Cellular responses to EGFR inhibitors and their relevance to cancer therapy

Abstract

EGFR is a trans-membrane receptor tyrosine kinase that belongs to the HER family of receptors. The EGFR family plays an essential role in normal organ development by mediating morphogenesis and differentiation. Unlike normal cells that have tight regulatory mechanisms controlling EGFR pathways, tumor cells often have dysregulated EGFR signaling through receptor overexpression and/or mutation. This leads to proliferation under adverse conditions, invasion of surrounding tissues, and increased angiogenesis as well as resistance to radiation and chemotherapy. Therefore, EGFR is a legitimate therapeutic target. Numerous EGFR inhibitors are under development, but to date only four of them are FDA-approved, including two that inhibit the receptor's intracellular tyrosine kinase activity (gefitinib and erlotinib) and two that block extracellular ligand binding (cetuximab, and most recently panitumumab). In this review, we focus on how these different inhibitors affect EGFR signaling and the mechanisms by which they potentiate the effects of chemotherapy and radiation therapy. Numerous clinical trials have been conducted with these agents either as monotherapy, in combination with chemotherapy, or concurrently with radiation. Unfortunately, many of the clinical trials reported so far have shown at best limited gains; therefore, understanding the actions of these agents is essential to improving their efficacy in the treatment of cancers.

INTRODUCTION

Despite significant advances in systemic therapies, radiation oncology, and surgical techniques, many patients with cancer are still incurable. A novel therapeutic approach has been to target the epidermal growth factor receptor (EGFR), which is often mutated and/or overexpressed in many tumors and regulates proliferation, apoptosis, angiogenesis, tumor invasiveness, and distant metastases [1, 2]. Specifically, inhibition of the EGFR signaling pathways has been accomplished extracellularly with specific antibodies to block ligand binding or intracellularly with small molecule inhibitors. This review will focus on the cellular responses to these EGFR inhibitors and their implications for cancer therapy.

EGFR SIGNALING

EGFR is a trans-membrane receptor tyrosine kinase that belongs to the HER family of receptors [3]. To date four member of this family have been identified including EGFR (HER1/erbB-1), HER2 (erbB-2/neu), HER3 (erbB-3) and HER4 (erbB-4). A wide variety of cancers express EGFR (Table 1). The N-terminus extracellular portion of EGFR can bind a variety of ligands. Based on their affinities for various receptors, these ligands are divided into three different groups. Epidermal growth factor (EGF), transforming growth factor β (TGF-β), and amphiregulin bind to EGFR; betacellulin, heparin-binding growth factors, and epiregulin can interact with both EGFR and ErbB4; and finally tomoregulins and heregulins bind to ErbB4 and occasionally to ErbB3 [4, 5]. HER2 has no known ligand, but it is the preferred heterodimerization partner for EGFR. Upon binding of ligand, multiple receptors, which normally exist as monomers, aggregate together at the cell surface to increase the probability of dimerization [6]. These molecules homodimerize with themselves or heterodimerize with other members of the family, bringing the intracellular C-terminal tyrosine kinase domains in close proximity to each other, resulting in autophosphorylation. Phosphorylation of these tyrosine residues allows docking of second messenger proteins that contain either Src homology 2 (SH2) or phosphotyrosine binding (PTB) domains to become activated [7], which then leads to activation of multiple downstream pathways.

TABLE 1.

