Vol. 9, 5813–5824, December 1, 2003
`
`Clinical Cancer Research
`
`5813
`
`Review
`Targeting the Epidermal Growth Factor Receptor in Non-Small Cell
`Lung Cancer
`
`Roy S. Herbst1 and Paul A. Bunn, Jr.2
`1Department of Thoracic/Head and Neck Medical Oncology, The
`University of Texas, M. D. Anderson Cancer Center, Houston, Texas,
`and 2Tobacco Related Malignancies Program, University of Colorado
`Cancer Center and Division of Medical Oncology, University of
`Colorado Health Sciences Center, Denver, Colorado
`
`Abstract
`Fifteen % or fewer of patients with non-small cell lung
`cancer (NSCLC) survive 5 years. The current standard of care
`for patients with locally advanced or metastatic NSCLC is
`systemic chemotherapy with a two-drug combination regimen
`that includes a platinum agent. Although systemic chemother-
`apy reduces the rate of death attributable to lung cancer,
`disease progression is inevitable and dose-limiting toxicities
`restrict their use. New molecularly targeted therapies aim to
`inhibit specific pathways and key molecules implicated in tu-
`mor growth and progression while sparing normal cells. Sev-
`eral therapies, which target signal transduction pathways in-
`volved in angiogenesis, metastasis, and apoptosis, are in clinical
`development to treat lung cancer. Among these targeted ther-
`apies are the oral, small-molecule epidermal growth factor
`receptor-tyrosine kinase (EGFR-TK) inhibitors gefitinib and
`erlotinib. Both therapies have been validated preclinically as
`new treatment approaches for NSCLC and have shown single-
`agent activity against advanced, chemorefractory NSCLC in
`clinical trials. This article focuses on the biology of the
`EGFR-TK signal transduction pathway, its role in the prolif-
`eration of solid tumors, and the rationale for the clinical de-
`velopment of EGFR-TK inhibitors. We also review clinical
`trials with EGFR-TK inhibitors in NSCLC and future direc-
`tions of investigation with these targeted agents.
`
`Introduction
`The TK3 activity of the EGFR has received considerable
`attention as a target for cancer therapy (1, 2). In recent clinical
`
`Received 5/28/03; revised 8/5/03; accepted 8/6/03.
`Grant support: Supported in part by an M.D. Anderson Cancer Center
`Physician Scientist Program Award and an American Society of Clinical
`Oncology Career Development Award (R. S. H.).
`The costs of publication of this article were defrayed in part by the
`payment of page charges. This article must therefore be hereby marked
`advertisement in accordance with 18 U.S.C. Section 1734 solely to
`indicate this fact.
`Requests for reprints: Roy S. Herbst, Department of Thoracic/Head
`and Neck Medical Oncology, M. D. Anderson Cancer Center, 1515
`Holcombe Boulevard, Unit 432, Houston, TX 77030. Phone: (713) 792-
`6363; Fax: (713) 796-8655; E-mail: rherbst@mdanderson.org.
`3 The abbreviations used are: TK, tyrosine kinase; EGFR, epidermal
`growth factor receptor; NSCLC, non-small cell lung cancer; SWOG,
`Southwest Oncology Group; ECOG, Eastern Cooperative Oncology
`Group; HER, human EGF-like receptor; TGF, transforming growth
`factor; PI3K, phosphatidylinositol 3⬘-kinase; MAPK, mitogen-activated
`
`trials, selective and orally active EGFR-TK inhibitors gefitinib
`[IRESSA (ZD1839); AstraZeneca] and erlotinib [Tarceva (OSI-
`774); OSI and Genentech] produced objective tumor responses
`and symptom improvement in some patients with NSCLC who
`had previously received chemotherapy (3–5). This was the first
`class of oral targeted therapies to produce such responses in
`advanced NSCLC. Although chemotherapy can result in life-
`threatening toxicities, the EGFR-TK inhibitors have far better
`safety profiles in patients with advanced NSCLC.
