WO2014182521A1

WO2014182521A1 - Diagnostic methods and treatments for cancer

Info

Publication number: WO2014182521A1
Application number: PCT/US2014/036104
Authority: WO
Inventors: Brandon Higgs; Jiaqi Huang; Chris Morehouse; Philip Brohawn; Yihong Yao; Liyan JIANG; Wei Zhu
Original assignee: Medimmune, Llc; Shanghai Chest Hospital
Priority date: 2013-05-06
Filing date: 2014-04-30
Publication date: 2014-11-13

Abstract

Described herein is a method for selecting a treatment for cancer in a patient in need thereof. In one embodiment, the method includes a step of determining whether one or more mutations are present in a tissue sample obtained from the patient; and administering to the patient an effective amount of an anti-cancer therapy tailored to treat cancer cells having one or more of the mutations.

Description

DIAGNOSTIC METHODS AND TREATMENTS FOR CANCER

INTRODUCTION Reference to a Sequence Listing

This application incorporates by reference a Sequence Listing submitted with this application as text file named KRAS-500WO1 SL created on April 25, 2014, and having a size of 1.38 MB. Field of the Invention

The present invention relates to cancer diagnostics and therapies. In particular, the present invention relates to methods for diagnosing tumors that include at least a subset of cells that may be resistant to EGFR (epidermal growth factor receptor) therapy. Background of the Invention

One of the major paradigm shifts over the last fifteen years in anti-cancer therapy is the introduction of targeted therapy. This approach differs from non-specific first line treatments by focusing on activated protein targets related to cellular growth and proliferation and anti- apoptotic processes. Such treatment strategies have proven to be highly effective, specifically when applied to aggressive cancers in certain patients, although clinical improvement is often not sustained. Recent findings show that low frequency variants within the genes that code for these targets may explain the lack of sustained benefit in patients (Diaz et al. (2012) The molecular evolution of acquired resistance to targeted EGFR blockade in colorectal cancers. Nature.

486:537-40).

Currently, physicians depend on traditional mutation-detection strategies to guide treatment decisions with targeted therapies - the gold standard for assays being either Sanger sequencing, allele-specific PCRs, for example, amplification refractory mutation system

(ARMS)-PCR, or fluorescence in situ hybridization (FISH). For example, KRAS mutational status is required in all metastatic colorectal cancer (mCRC) patients who are eligible to receive EGFR mAb therapy and detection of activating mutations within the EGFR gene in NSCLC patients is a routine diagnostic determinant for gefitinib or erlotinib treatment. Additionally, the Vysis ALK Break Apart FISH Probe Kit (Abbott Molecular, Inc.), which detects rearrangements of the ALK gene, was approved as a companion diagnostic test for crizotinib treatment.

However, even within patient subgroups harboring such variants, not all respond to the targeted treatment. First-line treatment response rates of EGFR- tyrosine kinase inhibitors (EGFR-TKIs) in patients with NSCLC containing EGFR mutations are between about 55% and about 90% (Neal and Sequist (2010) First-line use of EGFR tyrosine kinase inhibitors in patients with NSCLC containing EGFR mutations. Clin. Adv. Hematol. Oncol. 8(2): 119-26. Review), and many responding patients eventually develop resistance to the targeted therapy. Secondary mutations of EGFR and/or amplification of c-MET are responsible for only 50% of these EGFR- TKI resistance cases, leaving unknown molecular attribution for roughly half of the remaining patient population. Other biomarkers such as PIK3CA mutations and loss of PTEN have been investigated. However, their impact on clinical efficacies of EGFR therapies are not conclusive (Van Cutsem, et al. (2007) Open-label phase III trial of panitumumab plus best supportive care compared with best supportive care in patients with chemotherapy-refractory metastatic colorectal cancer. J. Clin. Oncol. 25: 1658-64; and Cappuzzo et al. (2008) Primary resistance to cetuximab therapy in EGFR FISH-positive colorectal cancer patients. Br. J. Cancer. 99(l):83-9.).

As such, there still remains a need for diagnostic methods to help guide anti-cancer therapies, particularly in the context of targeted anti-cancer therapies. SUMMARY OF THE INVENTION

Described herein is a method for selecting a treatment for cancer in a patient in need thereof. In one embodiment, the method includes a step of determining whether one or more mutations are present in a tissue sample obtained from the patient, wherein one or more mutations are selected from: (a) one or more mutations selected from a mutation in a cytoplasmic domain of MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof; (b) one or more mutations in an extracellular domain of ALK selected from H368P, F241L, VI 98 A, L170P, and combinations thereof; (c) a mutation in a kinase domain of ALK including Rl 181C; (d) one or more mutations selected from a mutation in an extracellular domain of EGFR selected from E66*, S170R, L173P, N182S, C236R, F420L, and combinations thereof; and (e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof; and administering to the patient an effective amount of an anti-cancer therapy tailored to treat cancer cells having one or more of the mutations.

In another embodiment, a method of detecting the presence of one or more mutations in a tissue sample is provided. The method includes, amplifying nucleic acid from the tissue sample; sequencing the amplified nucleic acid; and determining whether one or more mutations are present. In one embodiment, the one or more mutations are selected from(a) one or more mutations selected from a mutation in a cytoplasmic domain of MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof; (b) one or more mutations in an extracellular domain of ALK selected from H368P, F241L, V198A, L170P, and combinations thereof; (c) a mutation in a kinase domain of ALK including Rl 181C; (d) one or more mutations selected from a mutation in an extracellular domain of EGFR selected from E66*, S170R, L173P, N182S, C236R, F420L, and combinations thereof; and (e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof.