^*

Tumor type	Tumors Expressing EGFR (%)
Colorectal	25–77
Pancreatic	30–50
Bladder	31–48
Prostate	39–47
Ovarian	35–70
Glioma	40–63
Breast	14–91
Lung	40–80
Renal	50–90
Head and neck	80–100

^*Adapted from Grandis et al [2]

The EGFR family plays an essential role in normal organ development by mediating morphogenesis and differentiation through effects on cell proliferation, differentiation, apoptosis, invasion, and angiogenesis [8, 9]. Unlike normal cells that have tight regulatory mechanisms controlling EGFR pathways, tumor cells often have dysregulated EGFR signaling, allowing them to proliferate under adverse conditions, invade surrounding tissues, and increase angiogenesis. Normally EGFR must be activated by a ligand to initiate downstream signaling, but tumor cells can circumvent this requirement through a number of mechanisms [10]. First, many cancers overexpress wildtype EGFR, which leads to ligand-independent EGFR activation. Secondly, cancers can also develop EGFR mutations, causing the receptor to be inappropriately activated. These mutations include point mutations or deletions in the tyrosine kinase activation domain, sometimes seen in non-small cell lung cancer (NSCLC) [11, 12], and deletion in the extracellular domain (EGFRvIII variant), seen frequently in glioblastomas [13-15]. Lastly, some cancers overexpress TGF-β, a ligand for EGFR, thereby establishing their own autocrine loop [16].

Numerous signaling pathways originate downstream of the EGFR [17] (see Figure 1). EGFR stimulation leads to Ras activation [18], which activates the extracellular signal pathway regulated kinase (ERK)/mitogen-activated protein kinase (MAP) kinase pathway. Activation of EGFR also leads to increased PI3 kinase activity, either directly or through Ras. PI3 kinase phosphorylates PIP2 to generate PIP3, which then recruits the serine/threonine kinases PDK1 and Akt to the inner cell membrane [19, 20]. Akt is then activated by being phosphorylated at both Ser 473 and Thr 308, the latter in response to PDK1. PTEN is a phosphatase that opposes the action of PI3 kinase, thereby reducing the level of phosphorylated (activated) Akt [21]. EGFR activation also results in phospholipase C gamma (PLC-γ) phosphorylation and subsequent hydrolysis of phosphatidylinositol 4,5 biphosphate (PIP2) into inositol 1,4,5-triphosphate (IP3) and diacylglycerol (DAG), leading to protein kinase C (PKC) activation [22]. Other pathways activated by EGFR stimulation include the STAT3 pathway [23] and the p38 stress pathway.

Figure 1. Signaling pathways downstream of EGFR activation.

Binding of ligand (EGF, TGF-β, etc,) to EGF receptor leads to aggregation of multiple receptors (not shown in diagram). Dimerization of an EGF receptor with another ErbB family member leads to phosphorylation of tyrosine residues located in the intracellular domain. This allows for docking of proteins, leading to activation of the Ras/Raf/Erk, PI3K/Akt, PLC-γ, and JAK/Stat pathways.

CLASSES OF EGFR INHIBITORS

Numerous strategies have been explored to target EGFR to inhibit tumor growth including monoclonal antibodies (MoAbs; e.g. cetuximab, panitumumab), small molecule tyrosine kinase inhibitors (TKIs; e.g. gefitinib, erlotinib), ligand-linked toxins, and antisense oligonucleotides. MoAbs block ligand from binding to the extracellular domain of the receptor whereas TKIs target the ATP-binding pocket of the cytoplasmic domain to inhibit receptor phosphorylation. To date, the FDA has approved only two MoAbs for clinical use, cetuximab and panitumumab and two TKIs, erlotinib and gefitinib.

In the early 1980s, a murine monoclonal antibody to the EGFR was demonstrated to inhibit the in vitro proliferation of Harvey sarcoma virus-transformed A431, a squamous cell carcinoma line that expresses up to 200-fold higher levels of EGFR than non-malignant cells [24, 25]. Because many pathologic specimens from epithelial malignancies demonstrated elevated EGFR, it was postulated that inhibition of EGFR-induced signaling might prove useful therapeutically. Initial in vivo experiments with xenografts of human tumors expressing EGFR in athymic mice demonstrated dose-dependent growth inhibition [26]. This antibody, known as C225, was humanized to create cetuximab (Erbitux^R; ImClone System, Princeton, NJ). After the antibody binds to the EGFR, the receptor is internalized, then degraded, leading to receptor downregulation at the cell surface. The receptor is prevented from autophosphorylation and activation; therefore, downstream signaling is inhibited. However, studies by Mandic et al., have suggested that the antibody may transiently stimulate EGFR phosphorylation prior to downregulation [27].