`Lung cancer is the leading cause of cancer death in both
`men and women in the United States and throughout the world
`(6, 7). The 5-year survival rate for lung cancer patients remains
`very poor with 15% or less surviving 5 years (6). Nonetheless,
`this is improved compared with the 5% 5-year survival rate in
`the United States in the 1960s and the 5% rate still seen in many
`parts of the world. The major reasons for the poor survival rate
`for lung cancer are the lack of effective screening and early
`diagnosis procedures, the propensity for early metastasis, and
`the inability of systemic therapies to cure patients with widely
`metastatic disease. This is not to conclude that there have been
`no advances in lung cancer therapy. Systemic chemotherapy
`produces a 26–32% reduction in the hazard rate of death for
`patients with advanced stage III/IV NSCLC that includes ade-
`nocarcinomas, squamous carcinomas, and large cell carcinomas
`(8–10). Chemotherapy also reduces lung cancer-related symp-
`toms and improves quality of life in patients with advanced
`NSCLC (8–11).
`The current standard of care for patients with locally ad-
`vanced or metastatic NSCLC is systemic chemotherapy with a
`two-drug combination regimen that includes a platinum agent
`(8). Such two-drug combinations, developed in the 1990s, were
`shown to be more effective than the best supportive care or
`treatment with a single chemotherapy agent. These two-drug
`combinations were also shown to be as effective as, but less
`toxic than, combinations of three or more chemotherapy drugs.
`The efficacy is similar for several of these two-drug combina-
`tions. Trials of the SWOG and the ECOG compared five dif-
`ferent two-drug combinations in previously untreated patients
`with advanced NSCLC and found that they had similar efficacy
`in terms of tumor response rates and overall survival (Table 1;
`Refs. 12 and 13). Other large randomized trials from the United
`States and Europe have also shown the equivalence of a number
`of two-drug combination regimens, and superiority compared
`with single-agent chemotherapy regardless of whether the single
`agent is an older agent such as cisplatin or a newer agent such
`as paclitaxel, docetaxel, or gemcitabine (14–18).
`The development of NSCLC disease progression on chem-
`
`protein kinase; mAb, monoclonal antibody; IDEAL, IRESSA dose
`evaluation in advanced lung cancer; LCS, Lung Cancer Subscale;
`FACT-L, Functional Assessment of Cancer Therapy—Lung; INTACT,
`(cid:50)(cid:54)(cid:44)(cid:3)(cid:21)(cid:19)(cid:22)(cid:25)
`IRESSA NSCLC trial assessing combination therapy.
`(cid:36)(cid:51)(cid:50)(cid:55)(cid:40)(cid:59)(cid:3)(cid:57)(cid:17)(cid:3)(cid:50)(cid:54)(cid:44)
`(cid:44)(cid:51)(cid:53)(cid:21)(cid:19)(cid:20)(cid:25)(cid:16)(cid:19)(cid:20)(cid:21)(cid:27)(cid:23)
`
`

`

`5814 Targeting EGFR in NSCLC
`
`Table 1 Overview of efficacy results from large comparative trials
`of first-line chemotherapy regimens for advanced NSCLCa
`Trial end point
`SWOG 9509
`ECOG 1594
`Response, %
`25–28
`17–22
`Median survival, mo
`8.1–8.6
`7.4–8.1
`Time to tumor progression, mo
`4
`3.1–4.2
`1-year survival, %
`36–38
`31–36
`a See Refs. 12, 13.
`
`otherapy is inevitable, because these regimens do not result in
`cure. Until recently, there were no Food and Drug Administra-
`tion approved agents for use in the second-line setting. Do-
`cetaxel was approved on the basis of randomized trials in
`patients whose disease had progressed on platinum-based chem-
`otherapy (19, 20). The objective response rates of docetaxel
`were only 5–10% associated with a modest survival improve-
`ment. No agent produced tumor response in more than 5% of
`patients in the third-line treatment setting.
`The cytotoxic mechanism of action of chemotherapy
`agents imposes inherent limitations on their use. These agents
`nonspecifically kill normal proliferating cells and, as a result,
`are frequently associated with dose-limiting toxicities. Many of
`these effects, such as nausea, vomiting, and hair loss, are trou-
`bling to the patient but are not life threatening. Perhaps the most
`troubling effect is fatigue. Other frequent toxicities may be
`disabling even if not life threatening. Among these would be the
`neuropathy associated with paclitaxel and the severe fluid re-
`tention or effusion associated with docetaxel. All of the cyto-
`toxic chemotherapy agents produce hematological toxicities that
`are often life threatening and occasionally fatal. The careful
`observation of sequential blood counts and the i.v. infusions of
`chemotherapeutic agents and their supportive agents are expen-
`sive and inconvenient for the patient. Such toxicities often result
`in treatment adjustments such as dose reduction, delayed ad-
`ministration or, in some cases, discontinuation. Furthermore,
`with increasing rounds of chemotherapy, there is an increased
`risk that tumors will develop multidrug resistance, thus limiting
`future therapeutic options.
