V O L U M E 2 3 d N U M B E R 1 4 d M A Y 1 0 2 0 0 5
`
`JOURNAL OF CLINICAL ONCOLOGY
`
`R E V I E W A R T I C L E
`
`From the Department of Medical
`Oncology, Vrije Universiteit Medical
`Center, Amsterdam, The Netherlands.
`
`Submitted December 4, 2004; accepted
`February 5, 2005.
`
`Author’s disclosures of potential
`conflicts of interest are found at the
`end of this article.
`
`Address reprint requests to G. Giaccone,
`MD, PhD, Division of Medical Oncology,
`Vrije Universiteit Medical Center, 1117
`De Boelelaan, Amsterdam, The Nether-
`lands; e-mail: g.giaccone@vumc.nl.
` 2005 by American Society of Clinical
`Oncology
`
`0732-183X/05/2314-3235/$20.00
`
`DOI: 10.1200/JCO.2005.08.409
`
`Epidermal Growth Factor Receptor Inhibitors in the
`Treatment of Non–Small-Cell Lung Cancer
`Giuseppe Giaccone
`
`A
`
`B
`
`S
`
`T
`
`R
`
`A
`
`C
`
`T
`
`Remarkable developments in the systemic treatment of advanced non–small-cell lung cancer
`have taken place in the past few years. Targeted therapies have been largely employed in
`patients with far advanced disease, and some of them have demonstrated consistent activity
`in this setting. Epidermal growth factor receptor (EGFR) inhibitors cause dramatic response in
`approximately 10% of white patients who had received prior chemotherapy. Responses are
`higher in Asians. These findings are at least partly caused by the substantially higher incidence
`of EGFR mutations in Asians compared with whites. Studies of EGFR inhibitors in com-
`bination chemotherapy in front-line therapy of advanced non–small-cell lung cancer have,
`however, failed to improve survival, and better understanding of interactions between che-
`motherapeutic agents and EGFR inhibitors will be essential in the development of more effec-
`tive strategies.
`J Clin Oncol 23:3235-3242.  2005 by American Society of Clinical Oncology
`
`EPIDERMAL GROWTH FACTOR RECEPTOR IN
`NON–SMALL-CELL LUNG CANCER
`
`Lung cancer develops after a number of
`genetic hits, caused most commonly by
`years of smoking. A large number of chro-
`mosomal alterations and genetic defects
`have been identified in lung cancer, but a
`clear understanding of the chronology of
`these events in the carcinogenetic process
`is still rather unclear. Some molecular al-
`terations are visible before morphologic
`changes are apparent
`in the bronchial
`mucosa by light microscopy.1
`The epidermal growth factor receptor
`(EGFR) is a member of the ErbB family
`of transmembrane tyrosine kinase recep-
`tors, which includes ErbB1 (or HER-1, or
`EGFR), ErbB2 (or HER-2/neu), ErbB3 (or
`HER-3), and ErbB4 (or HER-4). The ex-
`pression of EGFR is common in a number
`of normal epithelial tissues and expression
`of EGFR is elevated in several solid tumors.
`In non–small-cell
`lung cancer (NSCLC),
`overexpression of EGFR has been reported
`to be present in over 50% of cases in several
`
`series. In addition to this, several retrospec-
`tive studies have identified the expression
`of EGFR as a negative prognostic factor
`in patients with resected early NSCLC.2
`
`CLINICAL STUDIES WITH EGFR
`INHIBITORS IN NSCLC
`
`Single Agents
`Several EGFR inhibitors have been de-
`veloped in recent years,2 which can be
`mainly categorized into two classes: mono-
`clonal antibodies to the extracellular do-
`main of the EGFR, or small molecules that
`are inhibitors of the intracellular tyrosine
`kinase (TKI) domain by interfering with
`autophosphorylation by adenosine triphos-
`phate (ATP). Other molecules, however,
`have also been developed, including vaccines
`and other approaches.3 In Table 1, the mol-
`ecules that have been studied or are planned
`in NSCLC are listed.
