”3
`
`
`
`EXHIBIT
`
`
`
`PERGAMON
`
`European Journal of Cancer 38 (2002) 2347—2361
`
`Review
`
`
`
`European
`
`Journal of
`Cancer
`
`
`www.cjconline.com
`
`The role of new agents in the treatment of non-small cell lung cancer
`
`Linda E. Broker, Giuseppe Giaccone*
`Department of Medical Oncology, VU University Medical Center, De Boelelaan 1117. 1081 HV Amsterdam, The Netherlands
`
`Received 24 July 2002; accepted 29 July 2002
`
`Abstract
`
`Lung cancer is one of the most frequent causes of cancer deaths worldwide. Non-small cell lung cancer (NSCLC) accounts for
`approximately 80% of cases and no curative treatment is available for the advanced stages of disease (stages III and IV), which comprise
`the majority of cases. Current treatment regimens with standard chemotherapy offer only a limited survival benefit, and, therefore, the
`development of new therapeutic strategies is needed. Novel chemotherapeutic drugs such as the epothilones, MEN 10755 and S-l are
`being studied in patients with advanced stages of disease. Furthermore, a large number of therapies targeted against critical biological
`abnormalities in NSCLC are being investigated in clinical trials. The latter approach includes inhibition of growth factors, interference
`with abnormal signal transduction, inhibition of angiogenesis and gene replacement therapy. Promising results have thus far been
`obtained with some of these therapies. This review describes the role of new therapeutic agents in the treatment of NSCLC.
`© 2002 Elsevier Science Ltd. All rights reserved.
`
`Keywords: Lung cancer; New agents; Treatment; NSCLC
`
`1. Introduction
`
`Lung cancer is one of the most commonly occurring
`malignancies in the world and is the leading cause of
`cancer-related death in men. It is generally divided into
`small cell
`lung cancer (SCLC), which accounts for
`approximately 20% of all cases, and non-small cell lung
`cancer (NSCLC) that can be subdivided into squamous
`cell carcinoma, adenocarcinoma and large cell carci-
`noma and represents approximately 80% of all lung
`cancers [1].
`Surgery remains the sole curative treatment modality
`for patients with NSCLC. However, less than one third
`of patients are candidates for surgical exploration and
`more than 50% of them will eventually relapse [2].
`Chemotherapy is broadly used for advanced stages of
`NSCLC and usually consists of a platinum—containing
`compound (cisplatin or carboplatin) combined with
`gemcitabine, a taxane (paclitaxel or docetaxel) or
`vinorelbine [3]. A recent randomised study among 1207
`patients showed that four platinum-based combination
`regimens were similarly effective with a response rate of
`17—21% and a l-year survival rate of 31—36% in pre-
`
`* Corresponding author. Tel; +31-20-4444321;
`4444079.
`E-mail address: g.giaccone@vumc.nl (G. Giaccone).
`
`fax: +31—20-
`
`viously untreated patients with stage IIIB or IV NSCLC
`[4]. When compared with best supportive care, chemo-
`therapy offers only a limited survival benefit often at the
`cost of substantial toxicity [5,6].
`Chemotherapy has not substantially altered the long-
`term outcome for most lung cancer patients in the past
`decade and it is likely that the results of chemotherapy
`have reached a plateau [7]. Therefore, novel treatment
`strategies are urgently needed in advanced NSCLC.
`New ways to improve the results of current treatment
`regimens appear to be the use of novel chemothera—
`peutic agents with more favourable toxicity and activity
`profiles and the use of biological agents that target for
`example abnormal signal transduction pathways, either
`alone or
`in combination with chemotherapy. This
`review describes the current status of novel biological
`and chemotherapeutic drugs for
`the treatment of
`NSCLC.
`
`2. Novel chemotherapeutic agents in the treament of
`NSCLC
`
`there are not that many novel chemo-
`In general,
`therapeutic agents being developed. Most novel agents are
`in fact targeted to specific molecular alterations. However,
`here we will examine some of the more interesting
`
`0959-8049/02/5 - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
`PII: s09s9-8049(02)oo457—4
`
`OSI 2045
`APOTEX V. OSI
`|PR2016-01284
`
`OSI 2045
`APOTEX V. OSI
`IPR2016-01284
`
`

`

`41
`
`2348
`
`LE. Bré'ker, G. Giaccone/ European Journal of Cancer 38 (2002) 2347—2361
`
`agents with cytotoxicity as the major mechanism of
`action, which may have activity in NSCLC and be fur-
`ther developed for the treatment of this disease.