`Toxicities associated with chemotherapy may interfere
`with the ability of some patients with advanced NSCLC to
`receive the standard two-drug combination chemotherapy regi-
`mens. Such patients include the elderly (ⱖ70 years of age),
`patients with poor performance status, and patients with comor-
`bidities. Several studies in elderly patients show that less than
`one-third receive therapy although it may prolong survival (21).
`In addition, patients with a poor performance status of 2 expe-
`rienced a high rate of serious adverse events in the ECOG 1594
`study of combination chemotherapy regimens (13). The study
`design was subsequently amended to include only patients with
`an ECOG performance status of 0 or 1, because patients with
`poorer performance status are, in general, more likely to expe-
`rience adverse events.
`There is a need for new therapies with novel mechanisms
`of action that are well tolerated, effective, and convenient. The
`molecularly targeted agents that are in clinical development aim
`at inhibiting specific pathways and key molecules in tumor
`growth and progression, sparing normal cells. Examples of such
`
`agents that were recently approved by the Food and Drug
`Administration include trastuzumab, a mAb targeting the HER2/
`neu receptor protein in breast cancer;
`imatinib [Gleevec
`(STI571); Novartis], a small molecule receptor TK inhibitor
`targeting Bcr/abl in chronic myelogenous leukemia and c-Kit in
`gastrointestinal stromal tumors, (22, 23); and gefitinib, an orally
`active EGFR-TK inhibitor used as monotherapy in the treatment
`of patients with advanced or metastatic NSCLC who have failed
`to respond to platinum-based and docetaxel chemotherapies.
`Depending on the specific molecule targeted and the mechanism
`of inhibition, these agents may offer novel clinical benefits
`compared with outcomes with cytotoxic chemotherapy, or at
`least the minimum comparable benefits with reduced general
`toxicity and improved convenience.
`A variety of new approaches to treat lung cancer that
`target signal transduction pathways involved in angiogenesis,
`metastasis, and apoptosis (24 –26) are in clinical develop-
`ment. These agents inhibit a wide variety of tumor-associated
`molecules including matrix metalloproteinase, farnesyltrans-
`ferase, and a number of protein kinases. The various thera-
`peutic approaches to inhibiting these molecules include
`mAbs, small-molecule inhibitors, antisense oligonucleotides,
`biological response modifiers, and vaccines (24 –26). Among
`these various approaches are small-molecule inhibitors of
`tumor cell TKs. Gefitinib and imatinib have been validated
`clinically as new treatment approaches for malignancies
`(3–5, 23). Furthermore, the EGFR-TK inhibitors gefitinib
`and erlotinib have shown single-agent activity against ad-
`vanced, chemorefractory NSCLC in clinical trials described
`below (3–5, 27).
`
`EGFR-TK: A Molecular Target in NSCLC
`EGFR-TK Biology and Signaling in Solid Tumors.
`The EGFR is a cell surface receptor encoded by the HER1 (HER
`type 1) or ErbB1 gene (1). EGFR belongs to a family of receptor
`TKs that
`includes HER2/neu (ErbB2), HER3 (ErbB3), and
`HER4 (ErbB4). EGF and TGF-␣ are the two predominant
`ligands for EGFR (28, 29). The binding of these ligands to the
`extracellular domain of EGFR results in dimerization of the
`receptor monomer with another EGFR molecule or another
`member of the HER family (Fig. 1). Dimerization produces
`structural changes in the intracellular portion of the receptor that
`activate the TK domain. The enzymatic activity of EGFR-TK
`transfers phosphate moieties from ATP to specific tyrosine
`residues in the cytoplasmic tail of the EGFR protein. These
`phosphotyrosine residues then act as docking sites for various
`downstream effectors (Fig. 2). Some of these effectors are
`adapter molecules, such as growth factor receptor-bound protein
`2 (Grb2) and Src homology collagen protein (Shc), which serve
`as platforms to assemble the downstream signaling elements
`necessary for activating cellular proliferation (30). Other mole-
`cules are enzymes that are activated on EGFR-TK-dependent
`phosphorylation, including son of sevenless (SOS), PI3K, and
`Grb 2-associated binder-1 (Gab-1). Multiple major signal trans-
`duction pathways are initiated by EGFR autophosphorylation,
`including the Ras-MAPK signaling cascade, Src, and the signal
`transducers and activators of transcription (STAT) pathways,
`which are widely used by growth signals to induce gene tran-
`
`

`

`Clinical Cancer Research
`
`5815
`
`EGFR-TK (34). In addition, EGFR-TK interacts with com-
`ponents of the integrin pathway involved in cell– cell adhe-
`sion, which is crucial for tumor cell invasion of adjacent
`tissues (29, 35, 36). EGFR-TK also promotes invasiveness
`through the up-regulation or activation of matrix metallopro-
`teinases and stimulates tumor cell motility that further con-
`tributes to metastasis (37–39).
