J Clin Oncol 28: by American Society of Clinical Oncology INTRODUCTION

Transcription

1 VOLUME 28 NUMBER 7 MARCH JOURNAL OF CLINICAL ONCOLOGY B I O L O G Y O F N E O P L A S I A Molecular Mechanisms of Resistance to Cetuximab and Panitumumab in Colorectal Cancer Alberto Bardelli and Salvatore Siena See accompanying article on page 1181 From the Laboratory of Molecular Genetics, Institute for Cancer Research and Treatment, University of Torino Medical School, Candiolo; Italian Foundation for Cancer Research Institute of Molecular Oncology; and The Falck Division of Medical Oncology, Department of Oncology, Ospedale Niguarda Ca Granda, Milan, Italy. Submitted June 16, 2009; accepted September 11, 2009; published online ahead of print at on January 25, Supported by grants from the Italian Association for Cancer Research, Italian Ministry of Health, Regione Piemonte, Italian Ministry of University and Research, Association for International Cancer Research, European Union FP6 Migratory Cancer Stem Cells Contract No , Cassa di Risparmio di Torino Progetto Alfieri, and Oncologia Ca Granda Onlus Fondazione. Presented in part at the 45th Annual Meeting of the American Society of Clinical Oncology, May 29-June 2, 2009, Orlando, FL, and the 100th Annual Meeting of the American Association for Cancer Research, April 18-22, Denver, CO. Authors disclosures of potential conflicts of interest and author contributions are found at the end of this article. Corresponding author: Alberto Bardelli, PhD, Laboratory of Molecular Genetics, Institute for Cancer Research and Treatment, University of Torino, Medical School, Str prov 142 Km 3.95, Candiolo, Italy; a.bardelli@ircc.it by American Society of Clinical Oncology X/10/ /$20.00 DOI: /JCO A B S T R A C T Personalized cancer medicine based on the genetic milieu of individual colorectal tumors has long been postulated, but until recently this concept was not supported by clinical evidence. The advent of the epidermal growth factor receptor (EGFR) targeted monoclonal antibodies cetuximab and panitumumab has paved the way to the individualized treatment of metastatic colorectal cancer (mcrc). Here we discuss the evidence that mcrcs respond differently to EGFR-targeted agents and that the tumor-specific response has a genetic basis. We outline how, from the initial observation that cetuximab or panitumumab as monotherapy is effective only in 10% to 20% of mcrcs, knowledge has being gained on the molecular mechanisms underlying primary resistance to these agents. The role of oncogenic activation of EGFR downstream effectors such as KRAS, BRAF, PIK3CA, and PTEN on response to therapy is discussed. We suggest that CRCs lacking oncogenic alterations in these four genes have the highest probability of response to anti-egfr therapies and are defined as quadruple negative. The rapid and effective translation of these findings into predictive biomarkers to couple EGFR-targeted antibodies to the patients who benefit from them is presented as a paradigm of modern clinical oncology. Finally, unresolved questions such as understanding the molecular basis of response as well the mechanisms of secondary resistance are presented as the future fundamental goals in this research field. J Clin Oncol 28: by American Society of Clinical Oncology INTRODUCTION Colorectal cancer (CRC) is the third most common cancer worldwide, and the metastatic disease accounts for 40% to 50% of newly diagnosed patients, which is associated with high morbidity. 1,2 Despite therapeutic advances, the prognosis for patients with metastatic CRC (mcrc) remains poor, with a median overall survival (OS) of 18 to 21 months. 3 Systemic chemotherapy in this setting has traditionally been based on fluorouracil, with the more recent introduction of other cytotoxics agents, such as irinotecan, oxaliplatin, and capecitabine. In the last 5 years, cetuximab and panitumumab, two monoclonal antibodies (moab) targeting the epidermal growth factor receptor (EGFR), have proven to be effective in combination with chemotherapy or as single agents for treatment of mcrc. 4-6 Both of these molecules bind to the extracellular domain of EGFR, thus leading to inhibition of its downstream signaling. EGFR is a transmembrane tyrosine kinase receptor that, on ligand binding, triggers two main signaling pathways. 7 These include the RAS-RAF- MAPK axis, mainly involved in cell proliferation, and the PI3K-PTEN-AKT pathway, mainly involved in cell survival and motility-invasion (Fig 1). MOLECULAR MECHANISMS OF RESPONSE AND RESISTANCE TO EGFR-TARGETED MOABS From the early studies conducted mainly in heavily pretreated chemotherapy-refractory patients and also in chemotherapy-naïve patients with mcrc, it became clear that only 10% to 20% of patients with mcrc clinically benefited from anti-egfr mo- Abs. 5,6 Initially, it was hypothesized that EGFRtargeted agents would be most effective in those tumors overexpressing the protein, based on the previous observation that trastuzumab, a moab against the human epidermal growth factor receptor 2 (HER2) target and a member of the EGFR family of receptors, was effective in patients with breast cancers overexpressing the HER2/neu protein. However, it was promptly documented that the levels of expression of the target molecule (EGFR) were not correlated with clinical by American Society of Clinical Oncology