`
`2.1. Epothilones '
`
`Taxanes are widely used in the treatment of NSCLC
`both as single agents (second-line) and in combination
`with other chemotherapeutic drugs such as cisplatin and
`carboplatin (first-line). However, a major drawback of
`this group of agents is the presence or the development
`of resistance. Drug resistance to taxanes is mostly due
`to upregulation of P-glycoprotein that can pump the
`drug out of the cell, or mutations in the cellular target,
`B-tubulin [8].
`Epothilones are new compounds that are structurally
`unrelated to the taxanes, but have the same mechanism
`of action, and stabilise the microtubules more potently
`than paclitaxel [9]. Interestingly, the epothilones are not
`good substrates for P-glycoprotein and they are active
`in paclitaxel-resistant cell lines with certain B-tubulin
`mutations [9,10].
`EPO906, an epothilone B analogue produced by
`Novartis Pharma AG, has been investigated in several
`phase I clinical trials, in which one objective response
`was observed in 31 patients with advanced solid
`tumours
`[11,12]. Another
`epothilone B—derivative,
`BMS-247550 produced by Bristol Myers Squibb, has
`also shown clinical activity in various phase I trials. Its
`side-efiects consist of myelosuppression, neurotoxicity
`and gastrointestinal
`symptoms
`[13—15]. Preliminary
`results from a phase II trial with this agent when given
`as a single agent every three weeks in 22 advanced
`NSCLC patients who had failed first-line platinum-
`based chemotherapy, showed a response rate of 18%
`with stabilisation of disease in another 46% [16]. Fur-
`thermore, this agent proved to be effective in some tax-
`ane—refractory breast cancer patients [17]. These results
`indicate that the epothilones may provide a valuable
`contribution in the treatment of NSCLC.
`
`2.2. Other new chemotherapeutic drugs
`
`Glufosfamide is a new alkylating agent in which iso-
`phosphoramide mustard, the alkylating metabolite of
`ifosfamide, is linked to B—D-glucose [18,19], which leads
`to drug stabilisation and preferential uptake of the
`compound via the transmembrane transport system of
`glucose [20]. This targeting mechanism, together with
`the accelerated metabolic rate and increased glucose
`consumption of
`tumour cells,
`suggests potentially
`enhanced tumour selectivity for glufosfamide. Phase I
`trials in patients with advanced solid tumours demon-
`strated clinical activity with objective responses in sev-
`eral tumour types [21,22]. The dose-limiting toxicity of
`this agent consists of reversible renal tubular acidosis
`
`include neutropenia, nausea,
`side-effects
`and other
`vomiting and alopecia. In a phase II trial in 39 patients
`with advanced NSCLC, 3 partial responses among 31
`assessable patients were observed after administration
`of this drug [23].
`Anthracyclines have not been very active in NSCLC
`[24], but in recent years, novel anthracycline analogues
`have been developed, that have a broader activity and a
`more favourable toxicity spectrum [25]. Among these is
`MEN 10755, an anthracycline disaccharide, which is
`active in doxorubicin-resistant xenografts,
`including
`lung cancer models, and causes less cardiotoxicity than
`doxorubicin and epirubicin in preclinical studies [26—
`30]. Its side-effects consist of neutropenia and throm-
`bocytopenia, and pharmacokinetic analysis revealed a
`shorter half time life and a much smaller volume of
`
`distribution than doxorubicin [31,32]. The clinical
`activity of MEN 10755 in NSCLC has been studied in
`phase II trials in patients with advanced stages of dis—
`ease, but final results are not yet available.
`8-1 is a novel oral fluoropyrimidine derivative con-
`sisting of tegafur (FT), a prodrug of 5-fluorouracil (5-
`FU), and two modulators, potassium oxanate and
`CDHP that inhibit the degradation of FT—derived 5-
`FU. In preclinical studies, this drug showed anti-tumour
`activity in a variety of tumours, including lung cancer
`models [33,34]. Side-effects of 8-1 consist of gastro-
`intestinal symptoms and myelosuppression in a few
`cases [35]. In a phase II study involving 59 NSCLC
`patients with advanced stages of disease, S-l showed a
`response rate of 22% with a median response duration
`of 3.4 months, indicating that this drug may represent a
`valuable contribution to the treatment of NSCLC
`
`[36,37].
