EGFR-Targeted Therapies in Colorectal Cancer
The management of colorectal cancer relies heavily on the combination of the pyrimidine analog antimetabolite 5- fluorouracil with the platinum-based drug oxaliplatin or the topoisomerase inhibitor irinotecan. Optimization of dosing and scheduling of these agents to improve response and survival continues to evolve. Meanwhile, the rational targeting of molecular signaling pathways that are involved in the etiology of malignancies is currently one of the most promising strategies in novel anticancer drug development. New classes of drugs that target the epidermal growth factor receptor are among the most clinically advanced molecular-targeted therapies and have shown efficacy in colorectal cancer. The current status of epidermal growth factor receptor-targeted therapeutic agents is reviewed, with emphasis on their role in the management of colorectal cancer.
[Key words: Epithelial growth factor receptor; Colorectal cancer; Targeted therapy; Combination drug therapy; Cetuximab; Erlotinib; Panitumumab; Gefitinib]
Treatment with 5-fluorouracil (5-FU)-based regi- mens has been central to the management of patients with colorectal cancer (CRC) for decades. Introduced in the 1950s, 5-FU is a prototypical example of rational drug design. The development of 5-FU was based on the discovery that rat livers containing hepatomas use uracil for ribonucleic acid (RNA) synthesis to a greater extent than do normal rat livers.1 Because fluorine substitution was believed to increase the toxicity of organic molecules, the effect of adding fluorine to uracil was investigated. Nearly 50 years later, 5-FU in combination with oxaliplatin or irinotecan remains the standard front- line treatment for advanced CRC.
The revolution in genetics and molecular biology during the last century has afforded substantial gains in our knowledge of the molecular basis of tumori- genesis. The current theory of cancer is that it is a genetic disease largely derived from the sequential accumulation of mutations.2 Dysregulation of dis- crete signaling pathways involved in processes, such as growth, apoptosis, differentiation, and vasculari- zation, is currently viewed as central to the develop- ment and progression of disease. Agents targeting biologic pathways uniquely disregulated in tumor cells, such as growth factor pathways in which components may be overexpressed or overactive in tumor cells relative to normal cells, present the tantalizing possibility of lower toxicities than ob- served with cytotoxic agents, which target both neoplastic and normal cells.
Rational targeting of receptors, such as the epider- mal growth factor receptor (EGFR) and various others, is a major strategy behind the design of novel anticancer agents. At present, two types of approaches targeting at the EGFR are undergoing clinical trials: anti-EGFR monoclonal antibodies and small-molecule EGFR tyrosine kinase inhibitors (TKIs). We review the current status of therapeutic agents targeting the EGFR, with an emphasis on the management of CRC.
METHODS
Epidermal Growth Factor Receptor
The EGFR (also known as ErbB1 or human epithelial growth factor, HER1) is a member of the ErbB family of receptors expressed on normal epithelial, mesenchymal, and neuronal cells.3 Since the initial discovery of EGFR in the 1980s, the ErbB receptor family has come to encompass four distinct receptors: ErbB-1 (EGFR), ErbB-2 (HER-2), ErbB-3, and ErbB-4. Structurally, they are receptor tyrosine kinases that consist of an extracellular ligand-binding domain, a hydrophobic transmembrane region, and an intracellular kinase domain (Fig. 1).
The primary ligands to EGFR include mitogens, such as epidermal growth factor (EGF) and trans- forming growth factor alpha (TGF-a). Ligand binding induces homodimerization or heterodimerization of EGFR-family receptors, leading to receptor activation via transphosphorylation of intracellular tyrosine residues. In addition to direct ligand-binding mediated receptor activation, EGFR-family signaling pathways can be activated via interactions with other signaling pathways, such as those relating to hormones, neuro- transmitters, lymphokines, and stress responses (^horizontal^ receptor activation).5 The activated EGFR transduces signals via a complex, multilayered signaling network (Fig. 2), including mitogen-activated protein kinase (MAPK), which is involved in cell proliferation, phosphotidylinositol-3 kinase (PI3K), which is involved in the cell cycle and survival, and the signal transducer and activator of transcription (STAT) family of proteins, which mediate cell division, survival, motility, invasion, and adhesion.