`EGFR-TK activation indirectly inhibits apoptosis in tu-
`mor cells, promoting tumor cell survival and resistance to
`cytotoxic therapies. This activity is mediated by PI3K, which
`activates Akt, an important signaling molecule in antiapo-
`ptotic pathways involving the transcription factor nuclear
`factor ␬B. Akt also regulates activity of the Ras-MAPK
`pathway, which is important for cellular proliferation (29).
`Interaction with signals from heterologous pathways, includ-
`ing those activated by stress inducers, neurotransmitters,
`hormones, and lymphokines, adds additional complexity to
`the EGFR-TK signaling network (31). These pathways in-
`volve G-protein-coupled receptors, which can transactivate
`EGFR. Cross-talk between EGFR and other receptors allows
`for EGFR-TK signals to activate other pathways.
`Increased expression of EGFR and its signaling pathways
`has been associated with a high percentage of tumors in the
`lung, breast, head and neck, colon, prostate, esophagus, and
`cervix (1, 2). These elevated levels of EGFR may be the result
`of transcriptional or posttranscriptional alterations or genomic
`mutation (34). Differences in the methodologies used and in the
`criteria for determining EGFR expression levels make it diffi-
`cult to compare study results (2, 40). Various methods of meas-
`
`Fig. 2 Key signal transduction pathways of activated EGFR-TK and
`the various pathways affected. Grb2, growth factor receptor-bound
`protein 2; SOS, son of sevenless; STAT, signal transducer(s) and acti-
`vator(s) of transcription; MEK, MAP/Erk (extracellular regulated ki-
`nase) kinase.
`
`Fig. 1 Homodimers and heterodimers within the human EGFR (HER)/
`ErbB family, and subsequent phosphorylation.
`
`scription and promote diverse cell responses. The particular
`dimer combinations that form at the cell surface after ligand
`binding determine which signaling molecules will be recruited
`to the surface (29–33).
`Ligand binding is the most extensively studied mecha-
`nism of EGFR-TK activation, but a variety of other cellular
`mechanisms are now known to influence EGFR-TK activity
`in tumor cells. For example, some mutations in the EGFR
`gene result in expression of EGFR proteins with constitu-
`tively activated TK activity, the most well-known being the
`EGFRvIII mutation. Defective inactivation mechanisms (e.g.,
`phosphatases, and receptor endocytosis and degradation) may
`also result in sustained signaling. Heterologous receptors and
`signal
`transduction pathways,
`including interactions or
`dimerization with other ErbB receptor types, have been
`shown to cross-activate EGFR-TK.
`Cellular proliferation as a result of EGFR-TK activation
`may occur via several signal transduction pathways; however,
`proliferation signals are strongly mediated by the MAPK path-
`way. After recruitment of adapter molecules on the activated
`EGFR complex, stepwise activation of Ras, Raf, MAP/Erk
`kinase (MEK1), and extracellular regulated kinase (Erk) pro-
`teins leads to increased activity of transcription factors, such as
`Elk1 and c-fos, key molecules that prime the cell for prolifera-
`tion and activate cell cycle progression (29). Activated EGFR
`has been shown to induce the expression of cyclin D, which is
`crucial in cell cycle progression and is commonly increased in
`solid tumors.
