32]. The drug
`is extensively metabolised and detected in plasma and
`urine. Its toxicity has previously been correlated to the
`formation of an active metabolite, N-acetyl-amonafide.
`In phase I studies using various administration sched-
`ules of amonafide, myelosuppression was the dose limit-
`ing toxicity. A phase II trial of amonafide at a dose of
`225 mg/m2 i.v. daily for 5 d was conducted by SWOG.
`Forty-three evaluable patients with measurable AIPC
`were treated. The most common toxicities were haema-
`tological including leucopaenia (72%), granulocytopae-
`nia (32.6%) and thrombocytopaenia (44.2%). There
`were no complete responses and 5 partial responses, giv-
`ing an overall response rate of 12% [33].
`Subsequent work determined that N-acetyltransfer-
`ase enzyme polymorphisms play an important role in
`the metabolism of amonafide (fast acetylators are at
`great risk of haematological toxicity from amonafide
`therapy). Ratain and colleagues [34] proposed a phar-
`macodynamic model based on acetylator phenotype
`using caffeine as a probe to optimise amonafide dosing.
`Observations from a phase II trial of amonafide in pa-
`tients with metastatic breast cancer that response rates
`appeared higher in patients experiencing more severe
`myelosuppression led to the hypothesis that optimisa-
`tion of amonafide dosing might yield evidence of higher
`anti-tumour activity [35]. Using a modification of Ra-
`tainÕs pharmacodynamic model, a phase I/II trial of
`NAT2 phenotype-based dosing of amonafide is ongoing
`in patients with metastatic AIPC treated with up to one
`prior systemic chemotherapy regimen.
`
`4. Nuclear receptor ligands
`
`4.1. Peroxisome proliferators-activated receptor gamma
`(PPARc) ligands
`
`Peroxisome proliferators-activated receptor gamma
`(PPARc) ligands are members of the nuclear receptor
`superfamily of ligand-dependent transcription factors
`and play a central role in lipid metabolism and adipo-
`cyte regulation. PPARs are widely expressed by prostate
`cancer cells. Synthetic PPARc agonists inhibit prostate
`cancer growth in vitro [36]. The thiazolidinediones
`(including troglitazone, rosiglitazone and piaglitazone)
`
`

`

`956
`
`J.M. Strother et al. / European Journal of Cancer 41 (2005) 954–964
`
`are a class of synthetic ligands for PPARc that have
`been developed primarily for the treatment of insulin-
`resistant type II diabetes mellitus. A phase II study of
`troglitazone in patients with asymptomatic advanced
`prostate cancer included 12 men with androgen-depen-
`dent disease and 29 men with AIPC. In the androgen-
`dependent group, 1 of 12 patients had a >50% decrease
`in PSA and 3 of 12 patients had a <50% decrease. In the
`AIPC group, no patient had a >50% decrease in PSA
`but 4 of 29 patients had a <50% decrease [37]. A second
`randomised, placebo-controlled trial of rosiglitazone
`was initiated. A total of 105 men with rising PSA after
`local therapy were randomised with a primary end-point
`of post-treatment
`change
`in PSA doubling time
`(PSADT). There was no difference between the 2 arms
`in the primary efficacy analysis. Complicating analysis
`was the higher-than-expected (40%) response rate in
`the placebo arm [38,39]. Given these results, it is unlikely
`that PPARc ligands alone will play a role in prostate
`cancer therapy. However, as the role of this receptor
`in prostate cancer progression is better understood,
`rationally designed combination of drugs that include
`PPARc ligands may emerge.
`
`5. Calcitriol
`
`Calcitriol, the most active metabolite of vitamin D,
`has significant anti-neoplastic activity in a broad range
`of pre-clinical cancer models including in vitro and
`in vivo activity in several models of prostate cancer
`[40–46]. Several mechanisms of activity have been pro-
`posed. These include inhibition of proliferation associ-
`ated with cell cycle arrest and,
`in some models,
`differentiation, reduction in invasiveness and angiogene-
`sis, and induction of apoptosis. Further synergistic and/
`or additive effects with cytotoxic chemotherapy, radia-
`tion and other cancer drugs have been reported. A con-
`temporary comprehensive review of
`spectrum and
`mechanisms of activity is available [47].