A number of other anti-EGFR MoAbs are in development including EMD72000 (matuzumab) and hR3 (nimotuzumab), both of which are humanized mouse MoAbs like cetuximab. In contrast, panitumumab (Vectoibix^R; Abgenix, Fremont, CA) is a fully humanized antibody. It blocks ligand binding and causes receptor internalization, but not degradation [28]. Theoretically, such a fully humanized antibody should elicit fewer side effects. Cetuximab is FDA-approved for use in combination with irinotecan in the treatment of metastatic colorectal cancer, in combination with radiation for locally/regionally advanced SCCHN, and as monotherapy for recurrent/metastatic SCCHN after failing platinum-based chemotherapy. Panitumumab was recently approved by the FDA for the treatment of metastatic colorectal cancer refractory to standard chemotherapy agents.

Gefitinib (ZD1839, Iressa^R, Astra Zeneca, Wilmington, DE) is a reversible TKI that has been tested extensively in both pre-clinical models and in clinical trials. The drug has been shown to inhibit EGFR tyrosine kinase activity as low as the nanomolar range by competitively blocking the intracellular ATP-binding domain of the receptor [29, 30]. In contrast to cetuximab, gefitinib induces neither EGFR internalization nor degradation and thus does not decrease the level of EGFR protein.

In vitro and in vivo studies have demonstrated growth inhibition of multiple cell lines by gefitinib [31]. Studies using xenografts of human tumors derived from, ovarian, colon, lung, vulval, breast, and hormone-refractory prostate cancers showed that gefitinib potentiated the cytotoxic effects of many chemotherapeutic agents [32]. However, as will be discussed later, clinical trials have shown only modest efficacy of gefitinib as both a single agent and as part of a combination regimen in the treatment of patients with NSCLC. Hence, although the FDA had initially given wider approval to gefitinib for the treatment of NSCLC, because of these unimpressive results, it is now available only for patients who have failed both platinum-based and docetaxel chemotherapy and had previously benefited from gefitinib.

Erlotinib (OSI-774, Tarceva™, OSI Pharmaceuticals in collaboration with Genentech and Roche) potently and reversibly inhibits EGFR tyrosine kinase activity of both wild-type EGFR and the constitutively active mutant EGFRvIII at concentrations at nanomolar concentrations in vitro. Like gefitinib, erolotinib does not decrease the level of EGFR protein. Many clinical trials have been conducted to test the efficacy of erlotinib, some of which show positive results. Side effects of the erlotinib in these trials included skin rash and diarrhea [33]. Erlotinib is currently FDA-approved for the treatment of patients with locally advanced or metastatic NSCLC after failure of first-line chemotherapy and as first-line therapy for locally advanced, unresectable or metastatic pancreatic cancer in combination with gemcitabine.

TKIs with different specificities than gefitinib and erlotinib are under development. PKI-166 (Novartis International, Basel, Switzerland), GW572016 (GlaxoSmithKline, Research Triangle Park, NC), and ARRY-334543 (Array BioPharma, CO) inhibit both EGFR and ErbB-2-reversibly. In contrast, EKB-569 (Wyeth-Ayerst, Madison, NJ) is an irreversible dual inhibitor of these two receptors. High levels of HER2/neu can potentiate EGFR signaling making it difficult for EGFR inhibitors to block EGFR phosphorylation [34], hence scientific rationale exists for inhibiting both receptors. Only the results of clinical trials will tell whether these newer inhibitors have better efficacy than the older TKIs.