`Activated EGFR-TK also influences the malignant pro-
`gression of solid tumors. TGF-␣ and EGF induce angiogen-
`esis by up-regulating the expression of vascular endothelial
`growth factor (VEGF) in tumor cells. Increased microvessel
`density has been found in tumors that express activated
`
`

`

`5816 Targeting EGFR in NSCLC
`
`Fig. 3 Expression patterns of EGFR in NSCLC. Scoring on immunohistochemistry scale, staining intensities: 0, none (A); 1⫹, mild (B); 2⫹,
`moderate (C); 3⫹, strong (D). All of the panels are stained with an anti-EGFR antibody.
`
`uring EGFR levels in tumor tissues include immunohistochem-
`istry (Fig. 3), immunoassays, and assessment of RNA levels.
`Although some studies have shown a correlation between high
`expression of EGFR and decreased survival times, most studies
`of NSCLC patients have failed to show that EGFR expression is
`independently prognostic of survival (2, 34). The prognostic
`value of EGFR expression is increased when analyzed in con-
`junction with its dimeric partners, such as HER2/neu/ErbB2, or
`with ligands, such as TGF-␣ or EGF (41–43). High levels of
`EGFR in tumors result in an increase in EGFR ligand-binding
`sites and higher levels of the TK enzyme, as well as an increase
`in initiation sites for signal transduction inside the tumor cell.
`These findings indicate that there is an important role for aber-
`rant EGFR signaling in the development and progression of
`various human tumors. In addition, they provide a strong ration-
`ale for EGFR-TK as a target molecule for the development of
`new cancer therapies.
`EGFR-TK Inhibitors. Different approaches to inhibit-
`ing EGFR have resulted in a number of EGFR-targeted agents in
`clinical development including small-molecule EGFR-TK in-
`hibitors, mAbs, vaccines, immunotoxins, and recombinant li-
`gand–toxin fusion proteins (1, 44).
`Small-molecule EGFR-TK inhibitors act by blocking the
`ATP binding site of the EGFR-TK enzyme inside tumor cells
`(Fig. 4). On the basis of this mechanism of action, EGFR-TK
`
`inhibitors have the potential to inhibit all mechanisms of
`EGFR-TK activation, including constitutively activating mu-
`tations and receptor cross-talk. EGFR-TK inhibitors were
`designed to selectively inhibit EGFR-TK relative to other
`kinase enzymes present
`in normal
`tissues (28). Gefitinib
`erlotinib and CI-1033 (Pfizer) are among the EGFR-TK
`inhibitors in clinical development (Table 2; Refs. 1 and
`45– 48). Both gefitinib and erlotinib selectively and revers-
`ibly inhibit EGFR-TK, whereas CI-1033 is an irreversible
`pan-ErbB family inhibitor. The small-molecule EGFR-TK
`inhibitors also inhibit signals induced by EGFR heterodimer-
`ization with other members of the ErbB family. Compared
`with anti-EGFR mAbs such as cetuximab [Erbitux (C225);
`ImClone], EGFR-TK inhibitors offer the advantages of oral
`bioavailability and once-daily treatment.
`The targeted agent cetuximab, [Erbitux (C225); ImClone],
`is a chimeric mAb directed against the extracellular, ligand-
`binding domain of EGFR that competes with ligand for receptor
`binding (1, 49, 50). Cetuximab was not studied as a single agent
`in NSCLC but is currently being evaluated in combination with
`carboplatin/paclitaxel and cisplatin/gemcitabine in untreated pa-
`tients with stage IV NSCLC, and with docetaxel in patients with
`chemotherapy-refractory tumors. ABX-EGF (Abgenix) is an-
`other anti-EGFR mAb in Phase I clinical trials. mAbs can also
`be coupled with various toxic agents such as bacterial toxins or
`
`

`

`Clinical Cancer Research
`
`5817
`
`Pharmacodynamic studies indicate that EGFR-TK inhibi-
`tors and anti-EGFR antibodies block cell cycle progression in
`the G1 phase by up-regulating p27Kip1, a cell cycle inhibitor, and
`down-regulating c-fos, a transcriptional activator that is promi-
`nent in EGFR-mediated signaling (45, 52–55). Elevated levels
`of p27Kip1 block cell cycle progression in the G1 phase of
`growth. This sustains the hypophosphorylated state of the reti-
`noblastoma (RB) gene product, which is necessary to keep cells
`from progressing in the cell cycle (37, 56).
`The inhibition of tumor growth seen with EGFR-TK inhi-
`bition is also accompanied by decreases in vascular endothelial
`growth factor (VEGF), basic fibroblast growth factor (bFGF),
`and TGF-␣, all potent inducers of tumor angiogenesis (57).