2 Molecular Mechanisms of Resistance to Cetuximab and Panitumumab DUSPs Grb2 Shc Cetuximab/panitumumab SOS EGFR Ras B-Raf MEK MAPK p85 PI3K response to cetuximab or to panitumumab. 4-6 These evidences, together with the side effects and the higher costs of moab therapies as compared with standard chemotherapy regimens, triggered a flourish of studies on the molecular mechanisms of primary resistance to cetuximab or panitumumab. MOLECULAR EVENTS ASSOCIATED WITH RESPONSE TO EGFR MOABS PDK PTEN Fig 1. Schematic representation of the monoclonal antibodies cetuximab/ panitumumab and of the epidermal growth factor receptor (EGFR) mediated intracellular signaling pathways. The molecules implicated in EGFR signaling and affected by oncogenic alterations are highlighted in red. AKT S6K GSK The initial hypotheses put forward to explain why only a small fraction of patients with mcrc benefit from cetuximab or panitumumab were that the clinical response was associated to (1) molecular alterations affecting the EGFR or (2) its immediate intracellular signal transducers. 8 Specific alterations of the EGFR gene, including somatic mutations and gene copy number variations, had been previously reported as the genetic events underlying the response to gefitinib or erlotinib (tyrosine kinase inhibitors of EGFR) and trastuzumab in non smallcell lung and breast cancer, respectively However, studies on large cohorts of untreated mcrc and on small cohorts of tumor samples taken from treated patients, quickly and unequivocally clarified that EGFR somatic mutations are extremely rare in CRC, and when they do occur, are not associated with response. 8,12 On the contrary, increased copy number of the EGFR gene was found to be associated with response, 8 and later studies on larger cohorts confirmed these findings Interestingly, in patients who experience a response, the copy number increase is modest (three- to five-fold), and cases with frank amplification ( 20-fold) are infrequently found. Importantly, the increased EGFR gene dosage does not seem to translate into increased expression of the corresponding proteins, 8,14,17 so how and why the copy number of the EGFR receptor is correlated with improved response remains largely unclear and warrants additional studies. Albeit confirmed, in patients with mcrc, the statistical correlation between the increased EGFR gene dosage and response to cetuximab or panitumumab is not strong enough to allow the clinical use of this biomarker for the predictive selection of patients. Methodologies to assess gene copy number (fluorescent in situ hybridization/chromogenic in situ hybridization) suffer from technical reproducibility, and the scoring system shows high interlaboratory variability, 17 further complicating the clinical application of this marker. Other molecular alterations have also been positively associated with response. Increased levels of expression of the EGFR ligands amphiregulin and epiregulin are present in responders as compared with individuals who do not experience a response. 18 Specifically, expression of high levels of mrna for either of these ligands was linked to sensitivity to cetuximab monotherapy, as shown by significantly improved disease control rate and longer progression-free survival. 18 One possibility is that an autocrine or paracrine loop generated by the increased expression of these epidermal growth factor (EGF) ligands is responsible for driving the growth of these tumors. Similar to EGFR copy number, so far, the levels of expression of amphiregulin and epiregulin have been difficult to assess systematically, and at present, these markers cannot be used to select patients eligible for cetuximab or panitumumab therapy. 19 In conclusion, it is safe to say that response to therapy does at least require the EGFR receptor target to be present. MOLECULAR EVENTS ASSOCIATED WITH RESISTANCE TO EGFR MOABS In addition to molecular alterations of the EGFR gene, oncogenic activation of EGFR downstream effectors was also investigated with respect to clinical outcome to cetuximab or panitumumab. The signaling cascade initiated by binding of the EGF and other ligands to the EGF receptor comprises two main axes. On one side, the KRAS-RAF- MAPK signaling pathway is thought to control cell growth, differentiation and apoptosis. Kirsten (K)RAS belongs to the gene family of oncogenes (KRAS, HRAS, and NRAS) encoding guanosine di/ triphosphate-binding proteins that act as self-inactivating intracellular signal transducers. 20 Analysis of more than 40,000 human tumors indicates an activating mutation rate of 22%, 8.2%, and 3.7% for KRAS, NRAS, and HRAS, respectively. 21 After Grb2/SOS-mediated activation, guanosine triphosphate-bound KRAS recruits the serine protein BRAF, thus starting a cytoplasmic phosphorylation cascade leading to the activation of transcription factors (Fig 1). The other axis involves membrane localization of the lipid kinase PIK3CA, which promotes AKT activation, ensuing a parallel intracellular propagation of the signal. Importantly, the two axes (KRAS/ BRAF and PIK3CA) are closely related and strictly interconnected, as the p110 subunit of PI3K can also be activated via interaction with RAS proteins 22 (Fig 1). MUTATIONAL STATUS OF KRAS Initial retrospective analyses on a small number of patients led to the breakthrough revelation that patients with CRC carrying activating KRAS gene mutations do not benefit from cetuximab or panitumumab therapy. 23,24 KRAS mutations have since emerged as the major negative predictor of efficacy in patients receiving panitumumab by American Society of Clinical Oncology 1255

3 Bardelli and Siena or cetuximab. Studies in the first line and subsequent lines of treatment have shown that patients with tumors harboring a KRAS mutation neither respond to EGFR-targeted moabs, nor experience survival or quality of life benefit from such treatment Specifically, patients with wild-type KRAS generally have doubled progression-free survival times compared with patients with mutant KRAS. Recently, three large, randomized, III cetuximab studies have been published, including the first-line OPUS (Oxaliplatin and Cetuximab in First-Line Treatment of Metastatic Colorectal Cancer) 28 and CRYSTAL (Cetuximab Combined With Irinotecan in First-Line Therapy for Metastatic Colorectal Cancer) 29 studies and the National Cancer Institute of Canada Clinical Trials Group (NCIC-CTG) monotherapy study conducted in relapsed/refractory patients or those with contraindications to chemotherapy (Table 1). 