`
`3. Biological drugs for NSCLC
`
`3.1. Targeting erbB receptor pathways
`
`Growth factor dependency drives cell proliferation
`and diflerentiation and it is now clear that tumour cells
`
`may overcome normal regulatory inhibition, of pro-
`liferation by an enhanced or inappropriate activation of
`protein tyrosine kinases such as the erbB receptor
`family [38,39]. This family includes four distinct mem-
`bers: HERl (Epidermal Growth Factor (EGF)-receptor
`or c-erbBl), HER2 (neu or c—erbB2), HER3 (c-erbB3)
`and HER4 (c-erbB4), which share an overall structure
`of two cysteine-rich regions in the extracellular domain
`and a cytoplasmic kinase pocket with a carboxy-term—
`inal tail that is responsible for the diversified stimulation
`of downstream signal transduction pathways. However,
`HER3 lacks intrinsic kinase activity and no direct
`ligand has thus far been identified for HER2, which acts
`instead as a co-receptor [40—42]. Upon binding of the
`
`

`

`L.E. Bré‘ker, G. Giaccone / European Journal of Cancer 38 (2002) 2347—2361
`
`2349
`
`ligand, the intracellular tyrosine kinase domain is acti-
`vated, resulting in tyrosine autophosphorylation which
`ultimately triggers a cascade of diverse physiological
`responses involved in the mitogenic signal transduction
`of the cells [39,43,44].
`It has been known for several years that the EGF-
`receptor
`(EGF R)
`is overexpressed in many lung
`tumours [39]. Squamous cell carcinomas overexpress
`EGFR most frequently (85%) and strongly, whereas
`adenocarcinomas and large cell carcinomas are positive
`in approximately 65% of cases
`(reviewed in [45]).
`Although some studies have showed that EGFR-
`expression may be correlated with a decreased survival
`[46—51], others have indicated that EFGR-expression is
`of no prognostic significance [52—58]. Her2/neu is over-
`expressed in approximately 25% of NSCLC (reviewed
`in [59]) and correlates with increased metastatic poten-
`tial, drug resistance and poor prognosis [52,60—69].
`Based on these molecular characteristics of NSCLC,
`several agents have been developed that target the erbB-
`receptor family [70]. A summary of studies in lung can-
`cer with erbB-inhibitors is given in Table 1.
`ZD1839 and OSI 774 are small molecules that target
`the HERl receptor, and both have demonstrated single
`agent activity with objective responses in heavily pre-
`treated NSCLC patients, They are well absorbed after
`oral administration and can be given chronically. Both
`drugs have similar side-effects that are usually mild to
`moderate and consist of dose-dependent acne-like skin
`rash and diarrhoea which represents the dose-limiting
`toxicity. Other adverse events include anorexia, nausea
`and a transient rise of the liver transaminases [71—76].
`Recently, a large randomised phase II multicentre trial
`with two doses (250 and 500 mg/day) of ZD1839 has been
`reported in 210 stage III or IV NSCLC patients, who
`
`Table 1
`
`failed one or more prior treatment regimens. Side-effects
`were generally mild and consisted of. skin rash, pruritus
`and diarrhoea, but were significantly more common and
`severe at the higher dose level. Remarkably, both dose
`levels were equally efficacious with response rates of
`18.4 and 19%, respectively [77,78]. In another rando-
`mised phase II trial of 250 or 500 mg ZD 1839 in 216
`patients with more extensive pre-treatment,
`response
`rates were 8.8—11.8% [79,80]. OSI 774 was investigated
`in a phase II trial in 56 patients with advanced NSCLC
`who failed prior platinum-based chemotherapy, and,
`unlike for the studies with ZD1839, were selected based
`on overexpression of EGF R (210% positive cells). In
`this study, OSI 774 was given continuously at a fixed
`dose of 150 mg/day, which produced an acneiform rash
`in almost 80% of patients. The response rate was 11%,
`whereas 34% of patients had stable disease on this
`treatment [81].