`Thus, inhibitors of EGFR/EGFR-TK may also inhibit tumor
`growth by interfering with angiogenesis (58, 59). These obser-
`vations suggest that by inhibiting EGFR-TK, gefitinib and er-
`lotinib treatment alters expression levels of key molecules in
`tumor cells that are important for stimulating proliferation, cell
`cycle progression, tumor angiogenesis, metastasis, and inhibi-
`tion of apoptosis.
`
`Clinical Trials of EGFR-TK Inhibitors in NSCLC
`Phase I Trials. Several anti-EGFR agents have been tested
`alone or in combination with other agents in Phase I trials that
`included patients with NSCLC. Phase I trials of gefitinib followed
`two escalating dose schedules: (a) once-daily gefitinib given con-
`tinuously for 28 days; or (b) intermittent gefitinib, with 14 days on
`and 14 days off treatment (60–62). In the intermittent-dosing trials,
`doses ranged from 50 to 925 mg/day. In the continuous-dosing
`trials, doses ranged from 150 to 1000 mg/day. Tumor EGFR status
`
`Agent
`ZD1839
`gefitinib)
`
`OSI-774
`(erlotinib)
`
`CI-1033 (none)
`
`Clinical status
`Phase III
`(Approved)
`
`Phase III
`
`Phase I
`
`Table 2 EGFR-targeted agents in clinical developmenta
`Mechanism of action in vitro
`IC50 HER1 (KBb,c; CGd)
`HER2 (KBc; CGd)
`EGFR-TK inhibitor
`HER1 (23–79 nM; ⱕ80 nM)
`HER2 (3.7–10 ␮M; N/A)
`EGFR-TK inhibitor
`HER1 (2–20 nM; ⱕ100 nM)
`HER2 (0.2 ␮M; ⬍3 ␮M)
`EGFR-TK/HER2 inhibitor
`HER1 (1.7 nM; 7.4 nM)
`HER2 (5 nM; N/A)
`Anti-EGFR antibody
`
`Phase II/III
`
`Anti-EGFR antibody
`
`Phase I/II
`
`Vaccine
`
`Immunotoxin
`
`Phase II/III
`
`Phase II
`
`C225
`(cetuximab)
`ABX-EGF
`(none)
`EGF-P64K
`(none)
`DAB389-EGF
`(none)
`a See Refs. 45–48.
`b KB, kinase/binding inhibition; CG, cell growth inhibition; N/A,
`not available.
`c KB varies based on the source of purified HER1/2.
`d CG varies based on cell line tested.
`
`Fig. 4 The mechanism of action of EGFR-TK inhibitors in blocking
`signal transduction through EGFR-TK.
`
`radioactive particles, which may be used as delivery devices.
`The binding of the mAbs to the extracellular domain of EGFR
`triggers endocytosis of the receptor-immunotoxin complex to
`the cytoplasm, in which the various toxins act to inhibit protein
`synthesis and induce apoptosis (51). In another approach for
`targeting toxins to EGFR-expressing tumor cells, chimeric mol-
`ecules are created by fusing portions of the genes for ligands
`(EGF, TGF-␣) with a toxin gene. One example of such a
`toxin-fusion protein, DAB389-EGF, is in Phase II clinical trials
`for NSCLC (24).
`Both EGFR-TK inhibitors and anti-EGFR antibodies are
`effective in preclinical models for inhibiting the growth of a
`variety of human tumor cell lines, including lung, colorectal,
`breast, and prostate, suggesting their potential for broad appli-
`cability for solid tumor types (1). Preclinical studies also
`showed that EGFR inhibition results in synergy with chemo-
`therapy agents or radiation therapy in cell lines that are sensitive
`to EGFR inhibitors (52). For example, in cell viability assays,
`gefitinib treatment was synergistic with the cytotoxic chemo-
`therapy agents, vinorelbine and paclitaxel, and had additive
`effects with cisplatin (52). Similarly, gefitinib in combination
`with radiation has shown growth-inhibitory effects ranging from
`synergistic to additive in gefitinib-sensitive cell lines (52). Lung
`tumor xenografts have also been inhibited by gefitinib alone or
`in combination with chemotherapy agents (53). Gefitinib, erlo-
`tinib, and cetuximab have all been shown to potentiate the
`antitumor effects of most cytotoxic agents, including platinum-
`based chemotherapy agents in preclinical models with cell lines
`sensitive to EGFR inhibition. Gefitinib also showed activity
`against NSCLC xenografts in combination with taxanes, doxo-
`rubicin, and antifolates (45, 52–54).