26,33 It should be noted that the NCIC-CTG monotherapy study did not allow patients from the control group who had disease progression to cross-over to moab treatment. This study design allowed for the best separation of Kaplan-Meier survival curves, thus unequivocally and definitively establishing KRAS mutations as a negative outcome predictor of cetuximabbased therapies in mcrc. The data from the OPUS and CRYS- TAL trials showed that the addition of cetuximab to first-line infusional fluorouracil, leucovorin, and oxaliplatin 28 or fluorouracil, leucovorin, and irinotecan 29 is not beneficial for KRAS-mutated tumors, although these benefit from chemotherapy alone (Table 1). Importantly, results from OPUS suggest that addition of EGFR-targeted treatment to chemotherapy may even be detrimental in such patients. 28 Given the results of these clinical trials, the European health authorities have restricted the use of panitumumab monotherapy, as well as cetuximab monotherapy or combination therapy with chemotherapy, to patients with wild-type KRAS mcrc only. 34 The American Society of Clinical Oncology has also put forward a provisional recommendation with guidelines similar to those of the European Medicines Agency. 35 Very recently, the US Food and Drug administration provided analogous guidelines specifying that panitumumab Table 1. Randomized Clinical Trials of the Effect of Treatment With Cetuximab or Panitumumab Patients With Metastatic Colorectal Cancer Harboring KRAS Mutations Patients With Effect of KRAS Results PFS OS Cetuximab or Name/Phase of Clinical Line of No. of Patients % Wild-Type KRAS Mutant KRAS Wild-Type KRAS Mutant KRAS Panitumumab on Study End Points Trial Treatment (ITT) First Author Treatment No. ITT HR 95% CI HR 95% CI HR 95% CI HR 95% CI in KRAS Mutants Monotherapy of cetuximab or panitumumab Third 463 Amado 27 Best supportive care panitumumab to to to to 1.39 PFS no difference ORR no difference NCIC-017 Third 572 Karapetis 26 Best supportive care cetuximab to to to to 1.37 PFS no difference ORR no difference Combination of cetuximab or panitumumab with chemotherapy PRIME CRYSTAL OPUS II trial PACCE CAIRO2 EPIC 181 First 1,183 Douillard 29 FOLFOX4 panitumumab 1, to to to to 2.05 PFS decreased ORR no difference First 1,198 Van Cutsem 29a FOLFIRI 1, to to to to 1.28 PFS no difference cetuximab ORR decreased First 337 Bokemeyer 28a FOLFOX to to to to 1.91 ORR decreased cetuximab PFS no difference First 1,053 Hecht 30 Oxaliplatin- or Oxaliplatin to to to to 1.54 PFS decreased irinotecanbased 664 chemo- ORR increased therapy, Irinotecan to to to to 5.59 ORR decreased bevacizumab 201 PFS decreased panitumumab OS decreased First 736 Tol 31 Bevacizumab, HR not HR not PFS significantly HR not HR not OS significantly PFS decreased capecitabine, reported reported reduced, reported reported reduced, OS decreased oxaliplatin P.003 P.03 ORR decreased cetuximab Second 1,298 Langer 32 Irinotecan to to to to 2.01 OS decreased cetuximab PFS no difference ORR increased Second 1,186 Peeters 32a FOLFIRI 1, to to to to 1.15 PFS no difference panitumumab ORR no difference Abbreviations: ITT, intention to treat; PFS, progression-free survival; OS, overall survival; HR, hazard ratio; ORR, objective response rate; FOLFIRI, folinic acid, fluorouracil, and irinotecan; FOLFOX4, folinic acid, fluorouracil, and oxaliplatin. HR: cetuximab versus no cetuximab or panitumumab versus no panitumumab. Results from the primary efficacy end point are in bold. A given end point is designated as decreased if there was a markedly lower result and as increased if there was a markedly higher result in the cetuximab or panitumumab group than in the control group. Statistically significant by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

4 Molecular Mechanisms of Resistance to Cetuximab and Panitumumab Fig 2. Graphic representation of a cohort of 100 patients with colorectal cancer treated with cetuximab or panitumumab. The genetic milieu of individual tumors and their impacts on the clinical response are listed. KRAS, BRAF, and PIK3CA somatic mutations as well as loss of PTEN protein expression are indicated according to different color codes. Molecular alterations mutually exclusive or coexisting in individual tumors are indicated using different color variants. The relative frequencies at which the molecular alterations occur in colorectal cancers are described. Responder (15%) KRAS (35%-45%) BRAF (5%-10%) PIK3CA and/or PTEN (15%-20%) KRAS/PIK3CA/PTEN BRAF/PIK3CA/PTEN Nonresponder (20%-25%) as well cetuximab should not be given to patients whose tumors harbor KRAS mutations in codon 12 or It is becoming increasingly clear that KRAS mutational status is a negative predictive marker for cetuximab only in colorectal tumors and not in non small-cell lung cancers. Specifically, in two independent s evaluating two different chemotherapy regimens with or without cetuximab, no significant correlation was found between objective tumor response to cetuximab and KRAS status of the individual lung tumors. 37,38 KRAS-MUTATED TUMORS MAY OCCASIONALLY RESPOND TO EGFR MOAB THERAPY On the basis of the above data, regulatory oversight is now firmly focused on the view that all patients carrying mutated KRAS mcrc should not receive anti-egfr moabs. However, it is intriguingly now coming to light that perhaps not all KRAS mutations are created equal in their impact on mediating EGFR resistance. For example, a very small number of patients carrying KRAS-mutated tumors have been reported to respond to either cetuximab or panitumumab. 