Epidermal growth factor receptor inhibition in lung cancer: the evolving role of individualized therapy

1 Introduction

Lung cancer is the leading cause of cancer-related deaths worldwide, accounting for more than 1 million deaths each year [1]. In the USA, lung cancer accounts for approximately 28% of all cancer-related deaths [2]. Non-small cell lung cancer (NSCLC) is the most common type and accounts for at least 85% of all lung cancer cases [2, 3]. Treatment options for NSCLC depend on stage of disease and include surgery, radiation, platinum-based doublet chemotherapy, and targeted therapies in some cases [3]. Most patients present with advanced or metastatic disease, for which chemotherapy is generally recommended as first-line treatment [3]; however, efficacy is modest and therapy is associated with significant toxicity [4, 5]. Adding bevacizumab, an anti-angiogenic agent, to standard first-line doublet chemotherapy regimens has been shown to increase efficacy, but with only minimal improvements in clinical outcomes [6, 7].

Epidermal growth factor receptor (EGFR) inhibitors have been investigated as first-line or subsequent therapy options for patients with NSCLC. Recent efforts are focused on identifying specific molecular markers that may predict treatment response, thus allowing for a more tailored approach for treating patients with NSCLC. This article will provide an overview of current understanding of the implications of EGFR signaling in the treatment of NSCLC and highlight strategies for its inhibition, with a focus on EGFR tyrosine kinase inhibitors (TKIs).

2 The EGFR signaling pathway

The EGFR family of receptor tyrosine kinases, also known as the HER or ErbB family, has four known members: EGFR or HER1/ErbB1, HER2/ErbB2, HER3/ErbB3, and HER4/ErbB4 [8]. Downstream signaling regulated by this family of receptor tyrosine kinases (RTKs) is complex and multidimensional, and aberrant activation of the pathway leads to downstream events that stimulate five of the six hallmarks of cancer, including evasion of apoptosis, self-sufficient growth, insensitivity to anti-growth signals, sustained angiogenesis, and tissue invasion and metastasis (the remaining hallmark is limitless replicative potential, which is not promoted by EGFR activation) [9] (Fig. 1). EGFR is overexpressed in many epithelial cancers, including NSCLC [10]; small cell lung cancer is one of the few solid tumors in which EGFR is not overexpressed [11]. Thus, EGFR has been the most intensively studied of the four family members, and has become a prototype of classical RTKs. However, deregulation of the pathway may occur at several nodal points, providing a multitude of targets for selection of individualized therapy [12]. Investigation of signaling pathways downstream of EGFR has demonstrated the far-reaching effect of this pathway on diverse cellular processes, such as proliferation, angiogenesis, and development [8].

Regulation of the EGFR pathway is complex, and a comprehensive review is beyond the scope of this article. Like other HER family members, EGFR is a transmembrane receptor activated in response to ligand (EGF and others) binding to the extracellular domain [13]. Ligand binding induces conformational changes that allow for the formation of receptor dimers. Both homodimer and heterodimer formation within the EGFR family have been verified, and the variety of pairing combinations is thought to provide an additional layer of signaling regulation. Activation of the kinase domain of the receptor leads to autophosphorylation and activation and the subsequent recruitment of adaptor proteins that mediate downstream signaling [13]. The EGFR pathway is also regulated on a higher level by several feedback loops. For instance, activation leads to increased cellular production of ligand and increased receptor internalization [13].

EGFR activates two major downstream intracellular signaling pathways—the Ras-Raf-mitogen-activated protein kinase kinase (MEK)-mitogen-activated protein kinase (MAPK) and the phosphoinositide 3-kinase (PI3K)-Akt/protein kinase B-mammalian target of rapamycin (mTOR) cascades [14–16]. The Ras-Raf-MEK-MAPK pathway modulates several cellular processes including gene transcription, G1/S cell-cycle progression, and cellular proliferation. EGFR tyrosine kinase activity leads to activation of the small GTPase Ras, which then exchanges GDP for GTP; activated, GTP-bound Ras then stimulates the Raf-MEK-MAPK cascade [17]. The PI3K pathway regulates anti-apoptotic and prosurvival signal cascades [17]. These pathways may also be modulated by other proteins such as c-mesenchymal-epithelial transition factor (MET), insulin-like growth factor 1 receptor (IGF-1R), LKB1-amp-activated protein kinase [14], and the echinoderm microtubule-associated protein-like 4/anaplastic lymphoma kinase (EML4-ALK) fusion protein [18]. While EML4-ALK is detected in less than 10% of lung cancers, it is most common in adenocarcinomas and in never or light smokers [14, 18]. It is also almost never detected along with activating EGFR or V-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations [18], suggesting that EML4-ALK is involved in this pathway. As a result, ALK fusion proteins are being investigated as potential therapeutic targets for NSCLC treatment [18].