RECEPTOR MUTATIONS AND RESPONSE TO INHIBITORS

Mutations in genes encoding for EGFR or its signaling molecules are observed in many tumors. Glioblastomas often express a mutant variant of EGFR, known as EGFRvIII, which constitutively activates PI3K and confers enhanced tumorigenicity [14]. Mellinghoff et al. studied patients with glioblastomas who had been treated with EGFR kinase inhibitors [35]. Their study demonstrated that patients with co-expression of EGFRvIII and PTEN were more likely to show a radiologic response to an EGFR inhibitor. Furthermore, glioblastoma cells co-expressing these two molecules were sensitive to erlotinib. A possible explanation is that loss of PTEN might activate the Akt pathway independently of EGFR and render it insensitive to EGFR inhibition. These results suggest that identification of patient populations with certain mutations may lead to specifically directed therapies.

EGFR is overexpressed in 80% of NSCL and mutated in a smaller percentage. Pao et al. found that 81% of patients with Stages I-IIIA NSCLC who responded to gefitinib or erlotinib had mutations either in or around exon 19 [36]. Paez, et al. analyzed NSCLC specimens from patients who did and did not respond to gefitinib. Tumors from patients whose disease progressed on gefitinib showed no mutations in exons 18 to 24. In contrast, specimens from gefitinib-responsive tumors contained mutations within the ATP-binding pocket located in the EGFR-kinase activation domain [12]. Likewise, Lynch et al. showed that 8 of 9 patients who responded to gefitinib had a mutation of EGFR in the kinase activation domain whereas all seven patients who failed to respond to gefitinib had wildtype EGFR [11]. These last two studies both examined the effect these EGFR mutations in cell lines. Cell lines with mutant EGFR showed enhanced tyrosine kinase activity in response to EGF, and a minimal concentration of gefitinib was able to completely inhibit EGFR autophosphorylation in these cells. In contrast, cell lines with wildtype EGFR required concentrations 100-fold higher to achieve the same level of inhibition. These three studies showed that certain groups of patients were more likely to have EGFR mutations in the kinase activation domain and consequently more likely to respond to gefitinib: women, patients with adenocarcinoma or bronchoalveolar cell carcinoma, non-smokers, and Asians.

These mutations in the kinase activation domain of EGFR activate the receptor. Tumors that contain these mutated receptors are felt to be “addicted” to them in the sense that their survival is absolutely dependent on this activation [37]. Hence if the pathway is disrupted through the use of inhibitors, these cell readily die. This term of oncogene addiction was first described in characterizing the response of chronic myelogenous leukemia to Gleevac monothearpy [37, 38].

ANTI-PROLIFERATIVE EFFECTS

EGFR inhibition has significant effects on cellular proliferation. Huang et al. showed that micromolar concentrations of gefitinib inhibited cell proliferation in a dose-dependent manner in SCCHN cell lines [39]. Even cell lines with relatively low levels of EGFR expression showed modest levels of inhibition. Flow cytometric analysis demonstrated that gefitinib treatment led to an accumulation of cells in the G₁ phase with a simultaneous decrease in cell numbers in S phase. This G1 phase arrest induced by EGFR inhibition is thought to be due to an increase in the levels of the cyclin-dependent kinase inhibitor p27 [40].

EFFECTS ON MIGRATION AND INVASION

EGFR activation results in PLC-γ phosphorylation and subsequent hydrolysis of phosphatidylinositol 4,5 biphosphate (PIP2) into inositol 1,4,5-triphosphate (IP3) and diacylglycerol (DAG) (see Figure 1). This in turns leads to the release of gelsolin and other actin-binding proteins that can modify the actin cytoskeleton [41, 42]. Hydrolysis of PIP2 occurs preferentially at the leading edge of a cell, which provides asymmetry in intracellular signaling and leads to the formation of lamellopodia during EGFR-mediated migration. Inhibition of PLC-γ blocks glioma cell motility and invasion into fetal rat brain aggregates [43] as well as EGFR-mediated migration of breast carcinoma cells [44].