16,24,26,39 These findings were initially regarded as erroneous due to inaccurate mutational analysis or to response to chemotherapy given in association with the EGFR-targeted moab in patients who were not truly chemotherapy-refractory. However, further analysis and the continuing identification of such patients now suggests that, in rare instances, KRAS mutations may not impair the response to cetuximab and panitumumab, and that differences in the specific KRAS mutation (eg, codon 12 v 13) could play a role. It is important to stress that these patients likely represent less then 1% of the KRAS-mutated population, and therefore current treatment guidelines are not significantly affected by these data. However, further research in this area is worthy for the following reasons: (1) to confirm the existence of clinically documented response of KRAS-mutated tumors, (2) to identify specific KRAS gene variants and/or concurrent molecular events that may be associated with the response in these tumors, and (3) to define clinically useful algorithms for the identification of patients with KRAS mutation who may benefit from cetuximab/panitumumab. MUTATIONAL STATUS OF BRAF The occurrence of KRAS mutations only accounts for approximately 35% to 45% of nonresponsive patients. Accordingly, although KRAS mutation can be considered a highly specific negative biomarker of response (specificity of 93%), it is also poorly sensitive (sensitivity of 47%). 40 The identification of additional genetic determinants of primary resistance to EGFR-targeted therapies in colorectal cancers is therefore important to prospectively identify patients who should not receive either cetuximab or panitumumab, thus avoiding their exposure to ineffective and expensive therapy. Recent work has therefore been focused on the molecular analysis of the molecules involved in downstream EGFR signaling, such as BRAF mutations and oncogenic deregulation of the PIK3CA/PTEN signaling (Fig 1). Mutations in BRAF have been recently shown to impair responsiveness to panitumumab or cetuximab in patients with mcrc. 41 The initial retrospective work was performed on a cohort of 132 patients, showing that none of the patients who experienced a response displayed BRAF mutations, whereas 11 (14.0%) of 79 nonresponders carried a BRAF V600E allele. 41 Of note, KRAS and BRAF mutations are known to be mutually exclusive in colorectal cancers. 42 Therefore, mutated BRAF tumors (approximately 10%) add algebraically to those carrying KRAS mutations (35% to 45%), thus further empowering the selection of patients eligible for cetuximab/panitumumab treatment. Of note, when considered together, the two markers can identify up to 55% of nonresponders 41 (Fig 2). Results from other groups recently reported at the 2009 annual meetings of the American Association of Cancer Research and the by American Society of Clinical Oncology 1257

5 Bardelli and Siena American Society of Clinical Oncology confirmed these data. 43,44 Prospective analysis of larger cohorts are now highly needed to confirm these data. Considering that genomic DNA suitable for mutational analyses should be available for the OPUS, CRYSTAL, and the NCIC-CTG studies, these confirmatory analyses should become available relatively soon. We believe that, similarly to KRAS, BRAF mutations should also be exploited to prospectively select patients who are unlikely to respond to cetuximab and panitumumab in mcrc. Importantly, it has recently emerged that mutated BRAF also represents a negative prognostic factor for mcrcs. According to the presently available data, it seems clear that BRAF mutations define a genetically distinct subset of CRCs characterized by an extremely poor response to both targeted and chemiotherapic therapeutic regimens. Although there are currently no drugs available for the specific and direct inhibition of KRAS, a number of agents designed to inhibit the kinase activity of BRAF are either already clinically approved or are progressing through the pipeline of I and II studies. 45 The concomitant treatment of patients with mcrc bearing BRAFmutated tumors with cetuximab/panitumumab in combination with a BRAF inhibitor are possible and supported by a strong rationale. Studies addressing this possibility are underway or have already been planned. ONCOGENIC ACTIVATION OF THE PIK3CA/PTEN SIGNALING PATHWAY In addition to KRAS and BRAF, the EGF receptor also activates the PI3K signaling pathway (Fig 1). The latter can be oncogenically deregulated either by activating mutations in the PIK3CA p110 subunit or by inactivation (often by epigenetic mechanisms) of the PTEN phosphatase. The role of deregulated PIK3CA/PTEN signaling on the response to cetuximab and panitumumab has therefore been investigated. A gross analysis of current data regarding the impact of PIK3CA mutations and PTEN loss on response is conflicting. 16,36,46-48 From the overall analysis of the published work, it seems that PIK3CA mutations are in fact associated with the resistance, although this correlation is nowhere close to that observed for KRAS or BRAF. However, most of the authors agree that PTEN inactivation is a negative predictor of response. 16,49 Because this alteration is currently being evaluated by immunohistochemistry, the lack of standardization in the clinical application of this analysis is likely to be responsible for the current equivocal results. Importantly, although KRAS and BRAF mutations seem to be mutually exclusive, PIK3CA mutations or PTEN inactivation can coexist (ie, occur in the same tumor) with KRAS/BRAF mutations (Fig 2). This may also explain why it is more difficult to definitively prove the individual contribution of PIK3CA mutations and PTEN loss to the lack of response than that of KRAS and BRAF mutations. Furthermore, it has been convincingly shown that PIK3CA mutations located in the exon 9 and 20 hotspots exert different biochemical and oncogenic properties and are differently activated by KRAS. 50 It is therefore conceivable that only some of the PIK3CA mutations may be associated with resistance to cetuximab and panitumumab. Considering their relatively low frequencies, the fact that they can co-occur with KRAS and BRAF, and the different oncogenic properties of the different PIK3CA mutations, it will be necessary to analyze large cohorts of patients before drawing definitive conclusions. In summary, it is likely that both the PIK3CA and PTEN molecular status will be useful to further optimize the algorithm used to identify eligible patients (Fig 2). However, these two markers are not yet ready to be used in the clinic. USE OF MULTIPLE MARKERS TO PREDICT CLINICAL OUTCOME TO CETUXIMAB OR PANITUMUMAB Current data suggest that the evaluations of not only KRAS mutations but also BRAF and PIK3CA/PTEN alterations could be useful for selecting patients with mcrc who are unlikely to respond to anti EGFR-targeted antibodies. 