EGFR and Cancer
Aberrations in EGFR signaling are implicated in various processes associated with carcinogenesis.6 These include proliferation, angiogenesis, invasion/ metastasis, antiapoptosis, and resistance to chemo- therapy or radiation therapy.In CRC, approximately 60 to 80 percent of tumors have been shown by immunohistochemistry (IHC) to express EGFR, although all CRC tumors likely express the receptor to some degree given that all epithelial cells constitutively express EGFR. Numerous retrospec- tive studies have evaluated the relationship between EGFR expression and prognosis. Early studies sug- gested correlations with grade, stage, and survival. In contrast, a recent large-scale study examining 249 colorectal adenocarcinomas found no association between EGFR expression and stage or survival.In addition, EGFR expression detectable by IHC has not proven to be a consistent predictor of response to EGFR-targeted therapy.8 Inaccuracies linked to IHC testing resulting from potential sample degradation or epitope loss during fixation have been proposed as a likely explanation for the lack of a consistent correla- tion. Based on these findings, the National Compre- hensive Cancer Network (NCCN) considers the evaluation of EGFR expression by IHC to not provide predictive value with which to determine patient eligibility for anti-EGFR monoclonal antibody therapy.9 In contrast to IHC-based detection of EGFR ex- pression, fluorescence in situ hybridization (FISH)- based detection of EGFR gene copy number may have a true predictive value in identifying patients likely to respond to EGFR-targeted therapies. A recent study examining EGFR gene copy number in 31 patients with metastatic CRC (mCRC) found a highly statistically significant correlation (P < 0.0001) between increased gene copy number and response to cetuximab or panitumumab treatment.10 An association between the development (or severity) of acneiform rash and increased survival has been noted in clinical trials of agents targeting EGFR pathways, suggesting that the development of rash could be used as a surrogate marker for efficacy of EGFR-targeted therapy. EGFR-Directed Monoclonal Antibodies for Colorectal Cancer At present, the Food and Drug Administration (FDA) has approved two anti-EGFR monoclonal antibodies: cetuximab and panitumumab. In addition numerous antibodies and TKIs targeting EGFR are in clinical development or under preclinical investigation (Tables 1 and 2).Cetuximab. Cetuximab (Erbitux\; ImClone Systems, Princeton, NJ) was approved in February 2004 for the treatment of patients with EGFR-detectable mCRC, as monotherapy in irinotecan (CPT-11)-intolerant patients or in combination with irinotecan in patients with tumors refractory to irinotecan-based therapy. This agent is among the few treatment options currently available for patients who are refractory to both irinotecan-based and oxaliplatin-based regimens. Cetuximab is a chimeric human-murine monoclo- nal antibody of the immunoglobulin G (IgG1) isotype, initially chosen for clinical development from a series of murine monoclonal antibodies that demonstrated activity against EGFR-bearing cells in vitro (reviewed in Mendelsohn and Baselga33). Cetuximab, which binds to the EGFR with a 2-log higher affinity than the cognate EGFR ligands TGF-a and EGF, induces EGFR internalization and degrada- tion without inducing receptor phosphorylation. As a result of degradation, receptor recycling is prevented, and the density of EGFR receptors on the cell surface is reduced.Cetuximab inhibits the growth of target cells via multiple mechanisms, including cell-cycle arrest in the G1 phase, potentiation of apoptosis, and inhibition of tumor angiogenesis. The antiangiogenic effect of cetuximab is related to inhibition of TGF-a, which has been shown to promote expression of VEGF.34 The IgG1 domain of the cetuximab antibody is capable of engaging the host immune response through antibody-dependent cell-mediated cytotoxic- ity (ADCC) and complement fixation mechanisms.35 Clinical Development of Cetuximab. In Phase I clinical studies, the most frequently occurring ad- verse events were acneiform rash as well as