`In general, ZD1839 and OSI 774 do not induce mye-
`losuppression, which makes them attracting for combi-
`nation studies with chemotherapy. Moreover,
`in vitro
`and in vivo studies have shown that ZD1839 potentiates
`the effect of several chemotherapeutic agents [82,83].
`Two large double-blind randomised studies with a
`combination of chemotherapy and ZD1839 have
`recently been completed,
`in which chemotherapy was
`added to 500 mg/day, 250 mg/day ZD1839 or to a pla-
`cebo [84]. Both studies accrued over 1100 patients and
`investigated two different treatment regimens: carbo-
`platin-paclitaxel, which is standard in North America
`and cisplatin-gemcitabine, which is more frequently
`employed in the rest of the world. Phase III trials of a
`similar design are currently underway using OSI 774
`[84]. Final analysis of the ZD1839 studies is expected
`shortly and the results of these important trials will help
`
`
`
` Current trials with inhibitors of the erbB-receptor pathway in NSCLC
`
`Compound
`Status
`Trial design
`
`Manaclonal antibodies
`
`HERI
`
`C225
`
`Phase II [96] Carboplatin/paclitaxel plus C225, first—line in advanced NSCLC
`Gemcitabine/carboplatin plus C225, first-line in advanced NSCLC
`Docetaxel plus C225, second-line in advanced NSCLC
`Dose-escalating study in patients with advanced solid tumours, including NSCLC [232]
`Dose—escalating study in patients with advanced solid tumours, including NSCLC [233]
`In progress [84]
`Chemotherapy + /—compound, first— and second-line in advanced NSCLC [100]
`
`Phase I
`ABX-EGF
`Phase I
`EMD72000
`Phase I
`GW2016
`Trastuzamab Phase II
`(Herceptin‘fi’)
`
`HERl—2
`HERZ
`
`Small molecules
`
`HERl
`
`ZD1839
`
`OSI 774
`
`Phase III,
`completed
`Phase III
`
`
`
`Carboplatin/paclitaxel or gemcitabine/cisplatin + /—compound (two different doses),
`first—line in advanced NSCLC [84]
`Carboplatin/paclitaxel or gemcitabine/cisplatin -- /—compound, first-line in
`advanced NSCLC [84]
`Carboplatin/paclitaxel or gemcitabine/cisplatin _- /—compound after 1—2 regimens in
`advanced NSCLC [84]
`
`pan-erbB Dose-escalating studies in patients with advanced solid tumours, including NSCLC [234—238] CI-1033 Phase I-II
`
`
`
`
`
`
`
`NSCLC, Non—small cell lung cancer. EGF, epidermal growth factor.
`
`

`

`a!)
`
`2350
`
`LE. Bré'ker, G. Giaccone/European Journal of Cancer 38 (2002) 2347—2361
`
`to define the role of these new agents in the management
`of NSCLC.
`
`C225, a chimeric antibody that blocks the tyrosine
`kinase activity of the erbBl-receptor, has been most
`extensively studied in head and neck and colorectal
`cancers. In head and neck cancers C225 has demon-
`
`strated considerable activity when given in combination
`with cisplatin to patients that are refractory to this drug
`[85—88]. Similar results were obtained when C225 was
`given in combination with CPT-ll to colorectal patients
`who were progressive on CPT-ll [89—91]. C225 has a
`half-life of approximately 7 days and can be given
`weekly with a loading dose of 400 mg/mz, followed by a
`maintenance dose of 250 mg/mz. Side-effects include
`acne-like skin rash, asthenia, and allergic reactions
`which occur in up to 4% of cases [92]. In preclinical
`NSCLC models, C225 was shown to potentiate the
`effect of chemotherapy and radiotherapy [93—95]. Phase
`II clinical trials in advanced NSCLC are ongoing to
`evaluate the efficacy and tolerability of combinations
`with chemotherapeutics [96].vPre1iminary results from a
`combination study with docetaxel show clinical activity
`with objective responses in 4 out of 20 patients with
`mild to moderate side-effects [97].
`Trastuzamab (Herceptin®), a humanised monoclonal
`antibody that binds to HER2,
`is registered for the
`treatment of breast cancer,
`in which it
`reached a
`response rate of 15% as single agent therapy in HER2-
`overexpressing tumours, which comprise 25—30% of
`breast cancers [98,99]. Side-effects consist of cardiac
`dysfunction in a few cases, which is worrying when
`trastuzamab is given in combination with anthracyclines
`[99]. Since HER2 is expressed in 20—66% of NSCLC,
`several trials have now been conducted to evaluate its
`
`effect in this tumour type [100]. However, the level of
`expression of HER2 is lower in NSCLC than in breast
`cancer, which makes the selection of patients rather
`cumbersome.