16,36,46-48 EGFR gene copy number detection and evaluation of the altered expression of receptor ligands and EGFR gene polymorphisms also seem relevant to positively identify responders. 8,12-16 At present, each of these markers has been mainly assessed as a single event, often in retrospective analyses. The fact that some of these events display an overlapping pattern of occurrence further complicate their analysis. As a result, a clear identification of which biomarkers should be used together with KRAS in the clinical setting remains to be defined. Comprehensive integrated analysis of the entire oncogenic pathway triggered by the EGFR is likely to enhance the prediction ability of the markers used individually. Recently published data indicate that when expression of PTEN and mutations of KRAS, BRAF, and PIK3CA are concomitantly ascertained, up to 70% of patients with mcrc unlikely to respond to anti-egfr therapies can be identified (Fig 2). 51 Similar to the nomenclature used to classify breast cancers, we propose to define these tumors as quadruple negative. The limited sensitivity of current sequencing methods in detecting point mutations may at least partially account for apparent absence of mutations. Oncogenic deregulation of the same four genes by a mechanism other than mutation (such as amplification) may also occur. The improvement of more sensitive technologies will address both of these issues. Another possibility is the occurrence of alterations in other key elements of the EGFR-dependent signal cascade (eg, AKT1 or MAPK) and the presence of genetic alterations in tyrosine kinase receptors other than EGFR, providing an alternate pathway of survival and/or proliferation (Fig 2). In summary, concomitant analysis of multiple genetic and epigenetic events involved in the cellular events triggered by the EGFR will likely improve tailoring of cetuximab and panitumumab therapies. MODELING RESISTANCE IN CELLULAR MODELS Up until now, investigators have mainly exploited retrospective analysis of treated patients as a means to study resistance and sensitivity to anti-egfr moabs. However, given that genetic manipulation techniques can be applied to cellular models, one can envisage developing in vitro tools to prospectively find new sensitivity/resistance biomarkers that can then be confirmed in patients and even be used to screen for rational drug combinations to reverse resistance. A few early proofof-concept studies using cellular models have been performed using CRC cells sensitive to cetuximab and panitumumab. In these, ectopic expression of either mutant KRAS or BRAF confers resistance to both of these agents. 24,52 However, these early models have the concern that the mutant alleles are driven by viral promoters to high expression levels, which may not recapitulate all aspects of patient responses to monotherapy or combination therapy. A new generation of cell-based tumor progression models has been recently developed. These include by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

6 Molecular Mechanisms of Resistance to Cetuximab and Panitumumab the somatic knock-out or knock-in of individual cancer alleles in normal or neoplastic cells. As a result, endogenous loci are targeted for mutation or correction, thus generating paired cell lines closely recapitulating the occurrence of cancer mutations in individual patients. 53,54 For example, we have introduced quadruple-negative colorectal cancer cells KRAS and BRAF mutations and observed that these mutations impart a marked degree of resistance to cetuximab or panitumumab. These studies strongly indicate that the correlation between the occurrence of KRAS and BRAF mutations and the lack of clinical response to EGFR-targeted antibodies is not generically a result of poorer prognosis of KRAS-BRAF mutated tumors; rather, the oncogenic activation of KRAS and BRAF could bypass the EGFRinitiated signalling cascade, thus conferring primary resistance to cetuximab and panitumumab. Supporting this evidence, it has been shown that CRC cell lines carrying preexisting alterations of KRAS- BRAF and/or PIK3CA-PTEN are generally not inhibited by cetuximab treatment. 46 Of note, in cancer cells carrying constitutively active KRAS and BRAF mutants, the pharmacologic inhibition of the MAPK signaling cascade has been shown to improve anti-egfr treatment with moabs. In this regard, it has been recently reported that treatment with the BRAF inhibitor sorafenib can restore sensitivity to panitumumab or cetuximab of CRC cells carrying the V600E allele. 