Ras Mutational Status Is a Biomarker for Resistance to EGFR Inhibitors in Colorectal Carcinoma

EGFR and Ras Oncogenes: Molecular Biology

Receptor tyrosine kinases (RTKs), such as EGFR and vascular endothelial growth factor receptor (VEGFR), are transmembrane proteins with an extracellular ligand-binding domain and an intracellular tyrosine kinase catalytic domain. On binding to their cognate ligands, most RTKs dimerise and become activated through autophosphorylation of intracellular tyrosine residues. Activation of RTKs results in up-regulation of multiple cellular signalling pathways that promote cell growth, survival and angiogenesis or environmental stimuli. Inappropriate activation of RTKs via mutation, overexpression or ectopic ligand production is a frequent feature of human tumour development and progression and is thought to be a major mechanism by which cancer cells subvert normal growth control (2-4). Consequently, in recent years, modulation of RTK signal transduction has been an active area in oncology drug discovery.

EGFR (also called erbB1) and other erbB family RTKs (erbB-2/HER-2-neu, erbB-3/HER-3 and erbB-4/HER-4) encoded by the c-erbB proto-oncogenes have been strongly implicated in cancer development and progression ((3) and reviewed in (5)). Several mechanisms cause aberrant receptor activation, resulting in tyrosine kinase activity, which is observed in cancer. Such mechanisms include receptor overexpression, mutation, ligand-dependent receptor dimerisation and ligand-independent activation. For erbB-2, where a specific ligand has not been identified, activation occurs by homo- or hetero-dimerisation alone, whereas erbB-3 does not have significant kinase activity (5, 6). However, on activation, all four receptors are capable of signal transduction, causing activation of the ras/MAP kinase pathway, the PI3K/Akt pathway, src family kinases and STAT proteins. Activation of these pathways promotes cell proliferation, survival and angiogenesis (7).

Several other downstream signalling mediators (e.g. Akt mTOR, src kinases, STAT proteins, K-ras and MEK1/2) have been reported to bypass EGF-R inhibition by constitutive activation of multiple pathways and some of them (mTOR and MEK1/2 inhibitors) are now being targeted in combination with EGFR inhibitors in early-phase clinical trials (8). Amongst them, the ras/MAP pathway is of potential clinical interest. The ras proteins are members of a large superfamily of guanosine-5′-triphosphate (GTP)-binding proteins that play a complex role in the normal transduction of growth factor receptor-induced signals (9). Stimulation of growth factor receptors, such as EGFR, causes activation of multiple regulatory molecules, including the ras protein. EGFR activates ras by stimulating its binding to GTP. Ras, in its active, GTP-bound state, binds several key target proteins, resulting in the subsequent activation of several downstream pathways, including those mediated by MAP kinase, PI3K and others (10). Engagement of these pathways leads to stimulation of cell-cycle progression, desensitisation of the cell to pro-apoptotic stimuli, changes in cytoskeletal organisation and invasion and other processes required for cell proliferation. Activating mutations in the K-ras gene, which result in EGFR-independent activation of the MAP-kinase pathway, are found in approximately 15-30% of patients with non-small cell lung cancer (NSCLC) and 40-45% of patients with colorectal cancer (CRC) (Table II), and their presence generally correlates with a worse prognosis with respect to the outcome of the cancer. In most cases, the somatic ras missense mutations found in cancer introduce amino acid substitutions at positions 12 (Gly->Val), 13 and 61. These mutations disable the endogenous GTPase activity of the ras protein and cause cancer-associated ras to accumulate in the active, GTP-bound conformation. This, in turn, results in activation of PI3K, MAP kinase and others, causing malignant transformation. Because ras is downstream from EGFR, aberrant ras signalling, like that occurring in cells with mutant K-ras, may lead to dysregulation of ras-dependent pathways and downstream signalling even when the upstream receptor is silenced by anti-EGFR monoclonal antibodies or RTKs (Figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Cascade of EGFR- and ras-modulated signal transduction.

In several studies, K-ras mutations have been significantly associated with a lack of response to EGFR tyrosine kinase inhibitors in patients with NSCLC and a lack of response to cetuximab or panitumumab in patients with advanced CRC, as detailed below. Both findings suggest that EGFR-independent, constitutive activation of the K-ras signalling pathway would impair the response to anti-EGFR drugs (11).

Clinical Data

Effect of K-ras mutation on response to anti-EGFR therapy. The clinical relevance of K-ras mutations has been evaluated retrospectively in several clinical trials investigating the effect of EGFR inhibitors such as cetuximab or panitumumab in first-line treatment of metastatic CRC. Van Cutsem et al. (12) reported the CRYSTAL trial which compared 5-fluorouracil, folinic acid and irinotecan (FOLFIRI) plus cetuximab to FOLFIRI alone. An analysis of 45% of the study population (540 out of 1198 patients) revealed a K-ras mutation in 35.6% of evaluable tumours. This study demonstrated that the addition of cetuximab to FOLFIRI significantly improved progression-free survival (PFS) in K-ras-wild-type (K-ras-WT) patients (hazard ratio (HR)=0.68, p=0.017), while no improvement was observed in patients with K-ras-mutant tumours (K-ras-Mut patients) (HR=1.07, p=0.47). Likewise, the overall response rate (ORR) was significantly improved in K-ras-WT patients (43% vs. 59%, p=0.0025), but not in the mutant population (40% vs. 36%, p=0.46).