`In an Eastern Cooperative Oncology
`Group (ECOG) study in 139 patients with NSCLC, 50
`patients were found to be HER2-negative, whereas only
`9% of patients were strongly positive [101]. Krug and
`colleagues found even fewer patients showing over-
`expression of HER2: among 84 patients screened,
`HER2 was 3+ in only 6 patients and a total of 19%
`were 2 or 3+ [102]. In this phase II trial, previously
`untreated patients received trastuzamab with either
`docetaxel or paclitaxel. The overall response rate was
`26% and did not differ significantly according to the
`HER2 status (overexpression 20 vs. 28% in others). To
`evaluate the effect of trastuzamab more closely, a ran-
`domised trial was conducted, using either trastuzamab
`plus gemcitabine and cisplatin, or gemcitabine and cis-
`platin alone [103]. Among 103 patients,
`the overall
`response rates in the control and trastuzamab arms were
`41.2 and 36%, respectively, indicating that trastuzamab
`was not likely to add any benefit to the standard ther-
`
`apy. However, only very few patients with strongly
`HER2-positive disease were included in this trial. This
`problem makes the investigation of the potential effect
`of trastuzamab in NSCLC particularly difficult, since
`this drug does not seem to be effective in patients
`that do not have a high expression of HER2 [104].
`Activity may be limited to cases that are strongly
`HER2 (3+) and/or fluorescent
`in situ hybridisation
`(FISH) positive.
`
`3.2. Farnesyltransferase inhibitors
`
`Post-translational modifications of proteins by the
`addition of a farnesyl group is critical for the function
`of a number of proteins involved in signal transduction.
`The best—studied proteins in this respect are probably
`the Ras proteins that play pivotal roles in the control of
`normal and transformed growth [105,106]. In approxi-
`mately 25—30% of all adenocarcinomas, mutations are
`present
`in Ras genes leading to the production of
`mutated proteins that remain in a locked, active state
`thereby relaying uncontrolled proliferative
`signals
`[107,108]. Essential for Ras-activity is the transfer of
`farnesyl isoprenoid to the cytoplasmic Ras c—terminus, a
`process catalysed by an enzyme called farnesyltransfer-
`ase. This understanding has led to the development of
`farnesyltransferase inhibitors (FTIs)
`that block the
`growth stimulating and regulatory effects of Ras. Addi-
`tionally, FTIs affect many other proteins, such as Rho,
`Rheb and CENP-E and F, that need to be farnesylated
`for their growth regulatory function [109].
`The FTIs can be divided into three groups: farnesyl
`diphosphate (FDP) analogues, which compete with
`FDP,
`the substrate for
`farnesyltransferase; CAAX-
`mimetics that target the CAAX-portion of the Ras-
`protein, and agents which combine features of both.
`Drugs that are in current clinical investigation, BMS
`214662, SCH 66336, R115777 and L778,123, all belong
`to the second class [110]. Side-effects of these agents
`include gastrointestinal toxicity, fatigue and,
`less fre-
`quently, myelosuppression. Biological
`studies have
`shown decreased farnesyltransferase activity in normal
`tissues and tumour cells after intravenous (i.v.) admin-
`istration [111—117]. In a phase I setting, three partial
`responses in solid tumours were seen after treatment
`with SCH 66336, and one patient with NSCLC respon-
`ded to this agent
`[111,118]. Although R115777 has
`demonstrated clinical activity in acute myeloid leukae-
`mia (AML) and glioma patients, no responses were
`observed in a phase II trial in 44 patients with NSCLC
`[119—121]. Similarly, L778—123 was not effective in 23
`patients with this tumour type [113]. Combination
`studies of FTIs with chemotherapeutic drugs are
`underway in NSCLC. A large phase III trial is about to
`start with carboplatin/paclitaxel plus SCH 66336 versus
`placebo in untreated NSCLC patients.