`Significantly supraphysiological concentrations of
`calcitriol are required for anti-neoplastic effects. The
`development of high-dose regimens, made possible by
`intermittent dosing of calcitriol [48,49], has led to several
`clinical trials in prostate cancer. Single agent weekly cal-
`citriol in hormone-naı¨ve prostate cancer appeared to
`lengthen PSADT in an uncontrolled study [50]. Weekly
`calcitriol yielded encouraging phase II results when
`added to docetaxel. Eighty-one percent of 37 patients
`had confirmed >50% reduction in serum PSA in a single
`institution trial [51]. These results were encouraging rel-
`ative to historical experience with single agent docetaxel
`and led to the development of a randomised trial of
`docetaxel with calcitriol or placebo. This trial (dubbed
`ASCENT) is using DN-101, a highly concentrated for-
`mulation of calcitriol developed specifically for oncol-
`
`ogy. ASCENT accrual has been completed and results
`should provide robust information about the potential
`of calcitriol to enhance chemotherapy for prostate can-
`cer. Studies of calcitriol with dexamethasone, paclitaxel
`and carboplatin are underway as well [52].
`The development of calcitriol analogues that might
`have anti-neoplastic activity but cause less hypercalca-
`emia has been proposed as another strategy to target
`the vitamin D receptor for cancer treatment. Several
`hundred such analogues have been synthesised [53–55].
`Several analogues, dosed daily, have entered clinical tri-
`als. Two of 25 AIPC patients had objective partial re-
`sponses to 1-alpha-hydroxyvitamin D2 [56]. Further
`studies of calcitriol analogues in prostate cancer are
`underway.
`
`6. Growth factor receptor antagonists
`
`6.1. Epidermal growth factor receptor
`
`The epidermal growth factor (EGF) system has been
`implicated as one of the mitogenic signals that may drive
`progression of AIPC. EGF receptor (EGFR) signalling
`pathways may be involved in prostate cancer invasion
`and angiogenesis [57,58]. Cetuximab is an anti-EGFR
`monoclonal antibody with high affinity that prevents
`activation of receptor tyrosine kinase. In pre-clinical
`studies cetuximab has demonstrated activity in a PC-
`3M-LN4 orthotopic model of AIPC, both as a single
`agent and in combination with paclitaxel [59]. A phase
`Ib/IIa study of cetuximab and doxorubicin in 18 patients
`with AIPC documented stabilisation in PSA for a med-
`ian of 4 months in 5 patients (28%) [60]. Gefitinib (Ires-
`sa) is an oral small molecule EGFR tyrosine kinase
`inhibitor that has demonstrated clinical utility in non-
`small cell lung cancer [59,61]. However, a phase II trial
`of gefitinib in AIPC has preliminarily found infrequent
`PSA responses and early progression in many patients
`[62]. Thus the EGF receptor is a target that has yet to
`be vetted in a clinical trial and its role in the future of
`prostate cancer therapy remains to be determined.
`
`6.2. HER-2
`
`Trastuzumab is a humanised monoclonal antibody
`that targets the extracellular domain of HER-2 and con-
`fers a survival benefit when combined with chemother-
`apy in patients with metastatic breast carcinoma that
`overexpresses HER-2 [63]. Subsequent analyses have
`suggested that most of the activity of this agent is seen
`in patients with HER-2 gene amplification and fluores-
`cent in situ hybridisation has largely replaced immuno-
`histochemistry as a tool for identifying breast cancer
`patients who are likely to benefit from trastuzumab.