fever and chills, asthenia, transaminase elevation, and nausea. Immunogenicity was low: < 4 percent of patients developed antichimeric antibodies, and a small number of patients (4 percent) showed Grade 3 anaphylactoid or Grade 4 anaphylactic infusion reactions.36 Postmarketing surveillance in the United States has documented infusion reactions in ap- proximately 1.5 percent of patients (ImClone Systems, Inc., data on file). Cetuximab has recently been associated with severe hypomagnesemia, with Grades 3 and 4 hypomagnesemia occurring in 23 percent of pateints.37 Because early data suggested that biologic agents administered alone have mostly cytostatic effects, cetuximab was first tested in combination with chemotherapy, based on observations suggesting that cetuximab could reverse irinotecan resistance in patients for whom irinotecan-containing regimens have failed. Patients with EGFR-detectable CRC refractory to both 5-FU and irinotecan (n = 121; median age, 56 years; median Karnofsky score (KPS), 90) were enrolled in a Phase II trial of cetuximab in combination with irinotecan.38 Patients were treated with cetuximab (400 mg/m2 initial dose followed by 250 mg/m2 weekly) plus irinotecan at the same dose and schedule as during progression. Toxicities attributable to cetuximab included acneiform rash (53 percent Grades 1 and 2,8 percent Grade 3) and infusion reactions (1–2 percent Grades 3 and 4). The rash resolved on discontinuation of therapy, and in subsequent studies the rash was correlated with tumor response and survival. Twenty-one patients (17 percent) achieved a partial response for a median duration of 84 days. Thirty-seven additional patients (31 percent) had stable disease or minor responses. To determine the single-agent efficacy of cetux- imab and confirm the response rate of cetuximab plus irinotecan, a second Phase II trial was under- taken in patients with EGFR-detectable, irinotecan- refractory mCRC.8 Patients were randomized in a 2:1 ratio, with patients in Arm A receiving cetuximab plus irinotecan at the same dose and schedule on which they had been progressing. Patients in Arm B received cetuximab alone with the option to switch to cetuximab plus irinotecan after failure of cetux- imab as a single agent. A total of 218 patients were enrolled in Arm A (143 males; median age, 71 years; 88 percent, KPS 80–100) and 111 in arm B (65 males; median age, 70 years; 85 percent KPS, 80–100). The response rate of patients in Arm A was 22.9 percent; median TTP was 126 days. The response rate in patients in Arm B was 10.8 percent; median TTP was 45 days. Additional Phase II studies combining cetuximab with both 5-FU and irinotecan have demonstrated response rates in the first-line meta- static setting of 44 percent for cetuximab/IFL (5- fluorouracil, leucovorin, bolus irinotecan) and 46 percent for cetuximab/FOLFIRI (5-fluorouracil, leu- covorin, continuous irinotecan).14,39 Two recent studies have evaluated cetuximab monotherapy in heavily pretreated patients. A large expanded access study involving 743 patients previ- ously treated with irinotecan, oxaliplatin, and fluo- ropyrimidine demonstrated a median treatment duration of 11.4 weeks, with 11.7 percent of patients remaining on treatment longer than six months.40 A similar open-label Phase II study evaluated the efficacy of cetuximab in 346 patients who had received a median of four regimens of previous therapy (range, 2–9). Twelve percent of patients achieved a partial response with a median survival of 6.6 months.11 The addition of bevacizumab to cetuximab with or without irinotecan has demonstrated promising ac- tivity. The interim results from the randomized Phase II trial (BOND-2 [Bowel Oncology with Cetuximab Antibody study 2]) of cetuximab, bevacizumab, and irinotecan (Arm A) compared with cetuximab and bevacizumab (Arm B) in irinotecan-refractory mCRC patients (n = 75; median, ECOG PS: 1; median age, 60 years) were reported recently. A total of 75 patients were entered; partial response rate was 35 percent in Arm A (n = 40) compared with 23 percent in Arm B (n = 35).12 A Cancer and Leukemia Group B (CALGB) Phase III trial randomizing mCRC patients to