The PI3K pathway, which is also downstream of EGFR activation, is involved in migration as well [45] through PIP3 accumulation, which leads to the activation of Rho family GTPases including Rac and/or cdc42 [46]. Activated Rac recruits a protein complex that leads to cycles of activation and inactivation of gelsolin. This leads to severing and dissolution of actin filament networks and creation of sites of actin reassembly. PIP3 may also enhance PLC-γ-catalyzed hydrolysis [47] and provide a mechanism by which the two pathways may cooperate to increase cell motility. Also, evidence suggests that PI3K can influence EGF-mediated motility by regulating translocation of PLC-γ to the leading edge of a migrating cell [44].

Other pathways that may also be involved in EGFR-mediated migration and invasion include the MAP kinase and protein kinase C delta (PKC-δ) pathways [48, 49]. PKC-δ was shown to be overexpressed in two invasive prostate cell lines relative to normal prostate epithelial cells [50]. Furthermore, inhibition of PKC-δ with a chemical inhibitor and specific siRNA were able to decrease invasion and migration of both cell lines. Whether these downstream pathways are also directly affected by EGFR inhibition remains to be determined.

ANTI-ANGIOGENIC RESPONSE

Decreased expression of vascular endothelial growth factor (VEGF), a key angiogenic factor, may account for some of the inhibition of tumor growth by EGFR blockade in vivo. EGF induces VEGF expression in many cell lines [51, 52]. Conversely, our own data and that of many other groups indicate that pharmacological inhibition of EGFR can decrease VEGF expression and consequently angiogenesis in many tumor types [31, 39, 53-58]. Particularly interesting is a study that showed that the development of resistance to anti-EGFR antibody therapy in human tumor xenografts correlated with VEGF overexpression [59]. Furthermore, this acquired lack of responsiveness to the anti-EGFR therapy could be mimicked by VEGF overexpression engineered by gene transfection.

C225 has been demonstrated to enhance the inhibitory effects of antisense VEGF in tumor xenografts of human GEO colon cancer cells in immunodeficient mice [54]. Mice treated with the combination of C225 and antisense VEGF demonstrated improved survival and local tumor control than with either agent alone. Based on this and other preclinical studies, clinical trials are now being conducted using erlotinib in combination with the anti-VEGF antibody bevacizumab (Avastin, Genentech, San Francisco, CA) in advanced NSCLC [60]. The combination of bevacizumab, conventional chemotherapy, and an anti-EGFR antibody (panitumumab or cetuximab) is being evaluated in metastatic colorectal carcinoma [61]. Furthermore, agents that are dual EGFR/VEGF inhibitors, such as ZD6474 (Astra-Zeneca, Wilmington, DE) have been studied pre-clinically [62] and are now in clinical trials.

RADIOSENSITIZATION

Growth factors have been shown to mediate cellular responses to radiation in numerous tissues. Some studies have shown that EGF treatment leads to radiosensitization [63-65]. However, in retrospect, it is very possible that radiosensitization with EGF in these studies occurred because prolonged exposure to EGF resulted in degradation of EGFR. In fact, most studies have shown the opposite, that activating the EGFR pathway leads to decreased sensitivity to radiation. For example, treatment of MCF-7 breast cancer cells with EGF prior to irradiation led to increased radioresistance, which was inhibited by antibodies to EGFR [66]. Furthermore, cell lines derived from head and neck tumors and xenografts that overexpress EGFR have been observed to be radioresistant [67, 68]. Akimoto et al. found an inverse correlation between EGFR expression and the radiocurability of murine carcinomas [69]. Transfection of the mutant EGFRvIII receptor into astrocytes led to radioresistance [70].