In another embodiment, a method for treating non-small cell lung cancer (NSCLC) or colorectal cancer (CRC) in a patient is provided, wherein the method includes: obtaining a tissue sample from the patient; determining whether the tissue sample includes at least a subset of cells including one or more mutations selected from(a) one or more mutations selected from a mutation in a cytoplasmic domain of MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof; (b) one or more mutations in an extracellular domain of ALK selected from H368P, F241L, V198A, L170P, and combinations thereof; (c) a mutation in a kinase domain of ALK including Rl 181 C; (d) one or more mutations selected from a mutation in an extracellular domain of EGFR selected from E66*, S170R, L173P, N182S, C236R, F420L, and combinations thereof; and (e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof; and if at least a subset of cells includes one or more of the mutations, administering to the patient an effective amount of an anti-cancer therapy tailored to treat cancer cells having one or more of the mutations. In another embodiment, a method for determining whether a tissue sample obtained from a patient includes at least a subset of EGFR anti-cancer therapy non-responsive cells is provided. In one embodiment, the method includes determining whether one or more mutations are present in the tissue sample, wherein one or more mutations are selected from: (a) one or more mutations selected from a mutation in a cytoplasmic domain of MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof; (b) one or more mutations in an extracellular domain of ALK selected from H368P, F241L, V198A, L170P, and combinations thereof; (c) a mutation in a kinase domain of ALK including Rl 181C; (d) one or more mutations selected from a mutation in an extracellular domain of EGFR selected from E66*, S170R, L173P, N182S, C236R, F420L, and combinations thereof; and (e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof; wherein the presence of one or more of the mutations in the tissue sample indicates that a least a subset of cells in the tissue sample may be resistant to EGFR anti-cancer therapy.

Also provided is a kit for determining whether a tissue sample includes at least a subset of EGFR-therapy resistant cells, wherein the kit includes: one or more reagents for determining a presence of one or more mutations selected from(a) one or more mutations selected from a mutation in a cytoplasmic domain of MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof; (b) one or more mutations in an extracellular domain of

ALK selected from H368P, F241L, V198A, L170P, and combinations thereof; (c) a mutation in a kinase domain of ALK including Rl 181C; (d) one or more mutations selected from a mutation in an extracellular domain of EGFR selected from E66*, S170R, L173P, N182S, C236R, F420L, and combinations thereof; and (e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof; and instructions for performing the assay. In one embodiment, the kit also includes a container for the reagents.

In another embodiment, a pharmaceutical composition for treating a cancer patient in which one or more mutations have been detected is provided. In one embodiment, one or more mutations are selected from: (a) one or more mutations selected from a mutation in a cytoplasmic domain of MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof; (b) one or more mutations in an extracellular domain of ALK selected from H368P, F241L, VI 98 A, L170P, and combinations thereof; (c) a mutation in a kinase domain of ALK including Rl 181C; (d) one or more mutations selected from a mutation in an extracellular domain of EGFR selected from E66*, S170R, L173P, N182S, C236R, F420L, and combinations thereof; and (e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof. In one embodiment, the pharmaceutical composition includes: (i) a therapeutically effective amount of an anti-cancer therapeutic agent tailored to treat cancer cells having one or more of the mutations; and (ii) a pharmaceutically acceptable carrier.

In one embodiment, one or more mutations are detected in a tissue sample that includes a genetically heterogeneous population of cancer cells. In one embodiment, one or more mutations are present in at least a subset of cells in the genetically heterogeneous tissue sample. In one embodiment, one or more mutations include low frequency mutations. In a more particular embodiment, one or more mutations have a frequency of less than about 10%. In another embodiment, one or more mutations have a frequency of less than about 5%. In one embodiment, one or more mutations have a frequency of at least about 1%.

In one embodiment, the presence of one or more mutations indicates that at least a subset of cells may be resistant to EGFR-therapy.

In one embodiment, one or more mutations include one or more mutations selected from a mutation in a cytoplasmic domain of MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof. In one embodiment, one or more mutations further include one or more low frequency K-RAS active domain mutations in combination with one or more low frequency EGFR kinase domain mutations. In one embodiment, the K-RAS active domain mutation includes one or more point substitutions in codons 12 or 13. In one embodiment, the K-RAS active domain mutation includes G12R. In one embodiment, the EGFR kinase domain mutation includes G719A. In one embodiment, the K-RAS active domain mutation includes G12R and the EGFR kinase domain mutation includes G719A.

In one embodiment, one or more mutations include one or more mutations in an extracellular domain of ALK selected from H368P, F241L, V198A, L170P, and combinations thereof. In one embodiment, one or more mutation further include more than one low frequency mutations in K-RAS and EGFR in combination with MET and ALK gene copy number increases.

In one embodiment, one or more mutations include a mutation in a kinase domain of ALK including Rl l 81C.

In another embodiment, one or more mutations include one or more mutations in an extracellular domain of EGFR selected from E66*, S170R, L173P, N182S, C236R, F420L, and combinations thereof. In another embodiment, one or more mutations include one or more mutations in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof. In one embodiment, E66* is a heterozygous stop code mutation located in an N- terminus of EGFR. In one embodiment, R832C is a homozygous deletion mutation in a kinase domain of EGFR.

In one embodiment, one or more mutations further include a mutation in an extracellular domain of MET. In one embodiment, the mutation includes C541G.