`
`

`

`L.E. Broker, G. Giaccone/ European Journal of Cancer 38 (2002) 2347—2361
`
`2351
`
`3.3. Inhibition of angiogenesis
`
`Growth of new blood vessels is required for solid
`tumours to expand beyond a volume of 1—2 mm3 [122].
`This process of angiogenesis is regulated by a balance
`between pro- and anti-angiogenic factors: vascular
`endothelial growth factor (VEGF), basic and acidic
`fibroblast growth factor (bFGF and aFGF), platelet-
`derived endothelial growth factor (PD-ECGF) and oth-
`ers stimulate neovascularisation [123], whereas angios-
`tatin [124], endostatin [125] and thrombostatin [126,127]
`are important
`inhibitors of this process.
`Increased
`tumour angiogenesis, identified by increased microvessel
`density and VEGF and PD-ECGF-expression, is asso-
`ciated with a worse clinical outcome in many solid
`malignancies, including NSCLC [128—136].
`The fundamental goal of anti-angiogenic therapy is to
`induce a ‘dormancy state’ of primary tumours and their
`(micro) metastasis by returning foci of proliferating
`microvessels to their normal resting state and preventing
`their re-growth [137]. Clinical responses induced by
`anti-angiogenic agents may therefore primarily be
`expected from their combination with chemotherapy
`[138]. Inhibition of angiogenesis is actively under study
`in NSCLC (Table 2). Rhumab-VEGF (bevacizumab), a
`recombinant humanised antibody against VEGF, has
`been administered in human studies as a single agent
`and in combination with chemotherapy. Side-effects
`associated with the VEGF-antibody were generally mild
`and consisted of headache, asthenia,
`low-grade fever,
`arthralgia, nausea, vomiting and skin rash [139,140]. In
`a randomised phase II trial in 99 chemotherapy-naive
`patients with advanced NSCLC, the effect of two dif-
`ferent doses of anti-VEGF (7.5 mg/kg and 15.0 mg/kg)
`plus carboplatin/paclitaxel was compared with carbo-
`platin/paclitaxel alone [141]. Patients treated at
`the
`higher dose of Rhumab-VEGF experienced a higher
`response rate than the patients treated at the lower dose
`or with chemotherapy alone (34.3 versus 21.9 and 25%,
`respectively). Additionally,
`time to progression was
`longer in this group (207 versus 124 and 181 days,
`
`respectively). However, six episodes of life-threatening
`haemoptysis were observed, four of which were fatal.
`Furthermore, several episodes of non-life—threatening
`epistaxis were seen in patients receiving Rhumab-VEGF
`(31
`and 44% in the high and low dose arms,
`respectively). Remarkably, pulmonary haemorrhage
`was most common in patients with squamous cell
`histology and central, cavitated tumours, whereas it was
`relatively mild in two patients with non-squamous cell
`carcinoma [142]. A subset analysis of 78 patients with
`non-squamous cell carcinoma showed very promising
`results with median survival times of 77 versus 61 and
`
`53 weeks in patients treated with high dose Rhumab-
`VEGF versus low dose Rhumab-VEGF and chemo-
`
`respectively [143]. Based upon these
`therapy alone,
`results, ECOG initiated a phase III study of chemo-
`therapy plus high dose anti-VEGF (15 mg/kg) versus
`chemotherapy alone in advanced non-squamous cell
`NSCLC [144].