Numerous studies have examined the effect of EGFR inhibition on the radiation response. Harari and colleagues found that the EGFR MoAb C225 enhanced the radiosensitivity of HNSCC as measured by clonogenic survival [71]. C225 also augments radiation killing in vivo in HNSCC xenografts implanted in nude mice [55]. The same group showed that the small molecule inhibitors gefitinib and erlotinib also increased both in vitro and in vivo radiosensitivity [39, 72]. Other groups have confirmed that C225 or gefitinib leads to enhanced killing in response to radiation in vitro and in vivo using diverse cell types including HNSCC, colon, ovarian, NSCLC, and breast cancer lines [73-76].

How EGFR inhibitors increase sensitivity to radiation is not completely understood. The C225 antibody causes an increase in the proportion of cells in G₁, which is a more radiosensitive phase, and a concomitant decrease in the proportion in the S phase, which is more radioresistant [71]. Gefitinib [77] and erlotinib [72] also cause this cell cycle redistribution, which could contribute to radiosensitivity. Another potential mechanism of radiosensitization is via increased apoptosis. Huang et al. showed that the C225 antibody increased radiation-induced apoptosis at the same time that it increased expression of Bax, a pro-apoptotic protein, and decreased expression of Bcl-2, an anti-apoptotic protein [71]. Likewise, gefitinib has also been found to increase radiation-induced apoptosis [39, 76]. However, in epithelial cells, apoptosis may constitute a relatively small component of cell deaths induced by radiation [78]

There are data suggesting that EGFR inhibition can decrease DNA repair following radiation. Based on their experimental data, Dittmann, et al. have proposed that ionizing radiation triggers EGFR import into the nucleus [79]. According to their model, EGFR interacts with DNA-PK, a protein involved in DNA damage repair, thus leading to increased nuclear activity of DNA-PK. The same group also showed that EGFR blockade with the C225 antibody abolished EGFR import into the nucleus and radiation-induced activation of DNA-PK, thus inhibiting DNA repair and increasing the radiosensitivity of treated cells [80]. Consistent with this model, Huang and Harari [39] showed that C225 decreased the repair of sublethal damage following radiation and caused redistribution of DNA-PK from the nucleus to the cytosol. This shift in DNA-PK from the nucleus to the cytosol has been documented by other groups in response to either anti-EGFR monoclonal antibodies or gefitinib [81].

Treatment interruptions are associated with worse local control for certain cancers, particularly SCCHN [82]. One explanation for this is that after many doses of radiation, cells that are not killed may actually be induced to proliferate faster. It has been hypothesized that the mechanism underlying this phenomenon termed “accelerated repopulation” may be the activation of EGFR signaling by radiation. In fact, irradiation of cells in vitro and tumors in vivo has been shown to lead to EGFR phosphorylation and activation of downstream signaling pathways [69, 83-85]. However, it is not clear whether this actually occurs in patients undergoing irradiation. Eriksen, et al. found some support for this hypothesis in their study showing that patients with SCCHN achieved better local control with an accelerated radiotherapy schedule, but only if their tumors overexpressed EGFR [86]. If ongoing studies continue to provide further evidence that EGFR overactivity may be responsible for the phenomenon of accelerated repopulation, then inhibition of downstream kinase activity may potentially be an alternative to accelerated radiotherapy for overcoming repopulation.

In addition to the mechanisms discussed above that are apparent in vitro (increased apoptosis, cell cycle redistribution, decreased DNA repair and inhibition of accelerated repopulation), there may be additional factors that are only important in vivo. As discussed previously, EGFR inhibition has effects on VEGF/angiogenesis and migration/invasion that could increase in vivo radiosensitivity. Radiation itself can upregulate the expression of VEGF, and there are reports in the literature that suggest that decreasing VEGF expression following radiation can augment tumor control in vivo [87-89].