cetux- imab, bevacizumab, or both in combination with investigator_s choice of FOLFOX or FOLFIRI is currently underway. The efficacy of cetuximab in combination with oxaliplatin was evaluated in Phase II studies, and cetuximab showed promising results in combination with FOLFOX4 in the international Phase II ACRO- BAT trial. A total of 42 EGFR-expressing chemother- apy-naive mCRC patients were enrolled; the results include an overall response rate of 72 percent and a stable disease rate of 23 percent.15 A small Phase II study investigating the combination of cetuximab, oxaliplatin, and capecitabine in a heavily pretreated mCRC population demonstrated acceptable tolerabil- ity and a partial response rate of 27 percent.13 The adjuvant role of cetuximab is currently under- going investigation in the ECOG N-0147 study in which patients are treated with cetuximab alone or cetuximab combined with FOLFOX for six months.Panitumumab. Panitumumab (Vectibix\; Amgen Inc., Thousand Oaks, CA) is an anti-EGFR mono- clonal antibody of the IgG2 isotype that was approved by the FDA in September 2006 for the use in EGFR- expressing colorectal cancers that were refractory to 5-FU, irinotecan, and oxaliplatin-based therapy. In contrast to cetuximab, the IgG2 isotype has limited ability to activate complement-dependent cytotoxicity or ADCC response mechanisms. Therefore, the cyto- toxicity of panitumumab is independent of immune function and results primarily from the blockade and internalization of EGFR.41 Panitumumab is a fully human antibody derived from transgenic mice; its clearance is slower and its immunogenicity is lower than that of chimeric antibodies, such as cetuximab. In the results of a multicenter Phase III study of panitumumab (6 mg/kg every 2 weeks) vs. best supportive care in 463 patients (ECOG PS 0–1 in 86 percent; median age, 62 years) with metastatic CRC, who had progressed on fluoropyrimidine, irinotecan, and oxaliplatin, were reported.19 Treatment with panitumumab resulted in an increased response rate (8 vs. 0 percent) and a 46 percent improvement in progression-free survival (60 vs. 96 days). In the panitumumab-treated group, 90 percent of the patients developed a skin rash, no patients had Grade 3/4 infusion reactions and no anti-panitumumab antibod- ies were detected. There was no difference in overall survival; however, 75 percent of patients in the best supportive care group crossed over and received panitumumab. Panitumumab (2.5 mg/kg once weekly) in combination with irinotecan/5-FU/leucovorin chemo- therapy (IFL) as first-line treatment in mCRC was investigated in 19 patients with a 47 percent response rate and 32 percent stable disease rate.42 The multi- center Phase III PACCE (Panitumumab in Advanced Colorectal Cancer Evaluation) study is currently evaluating the addition of panitumumab to bevacizu- mab and irinotecan-based or oxaliplatin-based thera- py in first-line mCRC. Matuzumab. Matuzumab (EMD 72000, EMD Phar- maceuticals) is a humanized IgG1 anti-EGFR mono- clonal antibody that has shown clinical activity in a number of Phase I studies in advanced solid tumors. Pharmacokinetic and pharmacodynamic testing sup- ports dosing once every three weeks with an optimal dose still undergoing evaluation. Near-complete inhi- bition of EGFR signaling activity occurs at 1,200 mg every three weeks, and a weekly maximum tolerated dose (MTD) has been established at 1,600 mg. A Phase I study in EGFR-positive patients (n = 22) with chemo- therapy-refractory solid tumors receiving matuzumab weekly demonstrated a 23 percent objective response rate and a 27 percent stable disease rate.20 A second Phase I study of weekly matuzumab in advanced refractory gastrointestinal cancers (n = 24; 87 percent CRC) showed a 29 percent response rate.21 Tyrosine Kinase Inhibitors Tyrosine kinase EGFR inhibitors are small mole- cules that inhibit the cytosolic tyrosine kinase domain of EGFR (Fig. 1). The FDA has approved two TKIs, and numerous other EGFR TKIs are still under study (Tables 1 and 2). Gefitinib. Gefitinib is a reversible EGFR TKI, (ZD1839, Iressa\; AstraZeneca, Wilmington, DE) currently undergoing investigation in CRC. It was