In one embodiment, the anti-cancer therapy includes at least one non-EGFR anti-cancer therapeutic. In one embodiment, the non-EGFR anti-cancer therapeutic is selected from:

chemotherapeutic agents, growth inhibitory agents, cytotoxic agents, agents used in radiation therapy, anti-angiogenesis agents, apoptotic agents, anti-tubulin agents, neutralizing antibodies, tyrosine kinase inhibitors, platelet derived growth factor inhibitors, COX-2 inhibitors, interferons, cytokines, organic chemical agents, and combinations thereof. In one embodiment, the non-EGFR anti-cancer therapeutic includes one or more chemotherapeutic agents selected from: an alkylating agent, an antimetabolite, an antitumor antibiotic, an antimitotic, a topoisomerase inhibitor, a proteasome inhibitor, tyrosine kinase inhibitors, and combinations thereof. In one embodiment, the non-EGFR anti-cancer therapeutic agent includes one or more chemotherapeutic agents selected from: (a) an alkylating agent selected from cisplatin, carboplatin, oxaliplatin, cyclophosphamide, nitrogen mustard, melphalan, chlorambucil, busulphan and nitrosourea; (b) an antimetabolite selected from antifolates, raltitrexed, gemcitabine, capecitabine, methotrexate, pemetrexed (Alimta), cytosine arabinoside and hydroxyurea; (c) antitumor antibiotics selected from anthracyclines, bleomycin, doxorubicin, daunomycin, epirubicin, idarubicin, mitomycin-C, dactinomycin and mithramycin; (d) antimitotic agents selected from vinca alkaloids and taxoids; (e) topoisomerase inhibitors selected from epipodophyllotoxins, irinotecan, amsacrine, topotecan and camptothecin; (f) proteasome inhibitors; (g) tyrosine kinase inhibitors selected from Axitinib, Bosutinib,

Cediranib, Dasatinib, Erlotinib, Gefitinib, Imatinib, Lapatinib, Lestaurtinib, Nilotinib,

Semaxanib, Sunitinib, and Vandetanib; and combinations thereof.

In one embodiment, the anti-cancer therapy includes combination therapy. In one embodiment, combination therapy includes an EGFR inhibitor in combination with a non-EGFR anti-cancer therapeutic agent. In one embodiment, the EGFR inhibitor is selected from cetuximab, panitumumab, erlotinib, gefitinib, and combinations thereof. In one embodiment, the EGFR inhibitor and the non-EGFR chemotherapeutic agent are administered simultaneously. In one embodiment, the EGFR inhibitor and the non-EGFR chemotherapeutic agent are

administered sequentially.

In one embodiment, cancer includes non-small cell lung cancer (NSCLC) or colorectal cancer (CRC).

In one embodiment, the patient is human. In one embodiment, the patient is treatment naive.

In one embodiment, the tissue sample includes a solid or fluid tissue sample. In one embodiment, the tissue sample is obtained from a cancerous tissue.

In one embodiment, the presence of the mutation is determined by amplifying nucleic acid from the tumor and sequencing the amplified nucleic acid. In one embodiment, the presence of the low frequency mutation is determined using next generation sequencing (NGS) technology selected from 454 sequencing, Solexa, SOLiD, Polonator, and HeliScope Single Molecule Sequencer technologies.

BRIEF DESCRIPTION OF THE FIGURES

Figures 1A and B are graphs showing (A) CRC patient cumulative curves summating the most prevalent nucleotide alternative to the reference genome (MPNAR) frequencies (blue lines) at each frequency threshold on the x-axis, with the proportion of CRC patients sharing similar breakpoints at each MPNAR threshold frequency on the x-axis, and (B) a histogram of the MPNAR frequencies for a representative CRC patient.

Figure 2 is a graph showing the distribution of mutations detected in 39 CRC patients across the coding regions of the KRAS gene. Each point corresponds to a patient and each black tick mark in the orange protein coding block indicates the presence of at least one mutation identified in at least a single patient.

Figures 3A-D are graphs showing variant frequencies for mutations identified in Patient A (blue), Patient B (red), and Patient C (green) tumor biopsy specimens in (A) MET, (B) KRAS, (C) EGFR, and (D) ALK. The y-axis in the barplots for each gene has been ceilinged to 5% frequency (except in the KRAS plot) to allow better visualization of the low frequency mutation differences between patients. For each plot, the chromosome ideogram, variant frequency values, gene structure, and physical coordinates are provided.

Figures 4A and B are pictures of fluorescent in situ hybridization (FISH) detection of EML4-ALK fusion using Spectrum Green labeled EML4 probe (green signal) and Spectrum Red labeled ALK probe (red signal). (A) EML4-ALK fusion gene was indicated with white arrow in patient B; (B) Both EML4 and ALK gene copy number were increased in patient C.

DETAILED DESCRIPTION

Introduction

Many targeted cancer therapies lack durability and demonstrate a limited overall efficacy in patients. The lack of durability and efficacy may be due to intra-tumor heterogeneity, which can result in the co-existence of several driver mutations within different, or even the same key oncogenes, which may impact the success of a targeted cancer therapy. A key to successful targeted therapy, therefore, may be detecting low frequency mutations in a tumor, for example, detecting mutations in a primary tumor at an early stage, and developing an appropriate therapy based on the presence or absence of particular low frequency mutations.

However, traditional mutation-detection strategies only reveal mutations occurring in most cancer cells, namely, the expanded clonal population, while neglecting low-frequency mutations. Further, technical limitations of detection sensitivity in traditional sequencing procedures may also prevent detection of low- frequency mutations. Although traditional Sanger sequencing has been effectively used for detection of treatment-relevant somatic mutations, Sanger sequencing often fails to detect low frequency mutations in a heterogeneous mixture of cancerous and normal tissue. In one study, Sanger sequencing failed to identify EGFR mutations in primary lung tumor samples with approximately 10% variant frequencies (Thomas et al. (2007) High-throughput oncogene mutation profiling in human cancer. Nat. Genet. 39(3):347-51. Erratum in: Nat. Genet. 2007;39:567).