CHEMOSENSITIZATION

EGF treatment of cells has been shown to be associated with resistance to chemotherapeutic agents including doxorubicin and topoisomerase II inhibitors [90, 91]. Fan, et al. found significant tumor regression of A431 squamous cell carcinoma xenografts in nude mice treated with C225 in combination with cisplatin whereas either agent by itself had no effect [92]. Likewise, studies with doxorubicin in combination with anti-EGFR monoclonal antibodies demonstrated marked tumor regression of A431 and MDA-468 xenografts in athymic mice, but only temporary growth inhibition when either agent was used by itself [93]. Chemosensitization has also been demonstrated with small molecule EGFR inhibitors. Sirotnak, et al. reported that the combination of gefitinib with platinum-containing chemotherapy resulted in several-fold higher growth inhibition in human tumor cell explants in mice compared to chemotherapy alone [32]. Gefitinib potentiated growth inhibition by chemotherapy irrespective of EGFR expression levels.

The effect of EGFR inhibition on chemosensitization may be related to suppression of DNA damage repair. Friedmann, et al. found that the addition of gefitinib delayed repair of DNA strand breaks after etoposide treatment and delayed repair of DNA interstrand cross-links after cisplatin treatment [94]. Gefitinib was also found to inhibit the removal of oxaliplatin DNA adducts [95].

Similar to ionizing radiation, chemotherapeutic agents such as cisplatin, [96] oxaliplatin, 5-FU [97] and paclitaxel [98] can lead to increased EGFR phosphorylation and activation. EGFR inhibitors could increase chemotherapy-induced killing by interfering with this response. Consistent with this idea, the ability of gefitinib to cooperate in cell killing by oxaliplatin or 5-FU correlated with the ability of these two chemotherapeutic agents to induce EGFR phosphorylation [97].

In the “Radiosensitization” section, it was mentioned that the accumulation of cells in the G1 phase by EGFR inhibitors could potentially explain their radiosensitization since cells are more radiosensitive in this phase of the cell cycle. However, this cell cycle redistribution could have an adverse effect in terms of response to chemotherapy. Chun, et al. found that gefitinib followed by gemcitabline was less toxic compared with the reverse combination [99]. This was hypothesized to be partly due to the fact that the former sequence led to an accumulation of cells in the G1 phase prior to exposure to gemcitabine, which has greater toxicity for cells in S phase. Similar sequence-dependent differences in cytotoxicty have been demonstrated for oxaliplatin [95] and a topoisomerase inhibitor [100] although these agents are not S-phase specific in their killing. Therefore, the precise sequencing of EGFR inhibitors in regard to chemotherapeutic agents could be an important factor in their efficacy in clinical trials.

CLINICAL TRIALS: RADIOTHERAPY AND EGFR INHIBITION

The addition of cetuximab to radiotherapy in SCCHN has led to improved overall survival and locoregional control in a clinical phase III trial. Bonner, et al., randomized 424 patients with locally advanced SCCHN to radiotherapy plus weekly cetuximab or radiotherapy alone [101]. After a median follow-up of 54 months, overall survival was 55% months at three years in the combined therapy arm versus 45% in the radiation alone arm (median 49 months vs. 29 months), a difference that reached statistical significance (p=0.03). Local control was also significantly improved in the radiation plus cetuximab arm at 50% versus 41% in the radiation alone arm (median 24 months versus 15 months). Cetuximab was associated with the development of acneiform rash in a majority of patients and hypersensitivity reactions in four patients. A major critique of this study is that none of the patients received chemotherapy, which is routinely used today in combination with radiation in patients with SCCHN. However, it is proof of principle that EGFR inhibition can improve the efficacy of radiotherapy in patients. There is currently a randomized trial underway (RTOG 0522) in which patients are randomized to radiation and cisplatin versus the same regimen plus cetuximab.