`
`Gefitinib
`Gefitinib (Iressa; AstraZeneca, Wil-
`mington, DE)
`is
`the first molecularly
`
`3235
`
`Downloaded from ascopubs.org by GEORGETOWN UNI MED CNTR DAHLGREN Library on July 18, 2017 from 141.161.013.208
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`APOTEX EX. 1051-001
`
`

`

`Giuseppe Giaccone
`
`Agent
`
`Class
`
`Target
`
`Characteristics
`
`Route
`
`Latest Studies
`
`Table 1. Development of EGFR Inhibitors in NSCLC
`
`Gefitinib
`
`Erlotinib
`
`Cetuximab
`
`Matuzumab
`(EMD72000)
`
`Small
`molecule
`
`Small
`molecule
`Monoclonal
`antibody
`Monoclonal
`antibody
`
`EGFR TK
`
`EGFR TK
`
`EGFR extracellular
`domain
`EGFR extracellular
`domain
`
`Panitumumab
`(ABX-EGF)
`MDX214
`
`EGF vaccine
`
`Monoclonal
`antibody
`Monoclonal
`antibody
`Vaccine
`
`EGFR extracellular
`domain
`EGFR extracellular
`domain
`EGF
`
`Pertuzumab
`
`Monoclonal
`antibody
`TheraCIM hR3 Monoclonal
`antibody
`Small
`molecule
`Small
`molecule
`
`Lapatinib
`(GW572016)
`Canertinib
`(CI-1033)
`
`Erb2
`heterodimerization
`EGFR extracellular
`domain
`EGFR and ErbB2
`TKs
`EGFR, ErbB2, and
`ErbB4 TKs
`
`ZD6474
`
`AEE788
`
`Small
`molecule
`Small
`molecule
`
`EGFR and VEGFR-2
`TKs
`EGFR, ErbB2, and
`VEGFR-2 TKs
`
`Specific reversible
`inhibitor
`
`Specific reversible
`inhibitor
`Chimeric
`
`Humanized
`
`Fully human
`
`Fully human
`
`hrEGF bound to
`protein and alum
`
`Humanized
`
`Oral
`
`Oral
`
`IV
`
`IV
`
`IV
`
`IV
`
`ID
`
`IV
`
`IV
`
`Dual reversible
`inhibitor
`Irreversible,
`pan-inhibitor of all
`ErbB members
`Inhibitor of EGFR
`and VGFR-2
`Inhibitor of EGFR,
`HER-2, VEGFR-2
`
`Oral
`
`Oral
`
`Oral
`
`Oral
`
`Phase III in advanced NSCLC
`completed; phase III in
`adjuvant setting ongoing
`Phase III in advanced NSCLC
`completed
`Phase III in advanced NSCLC
`started in December 2004
`Phase I-II in advanced NSCLC
`in combination with
`paclitaxel
`Phase II in advanced NSCLC
`ongoing
`Phase II in advanced NSCLC
`ongoing
`Phase II completed, needs
`confirmatory studies
`
`Phase II ongoing in advanced
`NSCLC
`Phase II planned in advanced
`NSCLC for 2005
`Randomized phase II in
`advanced NSCLC
`Randomized phase II in
`advanced NSCLC
`
`Randomized phase II trials
`in advanced NSCLC
`Phase II in advanced NSCLC
`
`Highest Phase in
`NSCLC
`
`Registered in United
`States, Japan, and
`other countries
`Registered in
`United States
`
`III
`
`II
`
`II
`
`II
`
`II
`
`II
`
`II
`
`II
`
`Study completed
`mid 2004
`
`Drug Company
`
`AstraZeneca
`
`OSI/Genentech/Roche
`
`Imclone/Bristol-Myers
`Squibb/Merck
`EMD/Merck KgGA
`
`Abgenix/Amgen
`
`Medarex
`
`CIMAB/Biocon/
`CancerVax/
`YMBioSciences
`Genentech/Roche
`
`YM BioSciences
`
`GlaxoSmithKline
`Beecham
`Pfizer
`
`Completed accrual
`
`AstraZeneca
`
`Planned in 2005
`
`Novartis
`
`Abbreviations: EGFR, epidermal growth factor receptor; NSCLC, non–small-cell lung cancer; TK, tyrosine kinase; IV, intravenous; EGF, epidermal growth
`factor; hrEGF, human recombinant epidermal growth factor; ID, intradermally; VEGFR-2, vascular endothelial growth factor receptor-2.
`
`targeted agent to be registered for advanced NSCLC.
