The epidermal growth factor receptor (EGFR) is a transmembrane protein with an extracellular ligand-binding domain and an intracellular ATP-dependent tyrosine kinase domain. Binding of ligand leads to autophosphorylation and activation of the signaling pathway, which regulates cell differentiation, proliferation, migration, protection from apoptosis, and angiogenesis.1,2 EGFR is either overexpressed or upregulated in a majority of colorectal cancers, and higher degrees of EGFR staining have been correlated with inferior survival in a variety of tumor types.3–6 Preclinical studies demonstrated the ability of monoclonal antibodies directed against EGFR to inhibit malignant cell proliferation and to act synergistically with chemotherapeutic agents in suppressing tumor cell lines.7,8 Although early clinical development of anti-EGFR antibody therapy focused on cetuximab (Erbitux), more recent data have also demonstrated similar benefit with panitumumab (Vectibix), and the spectrum of benefit seen in clinical trials has been found in both advanced and refractory patients as well as in previously untreated patients with metastatic disease.
In selecting therapy for a patient with advanced colorectal cancer, clinicians must now choose among several chemotherapeutic agents (fluoropyrimidines, oxaliplatin, and irinotecan) and decide how to integrate targeted agents with chemotherapy. Bevacizumab (Avastin)–containing regimens are well established, with phase III data demonstrating a survival benefit in the first-line setting with irinotecan-based therapy and in the second-line setting with oxaliplatin-based treatment.9,10 Choosing which EGFR-directed drug to use, which chemotherapy regimen to use as a backbone, and when in the sequence of treatment options to consider EGFR-directed therapy has become more complicated recently based on emerging data.
Cetuximab
Cetuximab is a chimeric IgG1 (immunoglobulin G1) monoclonal antibody that binds to EGFR with high specificity and affinity.7 It was initially approved by the US Food and Drug Administration (FDA) in February 2004 based on response rate data in combination with irinotecan for irinotecan-refractory patients. Although originally used only in patients with EGFR-expressing tumors, a number of studies have since suggested that response is independent of EGFR staining, and this is no longer used as a selection criterion for EGFR-directed therapy.11–13 Rather than EGFR expression, it is KRAS mutation status that has emerged as the most important predictor for lack of benefit from anti-EGFR therapy. Multiple trials demonstrate either no benefit or inferior outcomes with the addition of EGFR-directed therapy in the approximately 40% of advanced colorectal cancer patients with activating KRAS mutations.
Cetuximab is a chimeric IgG1 (immunoglobulin G1) monoclonal antibody that binds to EGFR with high specificity and affinity.7 It was initially approved by the US Food and Drug Administration (FDA) in February 2004 based on response rate data in combination with irinotecan for irinotecan-refractory patients. Although originally used only in patients with EGFR-expressing tumors, a number of studies have since suggested that response is independent of EGFR staining, and this is no longer used as a selection criterion for EGFR-directed therapy.11–13 Rather than EGFR expression, it is KRAS mutation status that has emerged as the most important predictor for lack of benefit from anti-EGFR therapy. Multiple trials demonstrate either no benefit or inferior outcomes with the addition of EGFR-directed therapy in the approximately 40% of advanced colorectal cancer patients with activating KRAS mutations.
In an early phase II trial, cetuximab, given as a single agent to irinotecan-refractory patients, produced a partial response rate of 9%.14 The pivotal BOND-1 trial randomized 329 patients whose disease had progressed after an irinotecan-based regimen to receive both cetuximab and irinotecan or cetuximab as monotherapy (Table 1).15 The response rate in the combination-therapy group was significantly higher (22.9% vs 10.8%; P = 0.007), and the median time to disease progression (TTP) was significantly greater in the combination-therapy group (4.1 months vs 1.5 months; P < 0.001). There was no difference in median survival (8.6 months vs 6.9 months; P = 0.48), although crossover from monotherapy to combination therapy was permitted. The ability of cetuximab to restore sensitivity to treatment with irinotecan was seen regardless of whether patients had been pretreated with oxaliplatin or not.
The CO.17 trial randomized 572 patients with metastatic disease previously treated with a fluoropyrimidine, irinotecan, and oxaliplatin to receive treatment with single-agent cetuximab versus best supportive care (Table 1).16 It was completed prior to knowledge regarding the predictive effects of KRAS status, and the primary endpoint was overall survival (OS). Treatment with cetuximab was associated with improvement in response rate, progression-free survival (PFS), and OS (from 4.6 months to 6.1 months; hazard ratio [HR] for death 0.77; P = 0.005). In a retrospective analysis, cetuximab therapy in KRAS wild-type tumors doubled PFS from 1.9 months to 3.7 months (HR = 0.40; P < 0.0001), whereas there was no effect of cetuximab in the KRAS-mutant group (Table 1).17
Given the popularity of using oxaliplatin in front-line therapy for metastatic colorectal cancer, there was interest in demonstrating a role for cetuximab with irinotecan in second-line treatment, after failure of oxaliplatin. The EPIC trial randomized patients in this setting to receive irinotecan with or without cetuximab, with OS as the primary endpoint (Table 1).18 Although there was a trend favoring the addition of cetuximab, OS was not improved (median survival, 10.7 months with cetuximab-irinotecan vs 10.0 months with irinotecan; HR = 0.975), perhaps confounded by the fact that almost half of the irinotecan-alone patients continued on to receive cetuximab later. Combination therapy did increase the response rate from 4% to 16% (P < 0.0001) and increased PFS from 2.6 months to 4.0 months (HR = 0.692; P < 0.0001).