CLINICAL TRIALS: EGFR INHIBITORS AS MONOTHERAPY OR IN COMBINATION WITH CHEMOTHERAPY

I. Gefitinib and non-small cell lung cancer

While some tumors clearly exhibit dramatic shrinkage in response to EGFR inhibitors [11, 12], clinical phase III randomized trials have not shown a survival advantage with gefitinib. In the phase III ISEL trial for locally advanced or metastatic NSCLC, gefitinib monotherapy failed to demonstrate a survival benefit in patients who had received one or two prior chemotherapy regimens [102]. Likewise, in the SWOG0023 trial for patients with Stage III NSCLC, gefitinib monotherapy did not improve survival after completion of induction and consolidation chemo-radiotherapy [103]. In the Iressa Dose Evaluation in Advanced Lung Cancer (IDEAL) trial gefitinib did not show single-agent activity in patients with NSCLC [104, 105]. Furthermore, gefitinib did not enhance the activity of standard chemotherapy in patients with advanced NSCLC in the Iressa NSCLC Trial Assessing Combination Treatment (INTACT) studies [106-108]. Because of the disappointing results in these trials, the use of gefitinib is currently restricted by the FDA to patients who had previously shown benefit from the drug.

II. Erlotinib and non-small cell lung cancer

In contrast to the studies with gefitinib, one large phase III randomized trial, the National Cancer Institute of Canada BR.21 study, has confirmed a survival benefit to erlotinib in NSCLC [109]. Patients with stage IIIB or IV disease who had failed one or two prior chemotherapy regimens were randomized to erlotinib or placebo. Progression-free survival in the two groups was 2.2 months and 1.8 months, respectively (P < 0.001), and overall survival was 6.7 months and 4.7 months, respectively (P < 0.001).

However, two other large clinical trials have failed to show a benefit to erlotinib in combination with standard chemotherapy. In the TALENT trial, patients with untreated advanced NSCLC were randomized to receive either erlotinib or placebo along with gemcitabine/cisplatin [110]. No statistically significant difference in overall survival or time to progression was observed between the two groups. In the TRIBUTE trial, patients with untreated stage IIIB or IV NSCLC were randomly assigned to erlotinib or placebo combined with up to 6 cycles of carboplatin and paclitaxel, followed by maintenance monotherapy. No difference was observed in objective response, time to progression, or overall survival. However, patients without a history of smoking demonstrated improved survival with erolitnib [111].

III. EGFR inhibitors and cancers of the gastrointestinal tract

Cetuximab was found to have clinically significant activity when given alone or in combination with irinotecan in patients with irinotecan-refractory colorectal cancer [112]. An international multi-center phase III trial compared panitumumab as a single agent to best supportive care in patients with metastatic colorectal cancer who had failed conventional chemotherapy. The results showed that the panitumumab arm significantly improved the progression-free survival over the control arm but did not affect overall survival [113]. A phase III trial in patients with advanced pancreatic carcinoma showed that the addition of erlotinib to gemcitabine improved overall survival and progression-free survival compared with gemcitabine alone [114].

In many of these trials, it has been observed that patients who develop the characteristic acneiform rash with treatment of EGFR inhibitors tended to have a higher response rates than those who did not develop a rash. In the phase III panitumumab trial for metastatic colorectal cancer, rash was reported in 90% of patients, and increased severity significantly correlated with improved median overall survival [113].

CONCLUSIONS

There is a wealth of preclinical data indicating why EGFR inhibitors should be effective in controlling cancers. These agents could block cancer growth via inhibition of tumor proliferation, apoptosis, angiogenesis, migration, invasion, and metastases. However, clinical trials in patients with cancer have shown only modest gains with these inhibitors, particularly when they have been given as monotherapy. There appears to be greater promise when these agents are combined with radiation or chemotherapy. A variety of mechanism may explain the potentiation of conventional cytotoxic therapy with EGFR inhibitors, including the suppression of DNA damage repair. A more thorough understanding of these mechanisms is required in order to maximize the benefit from EGFR inhibitors in the treatment of cancer.

ABBREVIATIONS

NSCLC

non-small cell lung cancer

SCCHN

squamous cell carcinoma of the head and neck

MoAb

monoclonal antibody