Other downstream targets of EGFR include the phospholipase C-protein kinase C (PKC) and Janus kinase/signal transducers and activators of transcription (STAT) pathways [16]. Phospholipase C enzymatically cleaves phosphatidyl inositol 4,5,-bisphosphate, which leads to release of cellular calcium stores and activation of PKC [19]. PKC in turn activates the Raf-MEK-MAPK pathway and other effector proteins [17]. STATs, which stimulate transcription of nuclear factors that promote cell survival and oncogenesis [17], are activated by EGFR signaling both directly by interaction with EGFR [20] and indirectly through Src family kinases [21]. Many protein families (e.g., Ras, Raf) contain multiple members, adding to the complexity and scope of activation of these pathways through EGFR. As known cellular processes influenced by EGFR signaling have continued to grow, the effects of the complete biochemical network associated with EGFR are not entirely known.

3 Implications of EGFR expression and activity in NSCLC

Owing to the variety of cellular processes regulated by EGFR signaling, its deregulation has been associated with carcinogenesis [22]. Aberrant activation of the EGFR pathway is thought to be due to at least three mechanisms: (1) enhanced production of ligands by cancer cells, (2) increased expression of EGFR on the cancer cell membrane, and (3) activating mutations of the EGFR gene or other family members [12, 15]. Several strategies for inhibition of EGFR, including tyrosine kinase inhibition, have been developed for treatment of human cancers, including lung cancer. EGFR is frequently overexpressed in NSCLC, and EGFR overexpression has been associated with poor prognosis [23, 24]. Total EGFR protein is detectable in approximately 80–85% of patients with NSCLC, though levels of expression vary widely on a continual scale [3]. Efforts toward implementing routine molecular profiling of tumors have been underway, with the hope that relevant correlates may predict patient response to EGFR blockade.

EGFR protein levels can be measured using several methods, including radioactive-labeled ligand binding, competitive immunoassay, western blotting, and immunohistochemistry (IHC) [25]. With the exception of IHC, however, these methods require complex laboratory equipment and are not easily modified for clinical use [25]. Moreover, analysis of EGFR levels by IHC is affected by many variables that decrease its reproducibility and quantitative value [26], and its use has been inconsistent in predicting response to EGFR TKIs [27–30]. Currently, IHC is not yet optimized for determining patient eligibility to receive EGFR TKI therapy, though standardization of this methodology may allow for clinical use in the future. As the phosphorylated form of the protein is the active form, measurements of phospho-EGFR may provide more clinically relevant information; however, the half-life of the phosphorylated form is short, and unless a specimen is optimally collected and processed, phospho-EGFR measurements may result in misleading findings [31].

Increased expression may be due to increased gene copy number via amplification or polysomy. Thus, increased copy number of EGFR in tumors may be meaningful for predicting response to EGFR TKIs. Although polysomy and amplification are often regarded as being of similar importance, chromosome 7 is the site not only of the EGFR gene but also of BRAF and MET. Therefore, polysomy may activate genes that have opposite effects on the EGFR pathway, resulting in complex interactions. Several methods are available to measure EGFR gene copy number, including fluorescence in situ hybridization (FISH), chromogenic in situ hybridization (CISH), and real-time polymerase chain reaction (PCR). FISH is widely used to detect copy number of specific genes in tissue sections [32], and EGFR gene amplification detected using this method has been shown to correlate with response to EGFR-targeted agents [27, 33, 34]. Gene number determined by FISH appears to be a more robust predictive marker of EGFR-TKI response than EGFR protein expression measured by IHC, but more data are needed regarding reproducibility and predictive power across laboratories [26]. In addition, FISH (and CISH) can distinguish between amplification and polysomy. There is also the potential that combined readouts from IHC and FISH may predict response to EGFR TKIs [35]. CISH, an alternative to FISH for measuring EGFR copy number, appears to be an accurate and reproducible method [26], but requires further evaluation and validation in clinical trial samples [25]. Finally, real-time PCR is a potential tool for detecting EGFR copy number [26]; however, measurements do not necessarily correlate with FISH, IHC, or mRNA expression [36]. Another method to detect increased copy number, in particular amplification, is comparative genomic hybridization (CGH), although it is not suitable for routine clinical use. An in vitro study using NSCLC cell lines demonstrated that FISH, quantitative polymerase chain reaction, and CGH gave comparable results [37]. Additionally, studies to date have not consistently shown that EGFR copy number is predictive of response to treatment, disease control, or progression-free or overall survival [36, 38–40].