Detection of low frequency mutations may be important for the development of treatment strategies based on combination therapy and may reduce the incidence of resistance to targeted therapy over time due to the pre-existence of low frequency mutations in oncogenes. Next- generation sequencing (NGS) technologies, such as ultra-deep sequencing, provide a viable alternative for detecting low frequency mutations.

Terminology

Unless otherwise defined, scientific and technical terms used herein shall have the meanings that are commonly understood by those of ordinary skill in the art. Further, unless otherwise required by context, singular terms shall include pluralities and plural terms shall include the singular. Generally, nomenclatures used in connection with, and techniques of, cell and tissue culture, molecular biology, and protein and oligo- or polynucleotide chemistry and hybridization described herein are those well-known and commonly used in the art. Amino acids may be referred to herein by either their commonly known three letter symbols or by the one- letter symbols recommended by the IUPAC-IUB Biochemical Nomenclature Commission.

Nucleotides, likewise, may be referred to by their commonly accepted single-letter codes.

As used in accordance with the present disclosure, the following terms, unless otherwise indicated, shall be understood to have the following meanings:

As used herein, the term "about" is used to modify, for example, the quantity of an ingredient in a composition, concentration, volume, process temperature, process time, yield, flow rate, pressure, and ranges thereof, employed in describing the invention. The term "about" refers to variation in a numerical quantity that can occur, for example, through typical measuring and handling procedures used for making compounds, compositions, concentrates or

formulations; through inadvertent error in these procedures; through differences in the manufacture, source, or purity of starting materials or ingredients used to carry out the methods, and other similar considerations. The term "about" also encompasses amounts that differ due to aging of a formulation with a particular initial concentration or mixture, and amounts that differ due to mixing or processing a formulation with a particular initial concentration or mixture. The term "diagnosis" is used herein to refer to the identification of a molecular or pathological state, disease or condition, such as the identification of cancer or to refer to identification of a cancer patient who may benefit from a particular treatment regimen. The term "prognosis" is used herein to refer to a prediction of the likelihood of disease or disease progression, including, for example, recurrence, metastatic spread, and drug resistance, of a neoplastic disease and the likelihood of clinical benefit from anti-cancer therapy. The term "prediction" is used herein to refer to the likelihood that a patient will respond either favorably or unfavorably to a particular anti-cancer therapy. In one embodiment, the prediction relates to the extent of those responses. In one embodiment, the prediction relates to whether and/or the probability that a patient will survive or improve following treatment, for example treatment with a particular therapeutic agent, and for a certain period of time without disease recurrence. The methods described herein can be used clinically to help inform treatment decisions by providing information on appropriate treatment for a particular patient. The methods described herein can also be used to predict whether or not a patient is likely to respond favorably to a treatment regimen, such as a given therapeutic regimen, including for example, administration of a given therapeutic agent or combination, or whether long-term survival of the patient, following a therapeutic regimen is likely. In a particular embodiment, the methods described herein can provide an indication as to whether or not a patient will favorably respond to EGFR therapy.

As used herein, the terms "EGFR inhibitor," "EGFR therapy," and "EGFR therapeutic" are used interchangeably to refer to a chemical composition that, upon administration to a patient, results in inhibition of a biological activity associated with activation of the epidermal growth factor receptor (EGFR) in the patient, including downstream biological effects resulting from the binding of a natural ligand to EGFR. In general, "inhibit" refers to the ability to decrease or reduce an activity, function, and/or amount as compared to a reference. In one embodiment, EGFR inhibitors include agents that can block EGFR activation or downstream biological effects of EGFR activation that are relevant to treating cancer in a patient. In one embodiment, an EGFR inhibitor can act by binding directly to the intracellular domain of the receptor, for example, to inhibit kinase activity. These types of EGFR inhibitors can also be referred to as tyrosine kinase inhibitors (TKIs). In other embodiments, an EGFR inhibitor can act by occupying the ligand binding site or a portion thereof of the EGF receptor, thereby making the receptor inaccessible to its natural ligand such that normal biological activity is prevented or reduced. In another embodiment, an EGFR inhibitor can act by modulating the dimerization of EGFR polypeptides, or interaction of EGFR polypeptide with other proteins, or enhance ubiquitination and endocytotic degradation of EGFR. In some embodiments, the EGFR inhibitor is a monoclonal antibody that binds to the extracellular domain of EGFR, preventing ligand binding and interrupting the signaling cascade. Antibody inhibitors may also be referred to as "anti-EGFR inhibitors." EGFR inhibitors also include, but are not limited to, low molecular weight inhibitors, antisense constructs, small inhibitory RNAs (i.e. RNA interference by dsRNA; RNAi), and ribozymes. In one embodiment, the EGFR inhibitor is a small organic molecule or an antibody that binds specifically to human EGFR. A variety of EGFR inhibitors have been evaluated in a variety of clinical settings and tumor types, including colorectal cancer, non- small-cell lung cancer (NSCLC), and squamous cell carcinoma of the head and neck (SCCHN).