52 Overall, these studies have provided the rationale for combination therapies, targeted simultaneously the EGFR and RAS-RAF-MAPK signaling pathways in patients with CRC. Accordingly, clinical trials aimed at simultaneous inhibition of EGFR (with cetuximab) and BRAF (with sorafenib) have been initiated. Additional studies coupling various inhibitors of the RAF-MAPK signaling cascade to anti- EGFR moabs are also planned. WHAT HAVE WE LEARNED SO FAR AND WHAT NEEDS TO BE DONE? What Makes Tumors Sensitive to Cetuximab and Panitumumab? The observation that a subset of CRCs respond to anti-egfr antibody therapies on the basis of knowledge of which other mutant alleles are also present has heralded the widespread realization that CRC medicine is now inevitably going to become more personalized. This has therefore stimulated research and development of clinically validated diagnostic tools for the prospective identification of responder patients. Remarkably, these studies have also increased our knowledge of the molecular basis of CRC progression. Although outstanding progress has been rapidly made, several questions remain unresolved and warrant further studies. In particular, although we have learned a lot about the molecular events that contribute to resistance, we do not really know what underlines clinical efficacy of cetuximab and panitumumab. We know that responsive CRC tumors carry wild-type KRAS/BRAF alleles and tend to have increased copy number of the EGFR receptor gene. In addition, studies indicate that responders show increased levels of amphiregulin and epiregulin and may carry a unique milieu of EGF and EGFR gene polymorphisms (SNPs). How do these factors contribute in determining the response? One possibility is that a subset of CRCs develop in the absence of KRAS/BRAF mutations because, at the cellular level, the effect of these alterations is substituted by an autocrine or paracrine loop involving EGF ligands. This would then generate tumors addicted/ dependent on the interaction between the amphiregulin/epiregulin EGFR ligand receptor pair. Cetuximab and panitumumab would then be effective by blocking this functional interaction, and this would explain the selective response of these tumors to the antibody therapy. What would then be the role of increased EGFR gene copy number and that of the EGF or EGFR polymorphisms in this scenario? Contrary to other situations in which the genomic locus corresponding to an oncogene is frankly amplified (PIK3CA, MET, and so on), thus resulting in largely augmented expression of the corresponding protein product, it seems clear that the increase in EGFR gene copy number is often modest (three- to five-fold) and does not translate into a significant increase of the receptor protein. It is possible that the limited resolution of the immunohistochemistry technique that is used to assess the expression levels of the receptor may have so far limited the detection of small EGFR protein increases. However, if this finding is confirmed (ie, the protein level does not match the gene copy), then the functional effect of increased gene copies may be qualitative rather then quantitative. At the moment, we can only speculate that a nonrandom duplication of mutant EGFR alleles that have so far remained elusive (affecting ligand receptor binding, for example) may represent the basis for this event. This is a testable hypothesis that could be addressed by genotyping large cohorts of tumors from treated patients in which the EGFR copy number is also known. Cellular models are similarly needed to functionally evaluate these hypotheses and to identify the molecular basis of response (and not only of resistance) to cetuximab and panitumumab. What Are the Mechanisms of Secondary Resistance? Another and possibly even more urgent question is to identify the mechanisms of secondary resistance to anti-egfr antibody therapies in mcrc. Clinical data indicate that even the best responses obtained in KRAS/BRAF wild-type tumors are transient, and even in the best cases, they do not last for longer than 12 to 18 months. 6 Virtually nothing is known regarding why the response is temporary and the tumors (in most cases, liver metastasis), after a massive initial reduction, rapidly begin to regrow and concomitantly become refractory to further anti-egfr treatment. The timeframe of these events is compatible with emergence of resistant clones through a Darwinian selection process. The difficulties in accessing samples from cetuximab- as well as panitumumab-resistant tumors are among the reasons that have thus far impaired research on this topic. It is clear that clinical studies designed to ethically obtain biopsies of liver metastasis from patients who have become refractory to therapy will be pivotal to address this question. Once available, these samples will allow the assessment of a number of hypotheses. Among them, the first will be the molecular status of the genetic markers (KRAS/BRAF/PIK3CA/ PTEN) already known or suspected to confer primary resistance. Should secondary resistance be based on mechanisms different from those contributing to primary resistance, then further analysis comparing the genetic and proteomic status of matched (from the same patient) sensitive versus resistant tissue samples will likely shed light on these mechanisms. In addition to studies on samples from treated patients, there is an urgent need for modeling secondary resistance to cetuximab and panitumumab by means of cellular models. In conclusion, the introduction of anti-egfr moabs as well as of other targeted agents for the treatment of common solid cancers (such as CRC and lung cancer) has led to the definition of new paradigms in clinical oncology. These can be summarized as follow: Therapies based on molecular targets are effective on subsets of patients whose tumors by American Society of Clinical Oncology 1259

7 Bardelli and Siena carry specific molecular alterations. The latter are being used to redesign the taxonomy of solid tumors (such as CRC) by moving it from a histologic- to a genetic-based level. Accordingly, the design of clinical trials to assess targeted agents should incorporate the analysis of the genomic milieu of individual tumors. The data generated from molecularly annotated clinical trials can then lead to the identification of genetic alterations underlying resistance and sensitivity that can be rapidly translated into clinically applicable molecular markers (such as KRAS). Considering the staggering costs of targeted therapies (especially of those based on biologic agents), the selection of patients will be key to their economic sustainability. The final goal will be to maximize the therapeutic index of novel agents (such as cetuximab and panitumumab), while at the same time providing the actual