`Gefitinib has recently received approval
`in Japan, the
`United States, and several other countries in the world
`for the treatment of patients who experienced treatment
`failure with chemotherapy. This registration has been
`based on two large randomized phase II studies, the Iressa
`Dose Evaluation in Advanced Lung Cancer (IDEAL) -1
`and -2 studies,4 where gefitinib was given orally in contin-
`uous dosing. In IDEAL-2, patients had to have received
`a platinum compound and docetaxel,5 and the study
`was performed in 30 centers in the United States. In total,
`221 patients were randomly assigned to receive either 250-
`or 500-mg gefitinib daily. The selection of the two doses
`was based on the doses at which activity was observed
`in phase I studies on the one hand, and tolerability of
`the higher dose when the drug is given continuously, on
`the other. Patients had to be symptomatic, and symptom
`improvement was assessed using the FACT-L instrument
`(Functional Assessment of Cancer Therapy-Lung). A total
`of 126 patients (58%) had three or more regimens in the
`past, and 65% had histology of adenocarcinoma. Symp-
`toms of NSCLC improved in 43% of patients receiving
`250-mg gefitinib and in 35% of those receiving 500 mg.
`There was no significant difference in response rate or sur-
`vival between the two dosages (approximately 10% in
`both). There was a good correlation between clinical
`
`response and symptomatic improvement. Interestingly,
`the symptomatic improvement rate approached 100% in
`patients with major responses and was in excess of 60%
`in patients with stable disease, suggesting that benefit
`can also be obtained in absence of a major response. How-
`ever, the 500-mg gefitinib dose was more toxic as it in-
`duced more acne-like rash (P Z .04) and diarrhea (P Z
`.006). Diarrhea was noted in 57% of patients receiving 250
`mg and 75% in those receiving 500 mg. Skin toxicity (rash,
`acne, dry skin, pruritus) was observed in 62% and 75%,
`respectively. Grade 3/4 toxicities were unusual, but
`more frequent in the 500-mg dosing (P Z .02). Dose re-
`ductions as a result of toxicity were also more frequent in
`the higher dose (P Z .009). In conclusion, gefitinib was
`well tolerated at 250 mg/d and it induced antitumor activ-
`ity in approximately 10% of patients. These results are
`impressive compared with chemotherapy, which induces
`far more adverse effects and probably even a lower level
`of activity.6
`In the second study (IDEAL-1), conducted mainly in
`Europe and Japan, patients with one or two prior chemo-
`therapy regimens, including a platinum compound, were
`also randomly assigned to receive gefitinib at 250 or 500
`mg/d.7 Response rate approached 20% and was similar
`in both arms, and symptom improvement was 40%, which
`was higher in patients who had an objective response.
`
`3236
`
`JOURNAL OF CLINICAL ONCOLOGY
`
`Downloaded from ascopubs.org by GEORGETOWN UNI MED CNTR DAHLGREN Library on July 18, 2017 from 141.161.013.208
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`APOTEX EX. 1051-002
`
`

`

`EGFR Inhibitors in the Treatment of NSCLC
`
`Again, adverse effects were, in general, well tolerated, but
`were more severe with the 500-mg dose. Similar toxicities
`were recorded in this study, as those reported by Kris et al.5
`Among the stratification factors for this study, ethnicity
`also played a role (Japanese v non-Japanese). This allowed
`more detailed assessment of the influence of some patient
`characteristics on efficacy. Japanese patients responded in
`27.5% versus 10.4% for non-Japanese (P Z .0023). How-
`ever, this factor lost significance in a multivariate analysis,
`where it became apparent that female patients, patients
`with histology of adenocarcinoma, and those who had pre-
`viously received hormonal or immunologic therapies had
`significantly higher chances of response.
`Gefitinib has been available in a large extended access
`program to many countries, and as a result, several reports
`have been published based on this experience. A few
`reports gave information on the results obtained in ad-
`vanced NSCLC patients who relapsed after chemo-
`therapy, but a number of untreated patients, who were
`either unfit or unwilling to receive chemotherapy, were
`also included in these reports. This was a rather broad pro-
`gram; patients were given 250 mg/d, and inclusion and
`assessments were relatively loose. In general, the response
`rate was similar or somewhat lower to the IDEAL studies
`in these reports.8-10 From these reports it became apparent
`that gefitinib is active in first-line therapy, with a level of
`activity not clearly higher than that observed after chemo-
`therapy failure,10-12 and that patients with brain meta-
`stases13-15 and elderly patients16,17 can benefit from this
`treatment. Patients with poor performance status may
`also be candidates for gefitinib treatment, and results
`were particularly promising in Chinese patients.17,18
`In a review of 139 patients treated with gefitinib
`250 mg/d, 21 had a partial response, for a response rate
`of 15%.19 By multivariate analysis, the presence of bron-
`chioloalveolar features and being a never-smoker were
`independent predictors of response. Gefitinib has been
`further investigated prospectively by the Southwest Oncol-
`ogy Group in a large phase II study of 138 patients with
`bronchioloalveolar histology.20 Gefitinib was given at
`500 mg/d; of 67 chemotherapy-naı¨ve patients with mea-
`surable disease, the response rate was 21%, whereas the re-
`sponse rate in 21 pretreated patients was 10%. Median
`survival was 12 and 10 months in chemotherapy-naive
`and pretreated patients,
`respectively. Subset analysis
`showed a prolonged survival in female patients (16 v 7
`months) and in those who developed a rash from the treat-
`ment (12 v 5 months).