The terms "EGFR-therapy resistant" or "non-responsive" refer to a patient, tumor or a subset of tumor cells in a patient that demonstrate(s) stable disease or progressive disease after administration of one or more EGFR therapies. In contrast, the term "responsive" means that a patient or tumor shows a complete or partial response after administering an EGFR therapeutic. "Responsiveness" or "non-responsiveness" can be assessed using any endpoint indicating a benefit to the patient, including, but not limited to: (1) inhibition, at least to some extent, of tumor growth, including slowing down and complete growth arrest; (2) reduction in the number of tumor cells; (3) reduction in tumor size; (4) inhibition (i.e., reduction, slowing down or complete stopping) of tumor cell infiltration into adjacent peripheral organs and/or tissues; (5) inhibition (i.e. reduction, slowing down or complete stopping) of metastasis; (6) enhancement of anti-tumor immune response, which may, but does not necessarily, result in the regression or rejection of the tumor; (7) relief, to some extent, of one or more symptoms associated with the tumor; (8) increase in the length of survival following treatment; and/or (9) decreased mortality at a given point of time following treatment. The term "increased resistance" refers to a decreased response to a standard dose or treatment protocol for a particular therapeutic agent or treatment option. The term "mutations indicative of EGFR-therapy resistance" refers to mutations in at least a subset of cells in a tumor or tissue sample from a patient that are associated with resistance to one or more EGFR-therapies.

The term "gene" is used broadly to refer to a nucleic acid associated with a biological function. Thus, genes include coding sequences and/or the regulatory sequences required for their expression. The term "gene" applies to a specific genomic sequence, as well as to a cDNA or an mR A encoded by that genomic sequence. Genes also include non-expressed nucleic acid segments that, for example, form recognition sequences for other proteins. Non-expressed regulatory sequences include "promoters" and "enhancers," to which regulatory proteins such as transcription factors bind, resulting in transcription of adjacent or nearby sequences.

The term "mutation" refers to a difference in the amino acid or nucleic acid sequence of a particular protein or nucleic acid (gene, RNA) relative to a reference sequence for the protein or nucleic acid, respectively. A mutated protein or nucleic acid can be expressed from or found on one allele (heterozygous) or both alleles (homozygous) of a gene, and may be somatic or germ line. Mutations can include point mutations and rearrangements. Point mutations are alterations in an amino acid sequence that involve one or a few nucleotide changes. Examples of point mutations include single-base substitutions and frameshift mutations. Single-base substitutions include transitions (the substitution of pyrimidine for another pyrimidine, or one purine for another purine) and transversions (the substitution of one base type for another base type, for example, a pyrimidine for a purine or vice versa). A "silent mutation" is a mutation that does not change the coding of a three-base codon. Frameshift mutations arise from additions or deletions of one or a few bases that result in the miscoding of one or more downstream codons. Sequence rearrangements can include deletions, inversions, translocations and duplications. As used herein, the term "mutations" can also refer to copy number increases, for example, gene copy numbers of at least about 2, 3, 4, 5 or more.

The term "native sequence" refers to a polypeptide or nucleotide having the same amino acid or nucleic acid sequence as a sequence derived from nature. Thus, a native sequence can have the sequence that is naturally occurring in any mammal. Native sequences can be isolated from nature or can be produced by recombinant or synthetic means. The term "native sequence" specifically encompasses naturally occurring truncated or secreted forms of a polypeptide (e.g., an extracellular domain sequence), naturally occurring variant forms (e.g., alternatively spliced forms) and naturally occurring allelic variants of the polypeptide. In one embodiment, the term "native sequence" refers to a polypeptide or nucleotide that does not include one or more mutations described herein.

The term "next generation sequencing" refers to a variety of high-throughput sequencing technologies that parallelize the sequencing process, producing thousands or millions of sequences at once. Next generation sequencing (NGS) is generally conducted with the following steps: First, DNA sequencing libraries are generated by clonal amplification by PCR in vitro; second, the DNA is sequenced by synthesis, such that the DNA sequence is determined by the addition of nucleotides to the complementary strand rather through chain-termination chemistry; third, the spatially segregated, amplified DNA templates are sequenced simultaneously in a massively parallel fashion without the requirement for a physical separation step. NGS parallelization of sequencing reactions generates hundreds of megabases to gigabases of nucleotide sequence reads in a single instrument run. Unlike conventional sequencing techniques, such as Sanger sequencing, which simply report the average genotype of an aggregate collection of molecules, NGS technologies digitally tabulate the sequence of many individual DNA fragments, such that low frequency variants (i.e., variants present at less than about 10%, 5% or 1% frequency in a heterogeneous population of nucleic acid molecules) can be detected. For this reason, NGS technologies are often referred to as "ultra-deep sequencing." The term "massively parallel" can also be used to refer to the simultaneous generation of sequence information from many different template molecules by NGS .

Next generation sequencing (NGS) strategies can include several methodologies, including, but not limited to: (i) microelectrophoretic methods; (ii) sequencing by hybridization; (iii) real-time observation of single molecules, and (iv) cyclic-array sequencing. Cyclic-array sequencing refers to technologies in which a sequence of a dense array of DNA is obtained by iterative cycles of template extension and imaging-based data collection. Commercially available cyclic-array sequencing technologies include, but are not limited to 454 sequencing, for example, used in 454 Genome Sequencers (Roche Applied Science; Basel), Solexa technology, for example, used in the Illumina Genome Analyzer (San Diego, CA), the SOLiD platform (Applied Biosystems; Foster City, CA), the Polonator (Dover/Harvard) and HeliScope Single Molecule Sequencer technology (Helicos; Cambridge, MA). Although these platforms are quite diverse in sequencing biochemistry as well as in how the array is generated, their work flows are conceptually similar. Other next generation sequencing methods include single molecule real time sequencing (Pacific Bio) and Ion semiconductor sequencing (Ion Torrent sequencing). See, Shendure J and Ji H. (2008) Next Generation DNA Sequencing. Nature Biotech. 26(10): 1135- 1145 for a more detailed discussion of NGS sequencing technologies. As used herein, the term "patient" or "subject" refers to any member of the subphylum cordata, including, without limitation, humans and other primates, including non-human primates such as chimpanzees and other apes and monkey species. Farm animals such as cattle, sheep, pigs, goats and horses; domestic mammals such as dogs and cats; laboratory animals including rodents such as mice, rats and guinea pigs; birds, including domestic, wild and game birds such as chickens, turkeys and other gallinaceous birds, ducks, geese, and the like are also non-limiting examples. The terms "mammals" and "animals" are included in this definition.

