EGFR Oncogene Addiction and
Resistance in Lung Cancer

Anthony C. Faber, PhD, and Jeffrey A. Engelman, MD, PhD
Massachusetts General Hospital Cancer Center, Boston, Massachusetts

The epidermal growth factor receptor (EGFR) has proven to be an effective therapeutic target in non–small cell lung cancer (NSCLC). Although many cancer types express abundant amounts of EGFR, NSCLCs are unique in that they sometimes demonstrate exquisite sensitivity to single-agent EGFR tyrosine kinase inhibitors (TKIs). Although most NSCLCs in the United States and Europe do not display such sensitivity, a rare subset does, and in East Asia a substantially larger proportion of cases exhibit sensitivity. The reason for this uncommon but sometimes extraordinary sensitivity was crystallized in 2004 with the discovery of activating EGFR mutations.^1-3 Indeed, cancers with two of the most common activating EGFR mutations, an in-frame deletion in exon 19 and an amino acid substitution in exon 21 (L858R), are highly enriched in cancers with enhanced response to EGFR inhibitors.^{7, 8, 13-19} Research over the past 4 years has identified two major reasons for this enhanced sensitivity.^20,21 First, these cancers are "addicted" to EGFR signaling, and inhibition of EGFR signaling leads to apoptosis (see below). Secondly, the EGFR kinase inhibitors, gefitinib and erlotinib, bind with higher affinity to the mutant receptors as compared to the wild-type receptor.²² This results in more complete inhibition of the mutant EGFR with the standard dosing regimens.

For our purposes here, "oncogene addiction" is defined as a condition in which disruption of one gene or protein leads to the death of the cancer cell. NSCLCs with EGFR mutations fit this definition well since these cancers undergo substantial apop tosis and exhibit tremendous clinical response to single-agent EGFR-TKIs. But what underlies and defines this addiction? Paradoxically, part of the answer comes from understanding how cancers become resistant to EGFR-TKIs. By determining how a cancer becomes resistant and avoids the deleterious effects of EGFR inhibition, the basis of their initial sensitivity is revealed.

Based on available data, we propose the "signaling hypothesis" of EGFR oncogene addiction and resistance. Cancers that are sensitive to EGFR inhibitors are unique in that EGFR is the sole regulator of key downstream signaling pathways, particularlythe phosphoinositide 3-kinase (PI3K)/AKT and the MEK/ERK pathways.²³ Thus, when EGFR is inhibited in EGFR mutant cancers, these pathways are turned off and the cells undergo apoptosis. In contrast to tumors that are highly sensitive to EGFR inhibitors, most other cancers do not have EGFR as the sole regulator of the PI3K/AKT and MEK/ERK pathways, and thus are resistant to the effects of EGFR inhibition.^24-25 This relationship between oncogene addiction and the master regulator of these down stream signaling pathways is not unique to EGFR addiction. This concept also applies to cancers sensitive to other receptor TKIs such as HER2-amplified cancers,²⁶ MET-amplified cancers,¹⁴ and those with EML4-ALK translocations.²⁷ Does this mean that EGFR has no important function in these cancers other than sole regulation of PI3K/AKT and MEK/ERK? Probably not, but these observations are among the most consistent and invariable. It is tempting to speculate that cancers that are the most sensitive to EGFR-TKIs are those in which there are absolutely no other regulators of these downstream pathways to compensate for loss of EGFR activity.

Although these therapies can be effective for some time, cancers invariably become resistant to EGFR-TKIs. Clinical studies have prospectively identified and treated over 200 patients with EGFR mutations. Together, they demonstrate a radiographic response rate ranging from 55% to 82% and median time to progression of 9.4 to 13.3 months.^28-33 These tumors become resistant when they reactivate downstream signaling despite the presence of the EGFR inhibitor.³⁴ The most commonly identified mechanism is an EGFR mutation at position 790 (T790M), resulting in substitution of a threonine residue with methionine, which abrogates the ability of gefitinib or erlotinib to inhibit EGFR. Other less common EGFR resistance mutations have been identified, and may be selected for in sanctuary sites such as the central nervous system.³⁵