`HER-2 expression in prostate cancer, as determined by
`
`

`

`J.M. Strother et al. / European Journal of Cancer 41 (2005) 954–964
`
`957
`
`immunohistochemistry, has produced very heteroge-
`neous results [64,65]. In one series of 70 patients with
`metastatic disease, the overall rate of prostatic-tissue
`sections presenting positive HER-2 immunostaining
`was 64.3% [66] and in another study all 39 patients with
`prostate cancer showed positive staining by immunohis-
`tochemistry [67]. Conversely, two other series identified
`HER-2 positivity in 40% (32 of 79) and 0% (0 of 40) of
`patients with prostate cancer [68,69]. A phase II trial in
`patients with progressive prostate cancer found that
`trastuzumab alone produced no responses in 23 patients
`with either HER-2 negative (17 patients) or HER-2 po-
`sitive (6 patients) disease as determined by immunohis-
`tochemistry. Additionally, four patients with HER-2
`positive disease received combination therapy with trast-
`uzumab and paclitaxel;
`two patients had a partial
`response by PSA and two patients had disease progres-
`sion [65]. A second phase II trial of trastuzumab and
`docetaxel in HRPC was closed early when only 7/72 pa-
`tients were found to have HER-2 overexpression by
`immunohistochemistry and none of the 4 patients trea-
`ted with trastuzumab alone demonstrated response
`[64]. Thus it appears that trastuzumab is unlikely to gain
`a foothold in prostate cancer therapy.
`
`6.3. PDGF receptor
`
`Platelet-derived growth factor (PDGF) and its recep-
`tor (PDGF-R) have been implicated as both paracrine
`and autocrine mediators of prostate cancer progression
`[70,71]. The PDGF-R is expressed in 80% of AIPC le-
`sions [72]. Imatinib (Gleevec) is a specific inhibitor of
`the BCR-ABL, c-kit and PDGF receptor tyrosine
`kinases. In a mouse model of prostate cancer bone
`metastases, PDGF-R inhibition by imatinib resulted in
`preservation of bone structure,
`significant
`tumour
`growth inhibition, a significant increase in apoptosis of
`malignant cells and a decrease in lymph node metastases
`[61]. Therefore, PDGF-R would appear to be a promis-
`ing target in prostate cancer therapy. A small study of 15
`patients with AIPC and painful bony metastases treated
`with imatinib and zoledronic acid failed to demonstrate
`PSA response and had no palliative or clinical activity
`[73]. A phase II trial of imatinib as a single agent in 17
`patients with hormone-naı¨ve PSA progression after lo-
`cal therapy demonstrated biochemical stable disease in
`6 patients (35%) [74]. At the present time, it is not clear
`whether this class of agents will prove useful. Although
`little single agent activity was seen, imatinib may prove
`more useful in combination therapy. A study at MD
`Anderson is examining imatinib in combination with
`docetaxel. As with all targeted therapies,
`it will be
`important that future studies include determination of
`PDGF-R expression whenever feasible, so that clinical
`outcomes can be interpreted in the context of knowledge
`about the status of the target.
`
`6.4. IGF receptor
`
`Interest in the IGF system and prostate cancer was
`stimulated by epidemiological investigations that sug-
`gest a relationship between elevated IGF-1 levels and
`prostate cancer risk [75,76] and recognition that signal-
`ling through the IGF-1 receptor inhibits apoptosis. Tar-
`geting the IGF signalling system is therefore a promising
`strategy in prostate cancer prevention and therapy.
`Agents that target the IGF system are in pre-clinical
`development.