The phrase "percent identical" or "percent identity" refers to the similarity between at least two different sequences. This percent identity can be determined by standard alignment algorithms, for example, the Basic Local Alignment Tool (BLAST) described by Altshul et al. (1990) J. Mol. Biol, 215: 403-410; the algorithm of Needleman et al. (1970) J. Mol. Biol, 48: 444-453; or the algorithm of Meyers et al. (1988) Comput. Appl. Biosci., 4: 11-17. A set of parameters may be the Blosum 62 scoring matrix with a gap penalty of 12, a gap extend penalty of 4, and a frameshift gap penalty of 5. The percent identity between two amino acid or nucleotide sequences can also be determined using the algorithm of E. Meyers and W. Miller (1989) CABIOS, 4: 11-17, which has been incorporated into the ALIGN program (version 2.0), using a PAM120 weight residue table, a gap length penalty of 12 and a gap penalty of 4. The percent identity is usually calculated by comparing sequences of similar length.

"Polynucleotide," or "nucleic acid," as used interchangeably herein, refer to polymers of nucleotides of any length, and include DNA and RNA. The nucleotides can be

deoxyribonucleotides, ribonucleotides, modified nucleotides or bases, and/or their analogs, or any substrate that can be incorporated into a polymer by DNA or RNA polymerase. A nucleic acid can be single-stranded or double-stranded. Unless otherwise indicated, a nucleic acid sequence encompasses complementary sequences, in addition to the sequence explicitly indicated. The term "oligonucleotide" refers to short, often single stranded synthetic

polynucleotides that are less than about 200 nucleotides in length. The terms "oligonucleotide" and "polynucleotide" are not mutually exclusive.

The term "primer" refers to a short single stranded polynucleotide, generally with a free 3'-OH group, that binds to a target potentially present in a sample of interest by hybridizing with a target sequence, and thereafter promotes polymerization of a polynucleotide complementary to the target. The term "Sanger sequencing" refers to a method of DNA sequencing based on selective incorporation of labeled chain-terminating dideoxynucleotides (ddNTPs) during in vitro DNA replication. Sequence information is obtained using cycles of template denaturation, primer annealing and primer extension. Each round of primer extension is stochastically terminated by incorporation of labeled ddNTPs. In the resulting mixture of end-labeled extension products, the label on the terminating ddNTP of any given fragment corresponds to the nucleotide identity of its terminal position. Sequence can be determined by high-resolution electrophoretic separation of the single-stranded, end-labeled extension products in a capillary-based polymer gel. Laser excitation of fluorescent labels as fragments of discrete lengths exit the capillary, coupled to four-color detection of emission spectra, provides the readout that is represented in a Sanger sequencing 'trace'. Software can translate these traces into DNA sequences, while also generating error probabilities for each base-call.

The phrase "targeted anti-cancer therapy" refers to administration of anti-cancer therapeutics that target known tumor associated antigens such that the exposure of malignant cells to the targeted therapeutic is preferentially increased, while reducing the exposure of normal cells to the targeted therapeutic. For example, targeted anti-cancer therapy can employ specific antibodies or other binding ligands that act on known targets or biologic pathways that, when activated or inactivated, may cause regression or destruction of the malignant process or by coupling cytotoxic agents to specific antibodies or binding agents that act on known cancer targets.

The phrase "tailored anti-cancer therapy" refers to the selection of one or more anticancer therapies based on the genetic composition of cells or subset of cells present in the tumor. For example, tailored anti-cancer therapy can refer to the selection of one or more anti-cancer therapies based on the genetic composition of cells or a subset of cells present in a heterogeneous population of cancer cells in either primary and/or metastatic tumors. The term "intratumor heterogeneity" refers to the diversity of genetic composition of a cancer tumor in a particular patient.

As used herein, the term "treatment naive" refers to a patient that has not received prior treatment for cancer. "EGFR-treatment naive" refers to a patient that has not received prior EGFR-therapy. "Treatment experienced" refers to a patient that has previously undergone treatment for cancer. "EGFR treatment experienced" refers to a patient that has previously received EGFR-therapy.

The term "target sequence" refers to a polynucleotide or protein of interest, the detection of which is desired. In one embodiment, the target sequence is a polynucleotide sequence of interest, in which a mutation is suspected or known to reside. "Detection" includes any means of detecting, including direct and indirect detection. In one embodiment, "detection" refers to sequencing, including the use of NGS technologies.

The terms "treat" and "treatment" refer to both therapeutic treatment and prophylactic or preventative measures, wherein the object is to prevent or slow down an undesired physiological change or disorder, such as the development or spread of cancer. Beneficial or desired clinical results include, but are not limited to, alleviation of symptoms, reducing extent of disease, stabilized state of disease, delay or slowing of disease progression, amelioration or palliation of the disease state, and remission (partial or total), whether detectable or undetectable. "Treatment" can also mean prolonging survival as compared to expected survival if not receiving treatment. Those in need of treatment include those already with the condition or disorder as well as those prone to have the condition or disorder or those in which the condition or disorder is to be prevented. The term "long-term survival" refers to survival for at least about 1 year, 5 years, 8 years, or 10 years following therapeutic treatment.