`Principal side effects of gefitinib are skin rash, acnei-
`form changes of the skin, and diarrhea. Diarrhea was ac-
`tually the dose-limiting toxicity in phase I studies. Most
`toxicities are Common Toxicity Criteria grade 1 or 2. In-
`terstitial lung disease has been observed in patients receiv-
`ing gefitinib. Worldwide, the incidence of interstitial lung
`
`disease is about 1% (2% in the Japanese postmarketing ex-
`perience and about 0.3% in a US expanded access pro-
`gram), with approximately one third of the cases being
`fatal.4 A recent article by Shah et al21 gives detailed infor-
`mation for the practicing oncologist on how to deal with
`gefitinib toxicities, and what to expect in terms of activity
`and patient selection.
`Several studies are still ongoing with gefitinib, includ-
`ing a large adjuvant study (BR.19) that randomly assigns
`radically resected NSCLC patients to receive gefitinib or
`placebo as adjuvant therapy. A study (Iressa Survival
`Evaluation in Lung Cancer) comparing gefitinib 250 mg
`daily versus placebo in patients experiencing treatment
`failure with chemotherapy recently concluded accrual.
`This study, which randomly assigned 1,692 patients, failed
`to demonstrate an improvement of survival with gefitinib
`(AstraZeneca press release, London, United Kingdom, De-
`cember 17, 2004) in the overall population (median 5.6 v
`5.1 months; P Z .11) or in patients with adenocarcinoma
`(6.3 v 5.4 months; P Z .07). There was suggested benefit in
`Asians and never-smokers. These results are disappointing
`and surprising in view of the positive results obtained with
`erlotinib (BR.21) recently reported. Although gefitinib and
`erlotinib have some differences in pharmacologic proper-
`ties, there are many similarities between the two com-
`pounds, and it seems plausible that the choice of dose
`may have played a role in determining the different out-
`come of these studies. The dose of erlotinib used was
`150 mg/d, which corresponds to about 600 to 700 mg/
`d of gefitinib. The fact that the IDEAL studies did not dis-
`cern differences in outcome between the 250- and the 500-
`mg/d gefitinib may have to do with the small sample size of
`these randomized phase II studies, which does preclude
`conclusions on survival. In light of these results, the on-
`going or planned studies with gefitinib are under recon-
`sideration. The
`adjuvant
`study
`(BR.19)
`and the
`maintenance studies will probably have to be amended,
`and a higher dose of gefitinib may have to be investigated.
`In addition, selection of patients, based on molecular
`markers and patient characteristics, should probably be
`considered for the continuation of these studies. For the
`practicing oncologist dealing with second- and third-
`line treatment of advanced NSCLC, these results suggest
`that erlotinib is a better drug than gefitinib at the doses
`used in these studies because it increases survival. How-
`ever, in this setting, where palliation and adverse effects
`are very important, one may wish to use low and relatively
`nontoxic doses of these compounds.
`
`Erlotinib
`Erlotinib (Tarceva; OSI Pharmaceuticals, Melville,
`NY) is another TKI, with slightly different pharmacologic
`characteristics from gefitinib. A phase II study in ad-
`vanced NSCLC was performed in 57 patients with prior
`
`www.jco.org
`
`3237
`
`Downloaded from ascopubs.org by GEORGETOWN UNI MED CNTR DAHLGREN Library on July 18, 2017 from 141.161.013.208
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`APOTEX EX. 1051-003
`
`

`

`Giuseppe Giaccone
`
`chemotherapy22; a response rate of 12.3% was obtained.