`
`

`

`5818 Targeting EGFR in NSCLC
`
`was not an eligibility requirement in these trials. Patients were
`selected based on having tumor types that were all known to
`express EGFR at a high rate, including NSCLC and colorectal,
`prostate, head and neck, ovarian, breast, renal, and pancreatic
`cancers.
`In total, 252 patients were recruited for Phase I trials of
`gefitinib (60–62). Almost all of the patients had received prior
`treatment with radiotherapy and/or chemotherapy, and many
`had received multiple prior chemotherapy regimens. Of the
`patients enrolled, 100 had advanced, previously treated NSCLC.
`The most common adverse events were diarrhea, acneiform
`rash, nausea, asthenia, and vomiting. The majority of adverse
`events were grade 1 or 2 and transient; grades 3 and 4 events
`were rare (62). Dose-limiting toxicities including reversible
`diarrhea and rash occurred at daily doses of 700 to 800 mg.
`Pharmacokinetics were consistent with once-daily dosing
`(60, 62).
`Histopathological studies of pretreatment and posttreat-
`ment skin biopsy specimens from Phase I trials showed that
`treatment with gefitinib suppressed EGFR phosphorylation, in-
`hibited MAPK activity, reduced the proliferation index as
`judged by staining for Ki67 (a nuclear proliferation-associated
`antigen), and increased both the apoptotic index and the expres-
`sion of p27Kip1 (55). The stratum corneum of the epidermis was
`significantly thinner in posttreatment skin samples (55). In a
`separate study, tumor biopsy samples obtained after 28 days of
`treatment with gefitinib showed decreased levels of activated
`signal
`transduction molecules compared with biopsies from
`baseline samples (63).
`In the Phase I trials of gefitinib in 100 NSCLC patients,
`partial responses were observed in 10% of patients, and disease
`stabilization was seen in 13% of patients (61). There were no
`obvious differences between the continuous- and intermittent-
`dose schedules with respect to response or toxicity. Antitumor
`activity occurred at all dose levels, with no clear dose-response
`relationship. Stable disease was observed at daily doses as low
`as 50 mg (n ⫽ 1), and partial responses were achieved at 150 mg
`(n ⫽ 2; Ref. 61). In many cases, there was anecdotal evidence
`of symptom amelioration with gefitinib treatment in the absence
`of an objective tumor response. Of the 100 patients with
`NSCLC, 28% remained on gefitinib for at least 3 months and
`20% for at least 6 months.
`Erlotinib was investigated in a Phase I trial of 40 patients
`with previously treated advanced solid tumors,
`including 4
`patients with NSCLC (64). Dose levels included 25, 50, 100,
`and 200 mg/day. Expression of EGFR in the tumor was not an
`eligibility requirement in this trial. However, EGFR expression
`was assessed immunohistochemically to determine whether pa-
`tients with EGFR-positive tumors were more likely to benefit
`from treatment with erlotinib (64). Tumor response did not show
`a correlation with EGFR expression (64). The most common
`adverse events were diarrhea and skin toxicities. These adverse
`events established the maximum tolerated dose at 150 mg/day.
`Most adverse events were grade 1 or 2 and were reversible. In
`this trial, there was one complete response (renal cell carci-
`noma), one partial response (colorectal), and several patients
`with stable disease (one of four with NSCLC).
`The mAb cetuximab, which is administered i.v., was in-
`vestigated in Phase I trials involving weekly administration of
`
`cetuximab, alone or in combination with cisplatin (65). Patients
`recruited for these studies had advanced solid tumors that over-
`expressed EGFR as documented through immunohistochemistry
`on tumor biopsies. The multiple-dose monotherapy study was
`conducted in 17 patients who had previously received chemo-
`therapy according to the standard of care for their particular
`tumor type. Another trial evaluated cetuximab in combination
`with chemotherapy in 22 patients with head and neck cancer or
`patients with NSCLC who had not been previously treated with
`a platinum agent. The dose levels of cetuximab investigated
`were 5, 20, 50, and 100 mg/m2 (65).