`
`7. Anti-angiogenic agents
`
`7.1. Thalidomide
`
`As is the case for most solid tumours, the recruit-
`ment of blood vessels is an important step in progres-
`sion and metastasis of prostate cancer. Inhibition of
`this complex process is an attractive approach to treat-
`ment and many anti-angiogenic agents are currently in
`clinical development. Thalidomide is a sedative, anti-
`inflammatory
`and immunosuppressive
`agent
`that
`blocks
`the activity of angiogenic agents
`including
`bFGF, VEGF and IL-6. The limb defects attributed
`to its teratogenic effects have been postulated to be
`secondary to an inhibition of blood vessel growth in
`the developing foetal limb bud. A recent clinical trial
`of thalidomide for AIPC randomised 63 patients to a
`low-dose arm of 200 mg daily or a high-dose arm,
`escalating to the highest tolerated dose (up to 1200
`mg) [77]. Only 13 patients were treated on the high
`dose arm that was terminated early due to a combina-
`tion of lack of efficacy and poor tolerance. In all, 9
`patients (15%) showed a PSA decline of over 50%,
`all of whom were in the low-dose arm. A second phase
`II study of low-dose thalidomide in 20 patients with
`AIPC identified 3 men (15%) with a decline in serum
`PSA of at least 50% [78]. A randomised phase II trial
`of docetaxel with or without thalidomide 200 mg daily
`in 53 patients with chemotherapy-naı¨ve AIPC reported
`that 19 of 36 patients (53%) in the combination arm
`and 6 of 17 (35%) receiving docetaxel alone had a
`PSA decrease of at least 50%. Combination therapy
`also resulted in improved progression-free survival
`(5.9 versus 3.7 months) and overall survival at 18
`months (68.2% versus 42.9%) [79]. Therapy was rela-
`tively well tolerated, although 12 of the first 43 pa-
`tients in the combined treatment group developed
`thromboembolic events; all subsequent patients re-
`ceived prophylactic anticoagulation with low molecular
`weight heparin and no further thromboembolic events
`were noted. Thus thalidomide appears to have modest
`single-agent activity and its activity with docetaxel
`appears additive.
`
`

`

`958
`
`J.M. Strother et al. / European Journal of Cancer 41 (2005) 954–964
`
`7.2. Lenalidomide
`
`CC-5013 (lenalidomide, Revlimid) is an immuno-
`modulatory thalidomide analogue, which in animal
`models is both anti-angiogenic and non-teratogenic [80].
`A phase I trial in patients with melanoma and other
`solid tumours provided evidence that CC-5013 can
`boost Th1-type cellular immunity. Figg and colleagues
`are currently conducting a phase I trial of CC-5013 in
`AIPC patients [81].
`
`7.3. Bevacizumab
`
`Bevacizumab (Avastin) is a recombinant, humanised
`anti-vascular endothelial growth factor (VEGF) mono-
`clonal antibody that blocks the binding of VEGF to
`its receptors. Recently approved in the United States
`for first-line use in metastatic colorectal cancer, bev-
`acizumab is also being examined in prostate cancer.
`CALGB 90006 is an ongoing phase II trial of bev-
`acizumab, docetaxel and estramustine in patients with
`AIPC. Initial results include a confirmed >50% decline
`in PSA in 13 of 20 patients (65%) with sufficient data
`and a partial response in 9 of 17 patients (53%) with
`measurable disease [82].
`
`7.4. Other
`
`PTK787/ZK 222584 is an oral angiogenesis antago-
`nist that inhibits all known VEGF receptors. A phase
`I study in 23 patients with AIPC showed that 3 (13%)
`had >40% reductions in PSA [83]. A recent randomised
`phase II trial of two doses of the matrix metalloprotein-
`ase inhibitor BMS-275291 in 80 patients with AIPC re-
`sulted in no responders [84].
`
`8. Proapoptotic agents
`
`8.1. Bcl-2 antisense
`
`The mitochondrial-associated protein Bcl-2 confers
`resistance to apoptosis and is overexpressed in AIPC
`[85,86]. G3139 (oblimersen sodium) is an 18-mer phosp-
`horothioate antisense oligonucleotide directed to the
`first six codons of the initiating sequence of the human
`bcl-2 mRNA [87]. In pre-clinical prostate cancer models,
`Bcl-2 antisense oligonucleotides have inhibited expres-
`sion of Bcl-2, delayed development of androgen inde-
`pendence and enhanced the effects of chemotherapy
`[88–91]. A phase II study of G3139 and docetaxel in
`31 men with metastatic AIPC (8 had received prior che-
`motherapy) yielded a confirmed >50% PSA reduction in
`15 patients (48%) and a PR in 4 out of 15 patients (27%)
`with measurable disease. Toxicities included grade 3–4
`neutropaenia in 42% of patients [92]. As this was a small
`
`study, it is difficult to determine whether this level of
`activity represents an improvement from that expected
`with docetaxel alone.