translation of individualized therapy into the clinical setting. AUTHORS DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a U are those for which no compensation was received; those relationships marked with a C were compensated. For a detailed description of the disclosure categories, or for more information about ASCO s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: None Consultant or Advisory Role: None Stock Ownership: None Honoraria: Alberto Bardelli, Amgen, Merck; Salvatore Siena, Amgen, Merck Research Funding: None Expert Testimony: None Other Remuneration: None AUTHOR CONTRIBUTIONS Conception and design: Alberto Bardelli, Salvatore Siena Financial support: Alberto Bardelli Collection and assembly of data: Alberto Bardelli, Salvatore Siena Data analysis and interpretation: Alberto Bardelli, Salvatore Siena Manuscript writing: Alberto Bardelli, Salvatore Siena Final approval of manuscript: Alberto Bardelli, Salvatore Siena REFERENCES 1. Ferlay J, Autier P, Boniol M, et al: Estimates of the cancer incidence and mortality in Europe in Ann Oncol 18: , Centers for Disease Control and Prevention: United States Cancer Statistics: US Cancer Statistics Working Group Poston GJ, Figueras J, Giuliante F, et al: Urgent need for a new staging system in advanced colorectal cancer. J Clin Oncol 26: , Saltz LB, Meropol NJ, Loehrer PJ Sr, et al: Phase II trial of cetuximab in patients with refractory colorectal cancer that expresses the epidermal growth factor receptor. J Clin Oncol 22: , Chung KY, Shia J, Kemeny NE, et al: Cetuximab shows activity in colorectal cancer patients with tumors that do not express the epidermal growth factor receptor by immunohistochemistry. J Clin Oncol 23: , Cunningham D, Humblet Y, Siena S, et al: Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med 351: , Ciardiello F, Tortora G: EGFR antagonists in cancer treatment. N Engl J Med 358: , Moroni M, Veronese S, Benvenuti S, et al: Gene copy number for epidermal growth factor receptor (EGFR) and clinical response to antiegfr treatment in colorectal cancer: A cohort study. Lancet Oncol 6: , Lynch TJ, Bell DW, Sordella R, et al: Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 350: , Paez JG: EGFR mutations in lung cancer: Correlation with clinical response to gefitinib therapy. Science 304: , Hudis CA: Trastuzumab: Mechanism of action and use in clinical practice. N Engl J Med 357:39-51, Barber TD, Vogelstein B, Kinzler KW, et al: Somatic mutations of EGFR in colorectal cancers and glioblastomas. N Engl J Med 351:2883, Sartore-Bianchi A, Moroni M, Veronese S, et al: Epidermal growth factor receptor gene copy number and clinical outcome of metastatic colorectal cancer treated with panitumumab. J Clin Oncol 25: , Cappuzzo F, Finocchiaro G, Rossi E, et al: EGFR FISH assay predicts for response to cetuximab in chemotherapy refractory colorectal cancer patients. Ann Oncol 19: , Personeni N, Fieuws S, Piessevaux H, et al: Clinical usefulness of EGFR gene copy number as a predictive marker in colorectal cancer patients treated with cetuximab: A fluorescent in situ hybridization study. Clin Cancer Res 14: , Frattini M, Saletti P, Romagnani E, et al: PTEN loss of expression predicts cetuximab efficacy in metastatic colorectal cancer patients. Br J Cancer 97: , Moroni M, Veronese S, Benvenuti S, et al: Copy number increase of the epidermal growth factor receptor gene is associated with clinical response of colorectal cancer to ANTI-EGFR antibody therapy. Ann Oncol 16:7, Khambata-Ford S, Garrett CR, Meropol NJ, et al: Expression of epiregulin and amphiregulin and K-ras mutation status predict disease control in metastatic colorectal cancer patients treated with cetuximab. J Clin Oncol 25: , Balko JM, Black EP: A gene expression predictor of response to EGFR-targeted therapy stratifies progression-free survival to cetuximab in KRAS wild-type metastatic colorectal cancer. BMC Cancer 9:145, Malumbres M, Barbacid M: RAS oncogenes: The first 30 years. Nat Rev Cancer 3: , Wellcome Trust Sanger Institute: Catalogue of Somatic Mutations in Cancer Gupta S, Ramjaun AR, Haiko P, et al: Binding of ras to phosphoinositide 3-kinase p110alpha is required for ras-driven tumorigenesis in mice. Cell 129: , Lièvre A, Bachet JB, Le Corre D, et al: KRAS mutation status is predictive of response to cetuximab therapy in colorectal cancer. Cancer Res 66: , Benvenuti S, Sartore-Bianchi A, Di Nicolantonio F, et al: Oncogenic activation of the RAS/RAF signaling pathway impairs the response of metastatic colorectal cancers to anti-epidermal growth factor receptor antibody therapies. Cancer Res 67: , Lièvre A, Bachet JB, Boige V, et al: KRAS mutations as an independent prognostic factor in patients with advanced colorectal cancer treated with cetuximab. J Clin Oncol 26: , Karapetis CS, Khambata-Ford S, Jonker DJ, et al: K-ras mutations and benefit from cetuximab in advanced colorectal cancer. N Engl J Med 359: , Amado RG, Wolf M, Peeters M, et al: Wildtype KRAS is required for panitumumab efficacy in patients with metastatic colorectal cancer. J Clin Oncol 26: , Bokemeyer C, Bondarenko I, Makhson A, et al: Fluorouracil, leucovorin, and oxaliplatin with and without cetuximab in the first-line treatment of metastatic colorectal cancer. J Clin Oncol 27: , a. Bokemeyer C, Bondarenko I, Hartmann JT, et al: Overall survival of patients with KRAS wildtype tumours treated with FOLFOX4 cetuximab as 1st-line treatment for metastatic colorectal cancer: The OPUS study. Eur J Cancer 7:S346, 2009 (suppl; abstr 6079) 29. Douillard J, Siena S, Cassidy J, et al: Randomized 3 study of panitumumab with FOLFOX4 compared to FOLFOX4 alone as 1st-line treatment (tx) for metastatic colorectal cancer (mcrc): The PRIME trial. Eur J Cancer 7:S6, 2009 (suppl; abstr 10LBA) 29a. Van Cutsem E, Rougier P, Köhne C, et al: A meta-analysis of the CRYSTAL and OPUS studies combining cetuximab with chemotherapy (CT) as 1st-line treatment for patients (pts) with metastatic colorectal cancer (mcrc): Results according to KRAS and BRAF mutation status. Eur J Cancer 7:S345, 2009 (suppl; abstr 6077) by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