There is growing evidence that EGFR mutations may be predictive of therapeutic efficacy [3]. The evidence that mutations predict for progression-free survival is much greater than for overall survival [12]. These findings may, in part, be related to the fact that TKI therapy is often a second-or third-line option. The most common EGFR mutations are deletions in exon 19 or a single point mutation in exon 21, both of which cause activation of the tyrosine kinase domain [41]. Presence of these activating mutations is predictive of treatment benefit from EGFR TKI therapy [26, 42–52]; however, the de novo existence or acquisition of some EGFR mutations (e.g., T790M) are associated with EGFR TKI resistance [53]. There are many techniques that may potentially be used for EGFR mutation analysis, the majority of which are PCR based [25, 26]. Mutations can be detected using a PCR assay and then confirmed by DNA sequencing [54]. Some large clinical screens for EGFR mutations have been performed [52]; however, more streamlined approaches are in development. For instance, it was recently demonstrated that detection of shed tumor DNA using the DxS EGFR mutation test kit [55] from the plasma of patients is sufficient for determination of EGFR mutation status; EGFR mutation status was also associated with patient outcome in this study [56]. Other potential indicators, such as KRAS mutations, EGFR truncations, expression levels of MET and HER2, and Akt phosphorylation state are also being investigated as predictors of response to EGFR-directed therapy [57].

Though various methodologies are available to assess potential molecular markers predictive of response to anti-EGFR therapy [26], more advancements will be necessary before these may provide widespread benefit to patients. As agents are developed that target downstream mediators of EGFR signaling, other mutational and expression assays will likely be evaluated. Ongoing randomized studies will continue to validate the assays that can predict patient outcome. It is likely that as molecular characteristics more routinely dictate treatment decisions, pathologists will begin playing a larger role in choosing the optimal therapy for individual patients [58]; testing of new biopsies when NSCLC patients relapse or begin a new treatment regimen will also be of importance. It is hoped that in the near future, extensive testing of patient tumors will become the standard of care for making treatment decisions. The ability to identify appropriate biomarkers to predict clinical efficacy would render clinicians one step closer to the provision of personalized medicine for patients with NSCLC.

4 Drugs targeting the EGFR pathway

Two classes of EGFR inhibitors, monoclonal antibodies (e.g., cetuximab) and small-molecule TKIs (e.g., erlotinib, gefitinib), have been studied in phase III trials and are currently in clinical use in NSCLC [15, 59–63]. Monoclonal antibodies targeting EGFR bind to the extracellular domain of EGFR and block ligand binding and receptor activation, while small-molecule EGFR TKIs compete reversibly with ATP to bind to the catalytic domain of the intracellular kinase domain to inhibit its activity.

Results from the phase III FLEX trial (N=1,125) showed that cetuximab in combination with chemotherapy (vinorelbine/cisplatin) improved overall survival compared with chemotherapy alone (11.3 vs. 10.1 months; P=0.04) in patients (performance status 0–2) with previously untreated advanced NSCLC that expressed EGFR [63]. Cetuximab is not approved by the US Food and Drug Administration (FDA) for NSCLC, but is recommended by the National Comprehensive Cancer Network (NCCN) in combination with vinorelbine/cisplatin for patients with advanced disease whose tumors express EGFR [3]. The Committee for Medicinal Products for Human Use, the scientific committee of the European Medicines Agency, has adopted a negative opinion for the use of cetuximab for such patients [64].