The OPUS trial selected a combination of 5-fluorouracil, folinic acid and oxaliplatin (FOLFOX4) as a chemotherapy backbone and investigated the addition of cetuximab in a randomised trial (13). Of 337 patients included in this first-line trial, 233 patients were evaluable for their K-ras status and a K-ras mutation was found in 42% of them. In K-ras-WT patients, the addition of cetuximab to FOLFOX4 caused a significant increase of ORR (61% vs.37%, p=0.011) and PFS (HR=0.57, p=0.016). In contrast, a negative impact on treatment efficacy was noted when cetuximab was applied in K-ras-Mut patients with regard to PFS (HR=1.83, p=0.0192) and ORR (33% vs. 49%, p=0.106).

A comparable effect was also noted in the CAIRO II trial which compared capecitabine/oxaliplatin (CapOx) plus bevacizumab to the same regimen plus cetuximab (14). The addition of cetuximab did not affect ORR or PFS in K-ras-WT patients. However, in K-ras-Mut patients, it induced a markedly shorter duration of PFS (8.6 months vs. 12.5 months, p=0.043) and OS (19.2 months vs. 24.9 months). An explanation for this apparently negative interaction of cetuximab with oxaliplatin-based chemotherapy in K-ras-Mut patients is far from clear.

The PACCE trial was designed to investigate double-targeting of VEGF and EGFR. Patients with metastatic CRC received first-line treatment with irinotecan- or oxaliplatin-based chemotherapy and were randomised to additional treatment with either bevacizumab plus panitumumab or bevacizumab alone (15). A subgroup analysis was performed for patients with irinotecan-based chemotherapy (n=200). In K-ras-WT patients (n=115), the addition of panitumumab to irinotecan/bevacizumab-based therapy induced an ORR of 54% compared to 47% without the EGFR inhibitor. Also in this study, no improvement of ORR was observed when panitumumab was given to K-ras mutant patients (30% vs. 38%). Taken together, the PACCE study and the CAIRO II study indicate that in the presence of VEGF-inhibition by bevacizumab, additional inhibition of EGFR does not provide further clinical benefit.

All four trials described above uniformly demonstrate that K-ras mutation confers resistance to anti-EGFR-directed antibodies. Furthermore, these data are supported by a large body of evidence from phase II and case-control studies showing the lack of efficacy of anti-EGFR antibodies in pre-treated patients (16-21). As a consequence, registration of cetuximab and panitumumab limits their use to patients with K-ras wild-type tumours. Determination of the K-ras mutational status is, therefore, required before the clinical application of anti-EGFR-directed antibodies. At present, multiple methods are available for the detection of K-ras mutations. Given that cross-validation of sensitivity, specificity and reliability is presently being evaluated, a single best method has not yet been defined.

Skin toxicity developing during the first weeks of treatment is an important predictor of response to anti-EGFR therapy (22), but is independent of K-ras mutational status. The greatest benefit from anti-EGFR therapy may, therefore, be expected in K-ras wild-type patients reacting to treatment with marked skin toxicity (23).

The results from these studies should therefore trigger retrospective analyses of K-ras mutations from all available trials in CRC (as well as NSCLC and pancreatic cancer). Analysis of K-ras mutations may also be important in the cetuximab-containing N0147 and the PETACC-8 adjuvant studies in CRC. These studies would enable further establishment of the correlation between K-ras mutation and resistance to cetuximab and panitumumab in CRC patients (22).

Effect of K-ras mutation status on response to anti-VEGF therapy. Given that VEGF is regulated downstream of EGFR and that inhibition of EGFR may cause a down-regulation of VEGF expression, it is of interest to investigate the effect of K-ras mutations on anti-VEGF-directed therapy (24, 25). Ince et al. (26) performed a retrospective analysis of the pivotal phase III trial which tested the addition of bevacizumab to first-line therapy with irinotecan, 5-fluorouracil and leucovorin (IFL) (26, 27). Micro-dissected tumours from 295 patients were available for determination of mutations in K-ras (35%), B-raf (5.6%), and p53 (68%). With regard to overall survival, K-ras- and B-raf wild-type patients had a better prognosis than patients with mutant tumours, but all subgroups showed a benefit from the treatment with bevacizumab. In patients who were wild-type for both K-ras and B-raf, HR in favour of bevacizumab treatment was 0.57 (95% confidence interval (95%CI)=0.31-1.06), while it was 0.67 (95% CI=0.37-1.20) in patients with mutant tumours. Considering the limitations of such a retrospective analysis, it was suggested that the survival benefit induced by bevacizumab is independent of K-ras-, B-raf- and TP53-mutation status (26).

Studies over the last few years have identified several anti-EGFR and anti-VEGFR resistance mechanisms. These findings have led to clinical trials using newly designed targeted therapies that can overcome these resistance mechanisms and have shown promise in laboratory studies. Ongoing research efforts will likely continue to identify additional resistance mechanisms and these findings will hopefully translate into effective therapies for different cancers.