8 Molecular Mechanisms of Resistance to Cetuximab and Panitumumab 30. Hecht J, Mitchell E, Chidiac T, et al: A randomized IIIB trial of chemotherapy, bevacizumab, and panitumumab compared with chemotherapy and bevacizumab alone for metastatic colorectal cancer. J Clin Oncol 27: , Tol J, Koopman M, Cats A, et al: Chemotherapy, bevacizumab, and cetuximab in metastatic colorectal cancer. N Engl J Med 360: , Langer C, Kopit J, Awad M, et al: Analysis of K-RAS mutations in patients with metastatic colorectal cancer receiving cetuximab in combination with irinotecan: Results from the EPIC trial. Ann Oncol 19:viii133, 2008 (suppl 18) 32a. Peeters M, Price T, Hotko Y, et al: Randomized 3 study of panitumumab with FOLFIRI vs FOLFIRI alone as second-line treatment (tx) in patients (pts) with metastatic colorectal cancer (mcrc). Eur J Cancer 7:S10, 2009 (suppl; abstr 14LBA) 33. Jonker DJ, O Callaghan CJ, Karapetis CS, et al: Cetuximab for the treatment of colorectal cancer. N Engl J Med 357: , European Medicines Agency: Committee for Medicinal Products for Human Use May 2008 Plenary Meeting monthly report Allegra CJ, Jessup JM, Somerfield MR, et al: American Society of Clinical Oncology Provisional Clinical Opinion: Testing for KRAS gene mutations in patients with metastatic colorectal carcinoma to predict response to anti-epidermal growth factor receptor monoclonal antibody therapy. J Clin Oncol 27: , GenomeWeb Daily News: FDA adds KRAS testing info to Vectibix, Erbitux Labels (Press Release, Genome Web News). dxpgx/fda-adds-kras-testing-info-vectibix-erbituxlabels 37. O Byrne KJ, Bondarenko I, Barrios C, et al: Molecular and clinical predictors of outcome for cetuximab in non-small cell lung cancer (NSCLC): Data from the FLEX study. J Clin Oncol 27:408s, 2009 (suppl; abstr 8007) 38. Khambata-Ford S, Harbison C, Woytowitz D, et al. K-Ras mutation (mut), EGFR-related, and exploratory markers as response predictors of cetuximab in first-line advanced NSCLC: Retrospective analyses of the BMS099 trial. J Clin Oncol 27:412s, 2009 (suppl; abstr 8021) 39. Loupakis F, Pollina L, Stasi I, et al: PTEN expression and KRAS mutations on primary tumors and metastases in the prediction of benefit from cetuximab plus irinotecan for patients with metastatic colorectal cancer. J Clin Oncol 27: , Linardou H, Dahabreh IJ, Kanaloupiti D, et al: Assessment of somatic k-ras mutations as a mechanism associated with resistance to EGFR-targeted agents: A systematic review and meta-analysis of studies in advanced non-small-cell lung cancer and metastatic colorectal cancer. Lancet Oncol 9: , Di Nicolantonio F, Martini M, Molinari F, et al: Wild-type BRAF is required for response to panitumumab or cetuximab in metastatic colorectal cancer. J Clin Oncol 26: , Rajagopalan H, Bardelli A, Lengauer C, et al: Tumorigenesis: RAF/RAS oncogenes and mismatchrepair status. Nature 418:934, De Roock W, De Schutter J, Prenen H, et al: PIK3CA mutations do not confer resistance to the EGFR-inhibitor cetuximab in chemorefractory metastatic colorectal cancer. 100th Annual Meeting of the American Association for Cancer Research, Denver, CO, April 18-22, 2009 (abstr 1896) 44. Laurent-Puig P, Cayre A, Pezet D, et al: Analysis of PTEN, BRAF and EGFR status in determining benefit from cetuximab therapy in patients with metastatic colon cancer in wild-type KRAS patients. 100th Annual Meeting of the American Association for Cancer Research, Denver, CO, April 18-22, 2009 (abstr 1897) 45. Garber K: Trial offers early test case for personalized medicine. J Natl Cancer Inst 101: , Jhawer M, Goel S, Wilson AJ, et al: PIK3CA mutation/pten expression status predicts response of colon cancer cells to the epidermal growth factor receptor inhibitor cetuximab. Cancer Res 68: , Sartore-Bianchi A, Martini M, Molinari F, et al: PIK3CA mutations in colorectal cancer are associated with clinical resistance to EGFR-targeted monoclonal antibodies. Cancer Res 69: , Prenen H, De Schutter J, Jacobs B, et al: PIK3CA mutations are not a major determinant of resistance to the epidermal growth factor receptor inhibitor cetuximab in metastatic colorectal cancer. Clin Cancer Res 15: , Perrone F, Lampis A, Orsenigo M, et al: PI3KCA/PTEN deregulation contributes to impaired responses to cetuximab in metastatic colorectal cancer patients. Ann Oncol 20:84-90, Zhao L, Vogt PK: Helical domain and kinase domain mutations in p110alpha of phosphatidylinositol 3-kinase induce gain of function by different mechanisms. Proc Natl Acad Sci U S A 105: , Sartore-Bianchi A, Di Nicolantonio F, Nichelatti M, et al: Multi-determinants analysis of molecular alterations for predicting clinical benefit to EGFRtargeted monoclonal antibodies in colorectal cancer. PLoS One 4:e7287, Di Nicolantonio F, Martini M, Molinari F, et al: BRAF V600E confers resistance to cetuximab or panitumumab metastatic colorectal cancer. Eur J Cancer 6:81, 2008 (suppl) 53. Rago C, Vogelstein B, Bunz F: Genetic knockouts and knockins in human somatic cells. Nat Protoc 2: , Di Nicolantonio F, Arena S, Gallicchio M, et al: Replacement of normal with mutant alleles in the genome of normal human cells unveils mutationspecific drug responses. Proc Natl Acad Sci U S A 105: , 2008 Journal of Oncology Practice Now Available on PubMed Central Beginning with the January 2010 issue, full text for all JOP articles will be available on the NIH PubMed Central archive 4 months after publication. Once available, content is then delivered to PubMed, a metadata repository overseen by the National Library of Medicine and used by many scientific, technical, and medical researchers, making JOP material discoverable in PubMed for the first time. Find yourself in PubMed. Submit your manuscript to JOP at jopsubmissions@asco.org by American Society of Clinical Oncology 1261