`The most common adverse events in the Phase I trials of
`cetuximab were fever and chills; asthenia; transaminase eleva-
`tion; nausea; and skin toxicities, including flushing, seborrhea,
`and acneiform rashes. The majority of adverse events were
`grade 1 or 2. The maximum tolerated dose was not reached in
`these trials. Because anti-EGFR mAbs such as cetuximab are
`administered i.v., gastrointestinal
`toxicities are not as pro-
`nounced as with the orally administered small-molecule
`EGFR-TK inhibitors. However, the mAb-based treatments oc-
`casionally induce immunological responses. Of the 189 patients
`treated with cetuximab in early-phase trials, 2% experienced
`grade 3 allergic reactions and 2% experienced grade 4 reactions.
`Allergic reactions to cetuximab were managed with standard
`interventions (66). Several patients in these trials experienced
`disease stabilization, particularly those with head and neck
`cancer (65).
`Randomized Phase II Trials in Advanced Refractory
`NSCLC. There were two large randomized Phase II trials that
`evaluated two doses of gefitinib in advanced, chemotherapy-
`refractory NSCLC patients. These trials were termed IDEAL-1
`and IDEAL-2 (3, 4). IDEAL-1 was a global
`trial with an
`enrollment of 210 patients with NSCLC who had failed one or
`more previous chemotherapy regimens (3). IDEAL-2 was a
`United States trial with an enrollment of 216 patients with
`NSCLC who failed two or more previous chemotherapy regi-
`mens that included a platinum agent and docetaxel (4). In both
`trials, patients were randomized to receive treatment with ge-
`fitinib at 250 mg/day or 500 mg/day. Unlike cytotoxic agents,
`for which the dose is dictated by toxicities, these once-daily oral
`doses of gefitinib were selected for study based on optimal
`biological doses that are well below the maximum tolerated
`dose (62).
`Objective tumor response (ⱖ50% inhibition of tumor
`mass), which was evaluated every 4 weeks by radiographic
`assessment, was a primary end point of both IDEAL trials. A
`co-primary end point in IDEAL-1 was safety, whereas symptom
`improvement was a co-primary end point in IDEAL-2. The LCS
`of the FACT-L questionnaire was used to assess symptom
`improvement. Because these novel targeted therapies are largely
`devoid of systemic toxicities, improvement in disease-related
`symptoms is an important basis for assessing their utility.
`The results of IDEAL-1 and IDEAL-2 are summarized in
`Table 3 (3, 4). In IDEAL-1, the following data were obtained for
`the 250 mg/day and 500 mg/day groups, respectively: objective
`tumor response rate, 18.4% versus 19%; disease control rate,
`54.4% versus 51.4%; progression-free survival, 2.7 months versus
`2.8 months; median survival, 7.6 months versus 7.9 months. Ob-
`jective tumor response rates were similar for patients who received
`
`

`

`Clinical Cancer Research
`
`5819
`
`Table 3 Results from Phase II studies of patients with NSCLC treated with gefitiniba
`IDEAL-1
`IDEAL-2
`
`Trial end point
`Objective tumor response, %
`Symptom improvement, %
`Disease control, %
`Overall survival, mo
`a See Refs. 3, 4.
`
`250 mg/day
`(n ⫽ 103)
`18.4
`40.3
`54.4
`7.6
`
`500 mg/day
`(n ⫽ 105)
`19.0
`37
`51.4
`7.9
`
`250 mg/day
`(n ⫽ 102)
`11.8
`43.1
`43.0
`6.1
`
`500 mg/day
`(n ⫽ 114)
`8.8
`35.1
`35.0
`6.0
`
`gefitinib as second-line versus third-line treatment (17.9% versus
`19.8%; Ref. 3). Adenocarcinoma, which characteristically ex-
`presses less EGFR than does squamous cell carcinoma, was iden-
`tified as one of the prognostic factors associated with objective
`response (3.5 times more likely to respond to treatment than other
`tumor histologies; Ref. 3). The overall symptom improvement rate,
`as measured by the LCS, was 40 and 37%, respectively, and
`improvement occurred rapidly with a median time to improvement
`of 8 days (3). Among responders, 78% exhibited symptom im-
`provement based on the LCS and 53% reported a quality-of-life
`improvement as measured by FACT-L (67).
`The most frequent adverse events were generally