Early clinical data showed that approximately 10% of unselected patients with NSCLC respond to gefitinib or erlotinib, possibly reflective of the fact that 10–15% of patients have activating EGFR mutations [30]. In a phase III study of patients (N=1,692) with refractory, advanced NSCLC [65], gefitinib did not show an overall survival benefit compared with placebo. As a result, gefitinib currently has a limited indication in the USA for the continued treatment of patients benefiting from gefitinib therapy. Efforts to determine the response to gefitinib in selected patient populations are ongoing. In contrast, erlotinib has shown significantly longer progression-free survival (2.2 vs. 1.8 months; hazard ratio [HR], 0.61; P<0.001) and overall survival (6.7 vs. 4.7 months; HR, 0.70; P<0.001) compared with placebo in patients with advanced NSCLC who had received prior chemotherapy [66]. Erlotinib is indicated by the US FDA for treatment of locally advanced or metastatic NSCLC that has progressed after at least one line of chemotherapy [67]. Erlotinib is recommended by the NCCN as second- and third-line therapy for NSCLC; it is also recommended for first-line treatment in patients with EGFR mutations, but this is supported by a low level of evidence [3]. The potential for primary resistance to currently available EGFR TKIs is an important consideration for treatment, but even in patients who initially respond, relapse often eventually occurs.

5 Drugs of the future: targeting multiple points in the EGFR pathway

Many strategies to inhibit EGFR signaling exist in addition to direct targeting of the receptor. Points within the pathway that are currently being investigated for intervention in NSCLC are illustrated in Fig. 3, as well as some key agents corresponding to their target. Other agents in early development that target EGFR and other pathways are listed in Table 1.

6 Conclusions

Targeting the EGFR pathway has demonstrated clinical benefit for a select group of patients with NSCLC. However, the multitude of actual or potential targets for individualized therapy offers the prospect that most, if not all, patients will eventually benefit from carefully selected therapy targeted toward the EGFR pathway or another critical signaling pathway. Whether such therapy targets a single pathway or gene, a combination of targeted therapies, or a combination of targeted and conventional therapies remains to be determined. Such an individualized approach requires acquisition and testing of tumor tissue shortly before the selection of therapy. Tumor heterogeneity, genomic instability, and the development of resistance make it mandatory that tumor tissue be obtained for drug selection shortly before individualized therapy. Thus, other than for first-line therapy, a further biopsy or other form of intervention may be necessary. Individualized therapy also requires “reflex” testing of tissues—e.g., certain clinicopathological features trigger automatic or “reflex” testing. For instance, a diagnosis of adenocarcinoma of the lung combined with one or more other characteristics of EGFR mutant tumors (female gender, never smoking status, or East Asian ethnicity) would trigger “reflex” testing for mutations or other forms of deregulation of the EGFR pathway, with therapy decisions based on the results obtained. Such an approach is already practiced at certain leading medical centers including M.D. Anderson Cancer Center, Houston, TX, USA, and Memorial Sloan-Kettering Cancer Center, New York, NY, USA. When patients are informed that additional procedures are required to obtain tissue to guide personalized therapy, the vast majority are willing to participate in such trials. The advent of rapid, relatively cheap next-generation sequencing technologies for all or large parts of the genome will accelerate the identification of individualized targets.

Ideally, EGFR TKI therapy could be tailored to patients who are most likely to experience benefit, allowing other patients to receive different therapies appropriate for their tumor profile. Several potential molecular indicators have been identified, and optimization of assay technologies is in progress, but these are not yet widely used to direct treatment choices in NSCLC patients. As the transition to personalized medicine occurs, pathologists will likely play a greater role in diagnostic decisions, and obtaining sufficient sample for testing upon initial biopsy will be critical. Moreover, it may be necessary to obtain subsequent biopsies when patients relapse or begin a new treatment regimen. In an effort to potentially overcome and prevent resistance to current EGFR-targeted agents, ongoing trials are evaluating new agents, such as EGFR/HER2 irreversible inhibitors and EGFR/VEGFR inhibitors. Hopefully, these new strategies for inhibiting multiple EGFR targets and/or multiple tumorigenic processes may eventually improve patient outcomes.

7 Key unanswered questions

The major unanswered questions are as follows:

• What are the best targets for therapeutic interventions in the EGFR signaling pathway?

• How do we identify the best targets in individual tumors? How do we get sufficient tissue for testing?

• Can the tests be performed in a clinically relevant time frame at an acceptable cost?

• What are the best drugs directed against these targets?

• How do we combine targeted therapies with each other and with conventional therapies?

• Are targeted therapies suitable for first-line therapies?

• How do we overcome drug resistance?