initially approved in May 2003 for the treatment of locally advanced or metastatic NSCLC, but a recent Phase III trial (ISEL) of gefitinib monotherapy resulted in no significant increase in survival or time to progression compared with placebo in the overall population.45 Thus, gefitinib is no longer indicated for patients with advanced NSCLC, and only patients who have previously benefited from the drug may continue to receive it. In preclinical studies, gefitinib blocked EGFR tyrosine kinase phosphorylation and downstream signaling, inhibited tumor cell growth, and enhanced the effects of chemotherapy. In contrast to antibody- based EGFR inhibition, gefitinib does not seem to induce EGFR internalization or degradation, as demonstrated by the lack of reduction in EGFR cell surface expression after gefitinib treatment.46 Numerous studies have investigated various com- binations of gefitinib and chemotherapy for CRC. Two Phase I studies evaluating gefitinib in combina- tion with IV 5-FU (n = 17) or oral capecitabine (n = 10) in 5-FU-refractory patients with CRC demonstrated no responses, although significant stabilization of disease was noted in 35 to 50 percent of patients.28,47 Combination therapy with irinotecan (bolus or continuous infusion) with or without 5-FU produced only limited activity in several small Phase I and II studies and was associated with dose-limiting toxicity resulting in significant irinotecan dose reductions. Combinations of gefitinib, 5-FU, and oxaliplatin have shown promise. In a Phase II study conducted at Stanford University, the combination of FOLFOX4 with daily gefitinib at 500 mg (IFOX) was investigat- ed in mCRC patients who were chemotherapy-naive (Group A) or had received previous therapy for metastatic disease (Group B).26 Response rates were 78 percent in Group A and 36 percent in Group B. Grades 3 and 4 toxicity included diarrhea (54 percent of patients), neutropenia (52 percent), nausea (22 percent), dehydration (16 percent), and fatigue (16 percent). A second Phase II study of gefitinib at 250 mg per day and FOLFOX6 conducted in 47 EGFR- positive patients with advanced CRC demonstrated a similar response rate of 74 percent (evaluable, n = 39).27 A Phase I trial of gefitinib in combination with oxaliplatin, however, did not demonstrate any objec- tive responses in 14 fluoropyrimidine-refractory and irinotecan-refractory patients.23 The exact contribution of gefitinib to oxalipatin/5-FU combinations requires further investigation in a randomized study. Unlike NSCLC, activating mutations of EGFR, which are thought to contribute to the activity of gefinitib in NSCLC, are extremely rare in CRC. Only one such mutation was found in an examination of 293 patients with CRC.48 Erlotinib. Erlotinib (OSI-774, Tarceva\, OSI Pharmaceuticals, Inc., Genentech, Inc.) is a reversible TKI approved for single-agent use in advanced NSCLC. Much like gefitinib, erlotinib inhibits both EGFR phosphorylation and tumor cell proliferation, and induces both G1-phase cell cycle arrest and apoptosis in human cell lines. Erlotinib at 150 mg per day represents the MTD and is the current approved dosage for use in NSCLC. Compared with gefitinib, erlotinib has been eval- uated as monotherapy in mCRC. Two Phase II studies demonstrated disappointing response rates of 0 and 4 percent, but stable disease was noted in 32 and 25 percent of patients, respectively.Erlotinib also has been evaluated in combination with chemotherapy, with encouraging results. A Phase Ib trial of erlotinib combined with capecita- bine and oxaliplatin in previously treated patients with mCRC resulted in 21.7 percent (5/23) PR and 60.8 percent (14/23) SD.49 In addition, interim data from a Phase II study (n = 22) evaluating this combination in first-line or second-line treatment of mCRC showed an 18 percent overall RR and 64 percent SD.31 A Phase I study evaluating the addition of erlotinib to FOLFOX4 in oxaliplatin-naive patients with mCRC demonstrated a response rate of 67 percent.50 Similar to findings with gefitinib, the combination of erlotinib and irinotecan is associated with dose-limiting toxicities, which resulted in early closure of a Phase