The term "tumor" refers to all neoplastic cell growth and proliferation, whether malignant or benign, and all pre-cancerous and cancerous cells and tissues. A tumor can include a genetically heterogeneous mixture of cancerous cells. The term "resistant" refers to cancer, cancerous cells, a tumor and/or at least a subset of cells within a tumor that do not respond completely, or loses or shows a reduced response over the course of cancer therapy to a cancer therapeutic. In one embodiment, a resistant tumor includes at least a subset of cells that are resistant to one or more EGF therapies.

The term "variant" with respect to a polypeptide refers to an amino acid sequence that is altered by one or more amino acids with respect to a reference sequence. Variants include, but are not limited to, allelic variants, splice variants, derivative variants, substitution variants, deletion variants, and/or insertion variants, fusion polypeptides, orthologs, and interspecies homologs. In certain embodiments, a variant may include additional residues at the C- or N- terminus, such as, but not limited to, leader sequence residues, targeting residues, amino terminal methionine residues, lysine residues, tag residues and/or fusion protein residues.

Diagnostic Methods/Methods of Treatment

Intra-tumor heterogeneity has been speculated to be at least partially responsible for the varied responses to certain cancer treatments that are observed. The traditional model of human cancers starts with a DNA mutation in a single cell, followed by malignant cell colonial expansion, and potential additional genetic aberrations. The continuing acquisition of genetic alterations can result in the emergence of subsets of cells with varying genotypes, which may result in phenotypic advantages such as invasion, proliferation, and/or the ability to colonize to different organs (Fidler and Kripke (1970) Metastasis results from preexisting variant cells within a malignant tumor Science. 197: 893-895). The presence of more than one genotype for cancer cells within a given tumor mass as well as the presence of different genetic alterations in different metastatic tumors from a single patient have been identified in several tumor types (Katona et al. (2007) Genetically heterogeneous and clonally unrelated metastases may arise in patients with cutaneous melanoma. Am J Surg Pathol. 31 : 1029-1037; Liegl et al. (2008) Heterogeneity of kinase inhibitor resistance mechanisms in GIST. J Pathol. 216: 64-74; and Maley et al. (2006) Genetic clonal diversity predicts progression to esophageal adenocarcinoma. Nat Genet. 38: 468-473).

Described herein are methods that include determining whether or not at least a subset of cells from a tumor or tissue sample, which may be present at low frequency in at an early stage (e.g., in the primary tumor) are present, and using this information to guide the choice of anticancer therapy. In one embodiment, NGS can be used to detect low-frequency and expanded clonal mutations in primary tumors to help better inform treatment decisions.

In one embodiment, the presence or absence of one or more mutations in a sample is used to help inform treatment of a cancer patient. In one embodiment, the presence or absence of one or more low frequency mutations in at least a subset of cells in a tissue sample or tumor is determined. The term "low frequency" refers to a mutation that occurs at a frequency of less than about 10%, 5%, 4%, 3%, 2% or 1% in a heterogeneous mixture of cancer cells, although, in some embodiments, mutations that occur at a frequency of less than about 1%, 0.5% or 0.1% may be discounted as noise or artifacts, for example, associated with sequencing errors. In one embodiment, the low frequency mutation is present at a frequency capable of being detected by traditional Sanger sequencing. In other embodiments, the low frequency mutation is present at a frequency that is not capable of being detected by traditional Sanger sequencing. In one embodiment, a tissue sample is a heterogeneous sample that includes at least a subset of cells with one or more mutations in one or more oncogenes. In one embodiment, one or more mutations can arise from different cellular clones.

In one embodiment, the presence or absence of one or more mutations is determined by amplifying nucleic acid from a tissue sample and sequencing the amplified nucleic acid. As used herein, the term "amplification," refers to the process of producing multiple copies, for example, at least 2 copies, and up to millions of copies of a desired sequence. The duplicated region (i.e., the amplified DNA) can be referred to as an "amplicon."

In one embodiment, the presence or absence of one or more mutations is determined using next generation sequencing (NGS) technology. In one embodiment, the NGS technology involves cylic-array technology, for example, cyclic-array technology selected from 454 sequencing, Solexa, SOLiD, Polonator, and HeliScope Single Molecule Sequencer technologies.

The term "cancer" refers to a physiological condition in mammals that is characterized by abnormal or uncontrolled cell growth and/or proliferation, immortality, metastasis, interference with the normal functioning of neighboring cells, release of cytokines or other secretory products at abnormal levels, suppression or aggravation of inflammatory or immunological response, neoplasia, premalignancy, malignancy, invasion of surrounding or distant tissues or organs, such as lymph nodes, and/or other characteristic morphological features. Often, cancer cells will be in the form of a tumor, but cancer cells may exist alone within an animal, or may circulate in the blood stream as independent cells, such as leukemic cells. Examples of cancer include, but are not limited to, carcinoma, lymphoma, blastoma, sarcoma, and leukemia or lymphoid

malignancies. More particular examples of cancers include kidney or renal cancer, breast cancer, colon cancer, rectal cancer, colorectal cancer (CRC), lung cancer including small-cell lung cancer, non-small cell lung cancer (NSCLC), adenocarcinoma of the lung and squamous carcinoma of the lung, squamous cell cancer (e.g. epithelial squamous cell cancer), cervical cancer, ovarian cancer, prostate cancer, liver cancer, bladder cancer, cancer of the peritoneum, hepatocellular cancer, gastric or stomach cancer including gastrointestinal cancer,