Recent data indicate that cancers also can become resistant by activating downstream signaling via alternative mechanisms. For example, MET amplification leads to persistent activation of the PI3K/AKT and MEK/ERK signaling pathways. In these cells there appears to be dual input into down stream signaling by MET and EGFR.^26,36 Thus, combined inhibition of both MET and EGFR is necessary to kill such tumor cells. Another mechanism of resistance involves activation of the IGF-1R pathway.³⁷ Unlike MET amplification, such tumors exhibit reactivation of the PI3K pathway. In gefitinib-resistant A431 squamous cancer cells, combined EGFR and IGF-1R inhibition were required to downregulate PI3K and ERK pathways and block cell growth. This underscores the importance of PI3K down regulation for receptor TKIs to promote cell death, a concept supported by numerous studies.^{21,24,26,38,39} In fact, a recent study showed that women with HER2-amplified breast cancers that had either mutational activation of PI3K (i.e., PIK3CA mutations) or inactivation of PTEN had a worse prognosis when treated with trastuzumab.³⁸

How can we use this information to better utilize EGFR inhibitors for treatment of lung cancer? First, there are likely some tumors in which EGFR regulates either PI3K/AKT or MEK/ERK pathways, but not both. Patients with such cancers may have stable disease or progress slowly on EGFR-TKIs. However, inhibition of both pathways might result in apoptosis and substantial tumor shrinkage. Thus, combining EGFR inhibitors with either MEK or PI3K inhibitors may increase the proportion of EGFR wild-type patients that respond. In addition, a combination approach may prevent the emergence of resistance that eventually occurs following initial response. Unfortunately, it appears that multiple mechanisms of resistance may occur simultaneously in the same patient.^26,36 Thus, treating one resistance mechanism (e.g., T790M with irreversible EGFR inhibitors) may have unacceptably poor clinical activity. Instead of waiting for resistance to occur, we can change the initial therapy to a combination approach that prevents the development of resistance. For example, initial therapy with an irreversible EGFR inhibitor and a MET inhibitor may block both T790M and MET amplification-induced resistance, and can prolong time to progression for patients. In preclinical studies, A431 tumor cells, which become resistant to gefitinib due to activation of the IGF 1R pathway, were injected into mice to develop xenograft tumors. These tumors became resistant to either single-agent gefitinib or an IGF1R antibody.³⁷ However, when gefitinib and the IGF1R antibody were combined, no resistant tumors were observed, even after discontinuation of therapy.

Thus, upfront therapy with combinations that overcome the potential resistance mechanism may lead to longer time to progression. Clinical trials assessing these strategies are underway.

Thus far, there has not been substantial evidence to suggest that adding an EGFR-TKI to chemotherapy is beneficial. Yet recent clinical studies suggest that inhibition of EGFR with the mono clonal antibody cetuximab may increase the survival of patients with advanced NSCLC.⁴⁰ Interestingly, cetuximab is relatively ineffective in NSCLC when used as monotherapy, and previous preclinical and clinical studies suggest that cetuximab is not as potent as EGFR-TKIs in cancers with L858R or exon 19 deletion EGFR mutations.⁴¹ However, studies in colorectal cancer suggest that cetuximab is most effective when the wild- type EGFR is activated by ligand.⁴² It is therefore interesting to speculate that some cancers may respond to the insults of cytotoxic chemotherapy by inducing autocrine activation of the EGFR that helps the cancer cell survive. Thus, chemotherapy may result in EGFR oncogene "addiction" in some tumors. These findings highlight that the mechanism of EGFR activation, such as mutation, gene amplification, or ligand-dependent activation, may impact which type of strategy (antibody or TKI) should be used to target the EGFR.

Further improvement of EGFR-targeted therapies will necessitate a more thorough understanding of both the critical downstream signaling events controlled by EGFR, as well as compensatory mechanisms cancers utilize to become resistant. When these factors are considered as a whole, treatment efficacy should improve, and rational combination-based targeted therapies may yield sustained responses.

< < PREV

NEXT > >