`
`8.2. Selective apoptotic anti-neoplastic drugs (SAANDs)
`
`Exisulind (sulindac sulphone), the oxidative metabo-
`lite of sulindac, is a member of a class of novel drugs
`that inhibit growth and induce apoptosis in prostate
`cancer cell lines by specifically inhibiting cyclic GMP
`phosphodiesterases but not cyclooxygenase-1 or -2
`[93,94]. Goluboff and colleagues [95] randomised 96 pa-
`tients with rising PSA levels after radical prostatectomy
`to 12 months of exisulind 250 mg orally or placebo twice
`daily and demonstrated that exisulind significantly sup-
`pressed the increase in PSA in all patients and prolonged
`the PSA doubling times (PSADT) in high-risk patients.
`Reduction in the PSADT persisted through 24-month
`follow-up on an open-label extension [96]. Additionally,
`pre-clinical studies have suggested synergistic interac-
`tions between exisulind and several chemotherapeutic
`agents [97–100] and a phase II study examining the com-
`bination of exisulind and docetaxel in men with AIPC
`has completed accrual [101].
`CP-461 is an exisulind analogue with broad anti-
`tumour activity. In a phase I study, 21 patients with a
`range of solid tumours (not including prostate cancer)
`received CP-461 twice daily for 28 d. Therapy was well
`tolerated and 4 patients exhibited disease stability after
`two cycles of treatment [93]. A phase II study of
`CP-461 in prostate cancer is underway.
`
`8.3. Clusterin antisense oligonucleotide
`
`Clusterin is a survival gene whose expression in-
`creases markedly in response to androgen-deprivation
`therapy [102]. The expression of clusterin confers a che-
`motherapy-resistant phenotype, probably due to reduc-
`tion
`in
`treatment-induced
`apoptosis
`[103,104].
`Oligonucleotide antisense molecules to clusterin have
`been developed and have entered clinical
`trials in
`prostate cancer [105]. Development in combination with
`chemotherapy is expected.
`
`9. Other novel targeted agents
`
`9.1. Bortezomib
`
`Bortezomib (PS-341, Velcade) is a novel boronic acid
`dipeptide
`that
`inhibits
`26S proteasome
`activity.
`Although this agent has been most extensively studied
`in multiple myeloma, there is some intriguing evidence
`of its potential utility in prostate cancer. Several investi-
`gators have demonstrated that bortezomib is active in
`
`

`

`J.M. Strother et al. / European Journal of Cancer 41 (2005) 954–964
`
`959
`
`pre-clinical models of human prostate cancer. Ikezoe
`and colleagues demonstrated that bortezomib blocks
`the androgen receptor signalling pathway and induces
`growth arrest and apoptosis in LNCaP cells [106]. Wil-
`liams and colleagues found that bortezomib appeared
`to have anti-tumour activity in both LNCaP and PC-3
`lines, but with distinctly different underlying mecha-
`nisms. The LNCaP lines demonstrated reduced micro-
`vessel density and VEGF secretion and high levels of
`apoptosis, while the PC-3 lines had direct increases in tu-
`mour cell death [107].
`Papandreou and colleagues conducted a phase I trial
`of bortezomib in 48 patients with AIPC. Doses ranging
`from 0.13 to 2 mg/m2 were administered weekly · 4,
`with cycles repeated every 5 weeks. Dose-limiting toxic-
`ity was noted in 2 of 5 patients treated at the 2 mg/m2
`dose level (grade 3 diarrhoea in both patients, grade 3
`syncope and hypotension in 1 patient). The maximally
`tolerated dose (MTD) recommended with this schedule
`was 1.6 mg/m2. Twenty-four patients were treated at
`the MTD (P1.45 mg/m2).