`Response did not appear to be correlated to the extent
`of prior exposure to chemotherapy. Survival was corre-
`lated to occurrence and degree of skin toxicity, and there
`was no relation to EGFR expression.
`Recently, the results of the randomized study BR.21
`were presented, where patients who had experienced treat-
`ment failure with one or two prior chemotherapy regi-
`mens were randomly assigned to receive placebo or
`erlotinib 150 mg daily.23 The study randomized 731 pa-
`tients in a 2:1 randomization ratio to receive erlotinib
`or placebo. Approximately half of the patients were ran-
`domly assigned to second-line treatment and the other
`half to third-line. In this study, an improvement in sur-
`vival was obtained in the erlotinib arm: patients on pla-
`cebo had a median survival of 4.7 months whereas those
`on erlotinib had a median survival of 6.7 months (P Z
`.001). The response rate to erlotinib was 8.9%. Toxicity
`was as expected, and acceptable, consisting mainly of
`skin toxicity and diarrhea. These adverse effects were in
`the range observed with the higher doses of gefitinib in
`the IDEAL studies.5,7 Major symptoms (cough, dyspnea,
`and pain) were significantly improved by erlotinib treat-
`ment compared with placebo. This important study
`brought erlotinib to registration by the US Food and
`Drug Administration on November 19, 2004 for the treat-
`ment of second- and third-line advanced NSCLC.
`
`Cetuximab
`So far, not much experience has been obtained with
`cetuximab in advanced NSCLC. To date, only one single-
`agent study has been performed in advanced NSCLC,24
`but results are still preliminary. In this study of 29 patients
`who experienced treatment failure after one or more regi-
`mens, there were two partial responses. Cetuximab has
`recently been registered in the treatment of irinotecan-
`refractory colorectal cancer patients.25 The main adverse
`effects of cetuximab are skin rash and acneiform skin tox-
`icity, and up to 4% are hypersensitivity reactions.26 Inter-
`estingly, diarrhea is not a common adverse effect of
`monoclonal therapy, in contrast to TKI treatment.
`
`Other EGFR Inhibitors
`An interesting study from Cuba reported results of
`vaccination with an epidermal growth factor–based cancer
`vaccine in 40 patients with advanced NSCLC.27 Although
`this was an unselected patient population, those patients
`who developed a good antibody response had a longer sur-
`vival. No substantial toxicity was reported with this vac-
`cine. These results, however, need confirmation.
`A phase II study is presently accruing patients with
`advanced NSCLC to receive pertuzumab as a single agent.
`Pertuzumab is an ErbB2 monoclonal antibody with a dif-
`ferent epitope than trastuzumab, which, as its main action,
`
`prevents homo- and heterodimerization of the receptor.28
`In doing so, it also blocks the EGFR pathway.
`A number of TKIs have been developed, which have
`a broader inhibitory spectrum on other tyrosine kinases
`of the Erb family or other receptor families (Table 1). CI-
`1033, a pan-HER inhibitor (inhibitor of all ErbB members)
`has recently completed a randomized phase II study of three
`different schedules of administration. The results are pres-
`ently being analyzed. Toxicities of this agent appear similar
`to EGFR inhibitors (skin rash and diarrhea), although some
`other adverse effects have been described (thrombocytope-
`nia and allergy).29 A randomized phase II study of lapatinib
`has recently been completed. Lapatinib inhibits both EGFR
`and ErbB2 with similar potency. The study randomly as-
`signed patients with untreated NSCLC to receive two differ-
`ent doses of
`lapatinib. However, the study was later
`amended to only include patients who were never-smokers
`or had adenocarcinomas with bronchioloalveolar carci-
`noma (BAC) features or pure BAC, and allowed one line
`of prior chemotherapy.
`Given the important issue of patient selection, based on
`molecular markers for EGFR action (eg, EGFR mutations),
`combination of inhibitory activities on other tyrosine
`kinases may increase the number of patients who may ben-
`efit from the treatment. Inhibitors of EGFR and vascular en-
`dothelial growth factor receptor-230 are being developed in
`advanced NSCLC. ZD6474 is undergoing large randomized
`phase II studies in advanced NSCLC, some of which have
`recently completed accrual. Phase II studies in advanced
`NSCLC with AEE78831 are planned to start in 2005.