`Another anti-angiogenic drug that has been actively
`investigated is SU5416, a synthetic antagonist of the
`VEGF-receptor type 2, Flk-l/KDR. This agent specifi-
`cally inhibits the phosphorylation of Flk-l that occurs
`in response to binding of its ligand VEGF,
`thereby
`inhibiting in vitro proliferation of endothelial cells and
`growth of in vivo models, including lung tumours [145—
`147]. Side-eflects of this agent consist of nausea, pro-
`jectile vomiting, headache, phlebitis and allergic reac-
`tions, possibly due to the vehicle cremophor® [148—
`150]. Responses after administration as a single agent
`were seen in squamous cell carcinoma of the head and
`neck and in AML [151,152]. In a feasibility study, two
`dose levels of SU5416, 85 and 145 mg/mz, given in
`combination with full doses of gemcitabine and cisplatin
`were investigated [153]. Besides the expected side-effects
`of chemotherapy and SU5416, a worrying increase of
`thromboembolic events was observed. In a total of 19
`
`patients with advanced solid tumours, 9 thromboem-
`bolic events were recorded in 8 patients. The explana-
`tion of these effects remains unclear, but may involve
`disturbances in the coagulation cascade and interaction
`
`Table 2
`Clinical studies with anti—angiogenesis agents in NSCLC
`
`
`
`Drug
`Mechanism
`Status
`Trial design
`
`Rhumab—VEGF
`
`VEGF—antagonist
`
`Thalidomide
`
`TNFa-antagonist
`
`SU5416
`
`Squalamine
`
`TNP-470
`
`Inhibitor of Flk-l receptor
`signalling
`Inhibits sodium-hydrogen
`pump (isoform NH3)
`Fumagillin analogue, broad
`anti—angiogenic activity
`
`Phase III
`
`Phase III
`
`Phase I,
`closed
`Phase II
`
`Phase I
`
`Carboplatin/Paclitaxel + /—Rhumab-VEGF in advanced non-squamous
`carcinoma [144]
`Carboplatin/Paclitaxel followed by radiotherapy + /—thalidomide in
`stage III NSCLC [144]
`Studies closed because of unfavourable toxicity and activity profile
`
`Carboplatin/paclitaxel plus squalamine in advanced NSCLC [239]
`
`Three arm study: TNP—470 continuous infusion in advanced NSCLC
`+ /—Carbop1atin/Paclitaxe1 and + /—bolus TNP-470 [240]
`
`VEGF, Vascular endothelial growth factor. TNch, tumour necrosis factor on.
`
`

`

`2352
`
`LE. Bré‘ker, G. Giaccone / European Journal of Cancer 38 (2002) 2347—2361
`
`with endothelial stability [153—155]. Because of these
`unexpected, severe side-effects, and other somewhat
`disappointing results in other tumours, all clinical stud-
`ies with SU5416 have been closed.
`Thalidomide was first marketed in the 19505 as a non-
`
`barbiturate sedative, but was eventually withdrawn in
`the early 1960s when it was found to be a potent ter-
`atogen. The recent return of thalidomide stems from a
`broad spectrum of pharmacological and immunological
`effects [156]. Although the exact mechanism of action is
`still largely unknown, it is now clear that its activity is
`regulated by a metabolite produced in the liver and
`involves inhibition of TFNot and inhibition of the
`
`angiogenic effects of bFGF and VEGF [157,158]. Tha-
`lidomide is active in several tumour types such as renal
`cell carcinoma, Kaposi’s sarcoma and glioblastoma
`multiforme [159—161]. In refractory multiple myeloma
`patients, response rates as high as 32% were observed
`after treatment with thalidomide as a single agent [162].
`A pilot safety trial of carboplatin, paclitaxel and thali-
`domide in advanced NSCLC patients revealed a good
`tolerance of this agent when given in combination with
`standard chemotherapy [163]. Side-effects associated
`with thalidomide consist of fatigue, myalgia, drowsi-
`ness, constipation and neuropathy [164,165]. A phase
`III trial
`is now being conducted by the ECOG in
`patients with locally advanced, stage IIIA or IIIB,
`NSCLC who receive neoadjuvant chemotherapy fol-
`lowed by radiotherapy with or without thalidomide.
`Many other anti-angiogenic agents are currently in
`several stages of development and some of these will
`certainly be tested in NSCLC as this disease is clearly an
`important target for novel treatment strategies, given its
`frequency and poor results obtained with present treat-
`ments. However, a major problem in the development
`of these agents is that no substantial anti-tumour activ-
`ity can be observed when they are given as a single
`agent. Early initiation of randomised trials to show sta-
`bilisation of disease and impact on survival is therefore
`necessary. Caution is, however, warranted, since toxi-
`cities, though non-overlapping with cytotoxic agents, do
`exist and may be severe, especially in combination
`therapy.
`
`3.4. Matrix metalloproteinase-inhibitors
`
`Matrix metalloproteinases (MMP) are a family of
`more than 20 zinc-endopeptidases, capable of degrading
`and remodelling the extracellular matrix, thereby facil-
`itating not only tumour invasion and metastasis, as ori-
`ginally thought, but also tumour proliferation and
`angiogenesis [166,167]. MMPs are produced as inactive
`zymogens by tumour cells and surrounding stroma cells
`and their activity is regulated by the action of other
`proteases and their endogenous inhibitors [168,169].