I trial of erlotinib plus reduced- dose FOLFIRI. Lapatinib (GW572016). Lapatinib is a reversible dual EGFR/Her-2 TKI, which was well tolerated in early testing. A dose-escalation study in EGFR-positive or Her2-positive patients with various solid tumors (n = 64, including 22 patients with CRC) afforded one complete response (CR) and SD in 22 patients.51 A multicenter Phase II study in second-line treatment of mCRC revealed only one PR in a total of 86 patients receiving lapatinib at 1,250 mg per day.32 Lapatinib also has been investigated in combina- tion with chemotherapy in Phase I trials. No re- sponses were seen in seven patients with advanced CRC enrolled in a Phase I trial of lapatinib in combination with capecitabine.52 Phase I studies conducted in patients with advanced solid malignan- cies revealed no dose-limiting toxicities when lapa- tinib was combined with FOLFOX4, but dose-limiting toxicities of diarrhea and myelosuppression were seen when lapatinib was combined with FOLFIRI. Other TKIs EKB-569 is an irreversible EGFR TKI currently undergoing clinical testing. An initial Phase I study of intermittent and daily dosing demonstrated good tolerability in a variety of advanced solid tumors. Two Phase I/IIa studies have investigated EKB-569 in combination with chemotherapy in patients with advanced CRC. Of 11 evaluable treatment-naive patients with advanced CRC treated with the combi- nation of EKB-569 and FOLFOX-4, 4 (36 percent) demonstrated a partial response and 6 had stable disease.53 A similar 38 percent response rate was seen in 39 evaluable first-line patients with advanced CRC treated with EKB-569 in combination with FOLFIRI.54 PD158780 is a multitargeted reversible TKI that inhibits the entire ErbB family of receptor tyrosine kinases. It is currently undergoing early clinical testing with limited information available in CRC. A Phase I trial in 32 advanced solid malignances revealed side effects of rash, stomatitis, and nausea with a dose-limiting toxicity of diarrhea/dehydration.55 Stable disease occurred in ten patients, and no objective responses were seen. The only Phase II study, conducted in advanced ovarian cancer, dem- onstrated 30 percent SD and 0 percent OR in 105 treated patients. CONCLUSIONS Erb-targeted therapy, in particular EGFR-targeted therapy, is undergoing extensive clinical evaluation. Current areas of active investigation include assess- ment of rash as a potential surrogate marker of response, determination of molecular markers to identify more accurately which patients might benefit from therapy, evaluation of EGFR-targeted agents in combination with chemotherapy or other biologic regimens, and the relevance of EGFR testing to clinical benefit. Antibody targeting of EGFR allows infrequent dosing, limited gastrointestinal toxicity, and potential activation of the immune system via IgG1 antibody isotypes. In contrast, small-molecule EGFR TKIs provide oral dosing, absence of severe allergic reactions, and the possibility for simultaneously inhibiting multiple Erb receptor family members. At present, the antibodies cetuximab and panitumumab are the only anti-EGFR therapies FDA approved for use in CRC. Several trials are currently evaluating cetuximab treatment in both the adjuvant and first-line meta- static settings. Panitumumab also is being studied in the first-line metastatic setting in combination with bevacizumab and either irinotecan-based or oxali- platin-based therapy. Critical issues related to the correlation between cetuximab dose, rash, tumor response, and survival will be investigated in the EVEREST trial, in which patients demonstrating no rash or Grade 1 rash on standard dose cetuximab/ irinotecan will be randomized to maintain or increase cetuximab dose until the development of a rash. Given the many EGFR-target therapies in late-stage clinical evaluation, randomized, controlled trials will be critical to determine the true clinical benefit of each agent. At present, no data exist that compare the different EGFR antibodies and TKIs. EGFR targeting in CRC has proven its clinical activity,BLU-945 but the optimal use of these various agents remains a work in progress.