`
`Combinations With Chemotherapy
`In a dose finding study, gefitinib was tested in first-line
`therapy in combination with carboplatin and paclitaxel.32
`In the first part of the study, patients were randomly as-
`signed to receive intermittent gefitinib with cycle 1 or 2
`of chemotherapy. In the second part of the study, the high-
`est dose of gefitinib, which was given without dose-limiting
`toxicity from the first part, was administered continuously
`beginning with the first cycle of chemotherapy. Three se-
`quentially enrolled cohorts received gefitinib 250 and
`500 mg (intermittently) and 500 mg (continuously). No
`unexpected toxicities were observed with either gefitinib
`doses and no pharmacokinetic interaction of gefitinib
`was observed on chemotherapy; however, a slight interac-
`tion was observed between chemotherapy on area under
`the curve and the maximum concentration of gefitinib.
`The response rate and survival were in the range of what
`is commonly observed with chemotherapy alone in this set-
`ting. This study was the basis for a large randomized study
`of gefitinib and chemotherapy (The Iressa NSCLC Trial As-
`sessing Combination Treatment [INTACT-2]).
`In another dose-finding study, the combination of cis-
`platin and gemcitabine has been studied in combination
`
`3238
`
`JOURNAL OF CLINICAL ONCOLOGY
`
`Downloaded from ascopubs.org by GEORGETOWN UNI MED CNTR DAHLGREN Library on July 18, 2017 from 141.161.013.208
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`APOTEX EX. 1051-004
`
`

`

`EGFR Inhibitors in the Treatment of NSCLC
`
`with gefitinib at 250 or 500 mg/d.33 This study confirmed
`that both doses could be combined safely with full-dose
`chemotherapy, and there were also no major pharmaco-
`logic interactions between the three drugs. This study
`constituted the rationale to run another large phase III
`study (INTACT-1).
`The INTACT-1 and -2 studies were large randomized
`studies of two doses of gefitinib (250 or 500 mg/d) or
`placebo, in combination with two different chemotherapy
`regimens. INTACT-1 used cisplatin and gemcitabine (cis-
`platin 80 mg/m2 on day 1 and gemcitabine 1,250 mg/m2
`on days 1 and 8 every 3 weeks),34 whereas INTACT-2
`used carboplatin and paclitaxel (carboplatin given at area
`under the curve of 6 and paclitaxel at 225 mg/m2 in
`3-hour infusion every 3 weeks).35 Chemotherapy was given
`for up to six cycles and gefitinib or placebo were continued
`in nonprogressing patients until progression. The main re-
`sults of these two studies are summarized in Table 2. A total
`of 1,093 and 1,037 patients were entered respectively in the
`two studies in less than 1 year of accrual. These two large ran-
`domized studies failed to demonstrate a survival increase
`with the addition of gefitinib to standard chemotherapy
`in first-line treatment of advanced NSCLC. A subset analysis
`of patients with adenocarcinoma who received 90 days of
`chemotherapy or more in the INTACT-2 study demon-
`strated statistically significant prolonged survival, suggest-
`ing a gefitinib maintenance effect. In general, treatment
`was well tolerated and the toxicity of chemotherapy did
`not overlap with gefitinib treatment (skin rash and diar-
`rhea), which made the studies feasible. However, as ex-
`pected, gefitinib 500 mg was associated with a higher
`degree of toxicity, as observed in the IDEAL studies, which
`led to more dose reductions and treatment interruptions.
`Two other studies employed the same combination
`chemotherapy regimens with erlotinib at 150 mg versus
`placebo. Also, these two large studies failed to demonstrate
`superiority of the erlotinib combination in terms of sur-
`vival benefit. The TALENT (Tarceva Lung Cancer Investi-
`gation) study enrolled 1,172 patients to be randomly
`
`assigned between cisplatin-gemcitabine given at the same
`doses as in the INTACT-1 study plus either placebo or er-
`lotinib 150 mg/d.36 The TRIBUTE (Tarceva Responses in
`Conjunction with Paclitaxel and Carboplatin)
`study
`randomly assigned 1,079 patients to receive carboplatin-
`paclitaxel (at the same doses as in the INTACT-2 study)
`plus either placebo or erlotinib at 150 mg/d.37 In these stud-
`ies, the oral medication was continued after termination
`of chemotherapy. Overall, these combinations were well
`tolerated and there was a positive correlation of survival
`with the degree of skin rash in the TRIBUTE study.37
`In none of these studies were patients selected based on
`EGFR expression, or any other marker of efficacy, and this
`lack of patient selection may have caused the lack of positive
`outcome. Interestingly, however, the time to progression
`curves and survival curves suggest
`that maintenance
`EGFR inhibition may be of help after termination of chemo-
`therapy. In some of these studies, landmark analysis has, in
`fact, supported this observation.35,37 These considerations
`would suggest that sequential therapies be the best approach
`to this disease for front-line therapy. Interestingly, in the
`TRIBUTE study, patients who were never-smokers had
`a longer survival if they received erlotinib,38 with a median
`survival of 23 versus 10 months (hazard ratio Z 0.49) for
`erlotinib plus chemotherapy versus chemotherapy alone,
`respectively, in the 69 and 44 never-smokers present, re-
`spectively, in these arms of the study.