`The level of expression of MMPs generally correlates
`
`with the stage of tumour progression [170]. MMP-2
`(gelatinase A) and MMP-9 (gelatinase B) are particu-
`larly expressed in NSCLC and were found to correlate
`with pathological invasiveness and survival [171—175].
`The important role of MMPs in tumour progression
`and metastasis has prompted rapid development of
`therapeutic agents that block enzyme activity in these
`processes [176,177]. Several matrix metalloproteinase-
`inhibitors (MMPIs) are currently being investigated in
`clinical trials to assess their efficacy in maintenance of
`remission after other treatment modalities or in combi-
`
`nation with standard chemotherapy [178]. MMPIs that
`have been studied in NSCLC include batimastat (BB-
`94), marimastat (BB-2516), prinomastat (AG-3340) and
`EMS-275291 [179]. The side-effect profiles of these
`agents are similar and consist of proximal musculoske-
`letal pain, arthralgia and morning stiffness (reviewed in
`[178]).
`Batimastat, a broad spectrum MMPI, was the first
`compound to enter clinical trials but its development
`was hampered by poor bioavailability. Nevertheless, in
`a phase I trial in 18 patients with malignant pleural
`effusion,
`3 of whom had NSCLC, batimastat sig-
`nificantly reduced symptoms and the frequency of ther-
`apeutic
`aspirations
`[180]. Of
`the
`orally
`active
`compounds that are currently in clinical trials, marima-
`stat (BB-2516) is in advanced stage of development.
`Although this agent provided clinical benefit in patients
`with pancreatic and colorectal cancer, activity could not
`be demonstrated in SCLC and breast cancer patients
`[181—184]. A phase III trial with marimastat is ongoing
`in NSCLC patients with minimal residual disease after
`chemotherapy,
`radiotherapy and/or
`surgery. BMS-
`275291 is also being studied in a phase III randomised
`trial, in which the additive effects of this agent vs. pla-
`cebo to a regimen with carboplatin and paclitaxel are
`evaluated in chemotherapy-naive NSCLC patients. Pri-
`nomastat (AG-3340) has recently been investigated in
`two large randomised phase III trials in which its activ-
`ity was studied in combination with paclitaxel/carbo-
`platin and gemcitabine/cisplatin in 686 and 362 NSCLC
`patients, respectively [185,186]. In both trials, patients
`were randomised to receive either prinomastat or a
`placebo, whereas the larger trial also included differ-
`ent dose levels of this agent. Disappointingly, neither
`trial showed an effect of prinomastat on the response
`rate, survival or progression-free survival. However,
`toxicities were more frequent in the prinomastat-con-
`taining arms and were more severe at the higher dose
`levels. Finally, neovastat (d-941), an aqueous extract
`derived from shark cartilage that has anti-VEGF and
`metalloproteinase
`inhibitory activity
`(reviewed
`in
`[187]),
`is undergoing phase II-III testing in NSCLC
`patients as a retrospective efficacy analysis of 48
`NSCLC patients who participated in 2 phase I-II
`trials
`suggested an improved survival
`time was
`
`

`

`‘27
`
`LE. Bré'ker, G. Giaccone/European Journal of Cancer 38 (2002) 2347—2361
`
`2353
`
`induced by this drug [188]. Current trials with MMPIs
`in NSCLC are summarised in Table 3. It appears that
`this class of agents has not produced, at least thus far,
`the promising results that were expected based on pre-
`clinical studies.
`
`3.5. Gene therapy in lung cancer
`
`Cancer is caused by multiple genetic alterations which
`together transform a cell and its progeny into a rapidly
`proliferating, invasive population of cells that has lost
`its ability to undergo programmed cell death. Common
`alterations of oncogenes and tumour suppressor genes
`in NSCLC include mutations in ms, myc, raf, erbB and
`p16, TP53, Rb and FHIT genes,
`respectively [189].
`Restoration of the normal function of specific genes is a
`particularly attracting treatment modality. Depending
`on the strategy, the gene therapy approaches for lung
`cancer can be divided into three groups, replacement of
`defective tumour suppressor genes; introduction of sui-
`cide genes and genetic immunopotentiation.
`A frequently used target for gene therapy in NSCLC
`is TP53 [190], which is mutated in approximately 50%
`of NS