`dosages in the range of
`Two patients (8%) demonstrated a P50% decline in
`PSA and 1 of 9 (11%) patients with measurable nodal
`disease achieved a partial response [108]. Based on pre-
`clinical data suggesting potential
`synergism with
`chemotherapy, phase II trials of bortezomib in combi-
`nation with docetaxel were initiated [109–112]. Dreicer
`and colleagues conducted a phase I/II trial (the phase
`II portion consisted of three separate cohorts of pa-
`tients treated at various dosages) of bortezomib plus
`docetaxel in 100 patients with metastatic AIPC. Data
`from the phase I portion of the study and the low ex-
`panded cohort were presented recently. Docetaxel was
`given at 25, 30, 35 and 40 mg/m2 (cohorts 1–4) i.v. over
`30 min on days 1 and 8. For the first 4 cohorts, bort-
`ezomib was given at a fixed dose of 1.3 mg/m2 i.v. push
`on days 2 and 9, while a fifth cohort evaluated a higher
`dose of bortezomib at 1.6 mg/m2 with 40 mg/m2 of
`docetaxel, with treatment in all cohorts repeated every
`21 d. In the phase I portion of the study no maximally
`tolerated dose was determined. A total of 32 patients
`were enrolled in the low expanded cohort. Twenty-
`one patients had prior chemotherapy, 11 with a tax-
`ane-containing regimen. Therapy was well tolerated;
`anaemia was the most common grade 3 event (2 if 32
`patients). Six of 25 patients (24%) evaluable for re-
`sponse had a confirmed PSA decrease of P50% from
`baseline. Three of 13 (23%) patients with measurable
`disease achieved a partial response [113].
`
`9.2. Atrasentan
`
`Endothelin-1 is a potent vasoconstrictor produced
`by prostate cancer and appears to have a role in pros-
`tate cancer progression and morbidity. Atrasentan is
`an oral, highly selective endothelin receptor antagonist.
`
`In a phase III, placebo-controlled trial that involved
`809 patients with metastatic AIPC, atrasentan did
`not prolong time to clinical progression (HR 1.14,
`95% CI 0.98–1.34) in an intent-to-treat analysis. Atras-
`entan therapy was associated with smaller increases in
`serum PSA and markers of bone turnover. A modest
`delay in time to clinical progression (HR 1.26, 95%
`CI 1.06–1.50) was noted in the protocol-compliant pa-
`tients [114].
`An intent-to-treat meta-analysis that included this
`study as well as a prior randomised phase II study dem-
`onstrated a delay in disease progression (HR 1.19,
`P = 0.013), but the overall clinical benefit was small.
`These studies are characterised by rapid early progres-
`sion of a subset of patients, which contributed to the dif-
`ficulty detecting a benefit in the individual trials. A
`phase III study of this agent in earlier stages of AIPC
`will probably determine its fate.
`
`10. Immunotherapeutic agents
`
`10.1. GM-CSF
`
`Granulocyte-macrophage colony stimulating factor
`(GM-CSF) is a cytokine that regulates the proliferation
`and differentiation of myeloid precursor cells and en-
`hances the function of both mature granulocytes and
`mononuclear phagocytes. GM-CSF can influence the
`recruitment, activation and survival of macrophages
`and dendritic cells. Additionally, experimental evidence
`suggests that GM-CSF may increase the efficiency of
`tumour antigen presentation and the subsequent acti-
`vation of
`tumour-specific cytotoxic T lymphocytes
`[115,116]. Small and colleagues administered GM-
`CSF (250 lg/d daily · 2 weeks followed by three times
`a week administration)
`to patients with metastatic
`AIPC to evaluate their hypothesis that apoptotic tu-
`mour cells could provide a source of tumour antigen,
`which might lead to T-cell cross-priming by dendritic
`cells. Therapy was well tolerated and a small subset
`of the patients experienced a greater than 50% decline
`in PSA, with one patient demonstrating both a pro-
`longed PSA response and objective improvement in
`bone scan imaging [117]. Dreicer and colleagues treated
`16 patients with advanced prostate cancer (7 hormon-
`ally-naı¨ve and 9 androgen-independent) with GM-
`CSF administered subcutaneously at 250 lg thrice
`weekly for up to 6 months. Although no patient
`achieved an objective response, 6 patients demon-
`strated a 10–15% decline in their baseline PSA that
`was maintained during the entire treatment period.