`Based on these considerations, sequential studies of
`EGFR TKIs (both gefitinib and erlotinib) following chemo-
`therapy have been planned or are presently accruing; these
`studies will investigate whether a TKI may increase outcome
`in patients who were not progressing on chemotherapy.
`It is also, however, interesting to note that in the ini-
`tial part of the survival curves, the EGFR TKI arms in all of
`these studies did worse than the placebo arms. This is un-
`likely as a result of toxicity, as even the higher doses of
`these drugs are relatively well tolerated in combination
`with chemotherapy. EGFR inhibitors act mainly by re-
`ducing proliferation in wild-type EGFR tumor cells, and
`
`Table 2. Four Randomized Studies of EGFR Tyrosine Kinase Inhibitors in Combination With Chemotherapy in First-Line Therapy of Advanced NSCLC
`No. of
`Response
`Median Time to
`Median
`1-Year
`Patients
`Rate (%)
`Progression (months)
`Survival (months)
`Survival
`
`Regimen
`
`Study
`INTACT-134
`
`INTACT-235
`
`TALENT36
`
`TRIBUTE37*
`
`Cisplatin-gemcitabine-placebo
`Cisplatin-gemcitabine-gefitinib 250 mg
`Cisplatin-gemcitabine-gefitinib 500 mg
`Carboplatin-paclitaxel-placebo
`Carboplatin-paclitaxel-gefitinib 250 mg
`Carboplatin-paclitaxel-gefitinib 500 mg
`Cisplatin-gemcitabine-placebo
`Cisplatin-gemcitabine-erlotinib
`Carboplatin-paclitaxel-placebo
`Carboplatin-paclitaxel-erlotinib
`
`363
`365
`365
`345
`345
`347
`586
`586
`540
`539
`
`47.2
`51.2
`50.3
`28.7
`30.4
`30.0
`29.9
`31.5
`19.3
`21.5
`
`6.0
`5.8
`5.5
`5.0
`5.3
`4.6
`5.6
`5.4
`4.9
`5.1
`
`10.9
`9.9
`9.9
`9.9
`9.8
`8.7
`10.1
`9.9
`10.5
`10.6
`
`44
`41
`43
`42
`41
`37
`42
`41
`43.8
`46.9
`
`Abbreviations: EGFR, epidermal growth factor receptor; NSCLC, non–small-cell lung cancer; INTACT, Iressa NSCLC Trial Assessing Combination Treatment;
`TALENT, Tarceva Lung Cancer Investigation; TRIBUTE, Tarceva Responses in Conjunction with Paclitaxel and Carboplatin.
`*In the TRIBUTE study, never-smokers did better.38
`
`www.jco.org
`
`3239
`
`Downloaded from ascopubs.org by GEORGETOWN UNI MED CNTR DAHLGREN Library on July 18, 2017 from 141.161.013.208
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`APOTEX EX. 1051-005
`
`

`

`Giuseppe Giaccone
`
`proliferating tumor cells are those most affected by che-
`motherapy,
`therefore an antagonistic effect between
`EGFR TKIs and chemotherapy is plausible. The influence
`of Ras mutations on the worse prognosis of patients
`treated with EGFR TKIs and chemotherapy should also
`be taken into consideration.
`Cetuximab given at 400 mg/m2 on day 1 and at 250
`mg/m2 weekly thereafter has been combined with cisplatin
`80 mg/m2 on day 1 and vinorelbine 25 mg/m2 on day 1
`and 8 in a randomized phase II study.39 A total of 101 pa-
`tients whose tumors expressed EGFR were included. At
`a preliminary assessment, the response rate was 31.7%
`in the cetuximab arm and 20% in the chemotherapy-alone
`arm. Based on these results, a large phase III randomized
`study has recently been launched to compare cisplatin/
`vinorelbine to the same chemotherapy in combination
`with cetuximab. Furthermore, there are two ongoing ran-
`domized phase II studies that use the chemotherapy regi-
`mens carboplatin/paclitaxel and cisplatin/gemcitabine with
`or without cetuximab. Each study will accrue 300 patients.