`Five of these 6 patients demonstrated a rise in their
`PSA following study completion. Therapy was well
`tolerated, with only one grade 3 event, which was
`probably not treatment-related [118].
`
`

`

`960
`
`J.M. Strother et al. / European Journal of Cancer 41 (2005) 954–964
`
`10.2. GVAX
`
`10.5. MVA-MUC1-IL2 vaccine
`
`GVAX cancer vaccines are comprised of tumour cells
`that have been genetically modified to secrete GM-CSF.
`GVAX was studied in a phase II dose escalation trial.
`Nineteen patients were treated at the highest dose. Of
`these patients, 6 (32%) had modest declines in serum
`PSA and 87% demonstrated immune response [119].
`Phase III studies of GVAX are planned.
`
`10.3. Provenge
`
`Provenge consists of autologous dendritic cells
`loaded ex vivo with a recombinant fusion protein of
`prostatic acid phosphatase (PAP) linked to GM-CSF.
`In early studies, Provenge was shown to induce immune
`responses to PAP in 38% of patients and occasional PSA
`reductions in excess of 50% [120]. A single durable re-
`sponse has been reported [121]. In the initial randomised
`study in AIPC, Provenge failed to improve progression-
`free survival, but a benefit in progression-free survival
`and overall survival was seen in a subset of patients
`whose Gleason score was less than 8 [122]. As a result,
`Provenge is now being evaluated in patients with meta-
`static AIPC that at initial diagnosis did not exceed Glea-
`son 7. It is also being tested in a placebo-controlled
`randomised study in patients with a rising serum PSA
`after prostatectomy.
`
`Over-expression, non-polarity and under-glycosyla-
`tion of the mucin glycoprotein MUC1 is associated
`with many epithelial neoplasms, making MUC1 a po-
`tential
`target of vaccine
`immunotherapy. MUC1
`expression has been reported in 60–90% of prostate
`cancer specimens evaluated [124,125]. Modified vac-
`cinia ankara (MVA) is a highly attenuated vaccinia
`virus that is non-propagative in most mammalian cells.
`TG4010 is a recombinant MVA expressing MUC1 and
`IL-2 that demonstrated an excellent safety profile in
`phase I evaluations. Interim results from a randomised
`phase II trial of two different schedules of TG4010 in
`patients with hormone-naı¨ve biochemical failure have
`been reported recently. Eligible patients were required
`to have a PSA doubling time of less than 10 months,
`with an absolute value of 2.0 ng/ml or greater without
`evidence of metastatic disease. Patients were randomly
`assigned to receive either weekly injections of 108 pla-
`que-forming units for the first 6 weeks then injections
`every 3 weeks until week 36 or progressive disease, or
`the same dosage administered every 3 weeks until week
`36 or progressive disease. At the time of the interim
`assessment, 29 patients had been treated on study,
`with a significant change in PSA doubling time from
`pre-treatment assessment
`favouring arm 1 (weekly
`therapy) [126].
`
`10.4. Vaccinia virus/fowlpox virus
`
`10.6. J591 monoclonal antibody
`
`Various investigators have pursued immunotherapy
`with viral vectors since they can mimic natural infection
`and produce potent immune responses. Investigators
`from the Eastern Cooperative Oncology Group
`(ECOG) have recently reported a phase II trial designed
`to evaluate the tolerability and feasib