`In a phase I/II trial of advanced NSCLC, patients
`treated with one prior regimen were given erlotinib and
`bevacizumab, a recombinant anti–vascular endothelial
`growth factor monoclonal antibody.40 Both agents could
`be given at the full dose of 150 mg/d of erlotinib and 15
`mg/kg bevacizumab every 3 weeks. Of 40 patients enrolled,
`the response rate was 17.5% and median survival was
`9.3 months. These results appear promising and have
`stimulated further studies of this combination in first-
`and second-line treatment of advanced NSCLC.
`
`MARKERS OF SENSITIVITY TO EGFR INHIBITORS
`IN ADVANCED NSCLC
`A study by Cappuzzo et al41 attempted to identify the value
`of immunohistochemistry for HER-2 and EGFR as predic-
`tors of response to gefitinib or symptom control. Neither
`marker was predictive; however, only 43 patients were
`assessed and partial remission was only observed in six
`patients (14%). Larger studies are necessary.
`Staining intensity was also not correlated with re-
`sponse in a phase II study of erlotinib.22 In this study,
`however, rash occurrence and intensity were correlated
`with survival. Although the studies performed with mono-
`clonal antibodies have selected patients with positive
`EGFR expression by immunohistochemistry, the level of
`expression was not correlated with outcome in a study
`in colorectal cancer.25
`Recently, EGFR mutations in the ATP binding site do-
`main have been described as conferring sensitivity to both
`gefitinib and erlotinib.42-44 A more detailed discussion of
`EGFR mutations and their biology is the topic of another
`section in this special issue. So far, no mutations have been
`described in nonresponders to TKIs. However, not all re-
`sponding patients harbor a mutation in their tumors,
`
`which suggests that other markers of sensitivity may be
`needed to select patients for these treatments in the future.
`Interestingly, EGFR mutations have been found more
`commonly in women, never-smokers, patients with ad-
`enocarcinomas, and Japanese patients.42 Within the
`adenocarcinoma type, those with bronchioloalveolar fea-
`tures appear to be particularly sensitive to these treat-
`ments. This may be due to the fact that women and
`never-smokers more often develop this type of lung can-
`cer, which has a somewhat more benign behavior than
`common NSCLC. The much higher incidence of EGFR
`mutations in Asians and nonsmokers is puzzling, and it
`may possibly be due to different types of carcinogenetic
`exposure in Asia versus the United States and Europe.
`More women are never-smokers and passive smokers in
`Asia than in the Western world. EGFR mutations may
`be washed out by the occurrence of a later genetic event
`(eg, Ras mutations) in active smokers.
`It is also interesting to note that EGFR TKIs have
`a higher toxicity in Asians than in whites. This is unlikely
`due to the presence of tumor EGFR mutations, which are
`somatic, not germline mutations. Recently, a significantly
`higher proportion of Asians were shown to have longer CA
`repeats polymorphism of EGFR intron 1.45 This difference
`might explain the higher toxicities, and may also be pos-
`sibly related to activity.46 Prospective studies will be nec-
`essary to confirm these findings.
`In looking at the BR.21 results, one may speculate that
`mutations may have been responsible for the survival im-
`provement obtained in erlotinib. However, major re-
`sponses were only obtained in less than 10% of patients
`and it is unlikely that this limited response rate is respon-
`sible for such an important shift in survival. However,
`a large proportion of patients had either minor response
`or stable disease, and considering that one of the major
`effects of EGFR inhibitors is, in fact, to reduce prolifera-
`tion, this may be a major effect indeed. Recent studies
`point out that phosphorylated Akt expression may be
`an indicator of tendency to respond to gefitinib,47 and
`for this it may be essential to have a high expression of
`EGFR. These data, like the EGFR mutation data, need val-
`idation in prospectively conducted clinical trials of pa-
`tients selected on the basis of biologic characteristics.
`For the time being, for clinical practice it may be prema-
`ture to exclude treatment of patients with an EGFR TKI
`based on EGFR mutations only, as their absence