gastrointestinal stromal tumors (GIST), pancreatic cancer, head and neck cancer, glioblastoma, retinoblastoma, astrocytoma, thecomas, arrhenoblastomas, hepatoma, hematologic malignancies including non-Hodgkins lymphoma (NHL), multiple myeloma and acute hematologic malignancies, endometrial or uterine carcinoma, endometriosis, fibrosarcomas, choriocarcinoma, salivary gland carcinoma, vulval cancer, thyroid cancer, esophageal carcinomas, hepatic carcinoma, anal carcinoma, penile carcinoma, nasopharyngeal carcinoma, laryngeal carcinomas, Kaposi's sarcoma, melanoma, skin carcinomas, Schwannoma, oligodendroglioma,

neuroblastomas, rhabdomyosarcoma, osteogenic sarcoma, leiomyosarcomas, urinary tract carcinomas, thyroid carcinomas, Wilm's tumor, as well as B-cell lymphoma, chronic

lymphocytic leukemia (CLL), acute lymphoblastic leukemia (ALL), Hairy cell leukemia, chronic myeloblastic leukemia, and post-transplant lymphoproliferative disorder (PTLD), as well as abnormal vascular proliferation associated with phakomatoses, edema (such as that associated with brain tumors), and Meigs' syndrome.

The term "sample" refers to a collection of cells obtained from a tissue of a subject or patient. The source of the sample may be solid tissue from a fresh, frozen and/or preserved organ or tissue sample or biopsy or aspirate; blood or any blood constituents; bodily fluids such as cerebral spinal fluid, amniotic fluid, peritoneal fluid, or interstitial fluid; cells from any time in gestation or development of the subject or plasma. The sample may also be primary or cultured cells or cell lines. In one embodiment, the sample is obtained from a cancerous tissue and/or organ. In one embodiment, the sample is a heterogeneous mixture of genetically diverse cells. In some embodiments, the sample includes compounds that are not naturally intermixed with the tissue in nature such as preservatives, anticoagulants, buffers, fixatives, nutrients, antibiotics, or the like. Samples may be obtained from a subject prior to commencement of treatment (i.e., from treatment naive patients) or after commencement of treatment (i.e., from treatment experienced patients). In one embodiment, treatment includes EGFR-therapy. In one embodiment, the sample is a clinical sample. In another embodiment, the sample is used in a diagnostic assay. In some embodiments, the sample is obtained from a primary or metastatic tumor. Tissue biopsy is often used to obtain a representative piece of tumor tissue. Alternatively, tumor cells can be obtained indirectly in the form of tissues or fluids that are known or thought to contain the tumor cells of interest. For instance, samples of lung cancer lesions may be obtained by resection, bronchoscopy, fine needle aspiration, bronchial brushings, or from sputum, pleural fluid or blood. In a more particular embodiment, the method includes a step of determining the presence or absence of one or more mutations in a sample obtained from a patient. In one embodiment, a method for selecting a treatment for cancer in a patient is provided, wherein the presence or absence of one or more mutations in a tissue sample obtained from a patient is determined, and, if one or more of the mutations are present, treating the patient with an anti-cancer therapy tailored to treat cancer cells having one or more of the mutations. In one embodiment,

"tailoring" the anti-cancer therapy means that one or more anti-cancer therapies are selected based on the genetic composition of cells or a subset of cells present in the tissue sample. In one embodiment, "tailoring" the anti-cancer therapy means selecting an anti-cancer therapy that includes at least one non-EGFR therapeutic agent. In another embodiment, a method for predicting whether or not at least some cancer cells in a patient will be resistant to a selected anti-cancer therapy is provided. In a more particular embodiment, the method includes predicting whether or not at least some cancer cells in a patient will be resistant to EGFR- therapy. In another embodiment, a method for distinguishing between EGFR therapy responsive and EGFR therapy non-responsive cancer cells is provided, in which the presence or absence of one or more mutations indicative of EGFR-therapy resistance is determined. In another embodiment, a method of predicting whether a cancer patient will be responsive or

nonresponsive to anti-cancer therapy that includes at least one EGFR-inhibitor is provided, wherein the presence or absence of one or more mutations indicative of EGFR-therapy resistance in a tissue sample obtained from the patient is determined, wherein the presence of one or more mutations indicative of EGFR-therapy resistance indicates that the patient may be nonresponsive to treatment with an EGFR inhibitor.

In one embodiment, a method for treating non-small cell lung cancer (NSCLC) in a patient is provided, which includes a step of determining whether the patient has NSCLC that includes at least a subset of cells having one or more of the mutations described herein, and if the patient has NSCLC that includes one or more of the mutations, administering to the patient an effective amount of an anti-cancer therapeutic tailored to treat cancer cells having one or more of the mutations. In a more particular embodiment, a method for treating non-small cell lung cancer (NSCLC) in a patient is provided, which includes a step of determining whether the patient has NSCLC that includes at least a subset of cells having one or more mutations indicative of EGFR-therapy resistance, and if the patient has NSCLC characterized by one or more mutations indicative of EGFR-therapy resistance, administering to the patient an effective amount of a non-EGFR anti-cancer therapeutic. In one embodiment, NSCLC is characterized by one or more low frequency mutations indicative of EGFR-therapy resistance.

In one embodiment, a method for treating colorectal cancer (CRC) in a patient is provided, which includes a step of determining whether the patient has CRC that includes at least a subset of cells having one or more of the mutations described herein, and if the patient has CRC that includes one or more of the mutations, administering to the patient an effective amount of an anti-cancer therapeutic tailored to treat cancer cells having one or more of the mutations. In another embodiment, a method for treating colorectal cancer (CRC) in a patient is provided, which includes a step of determining whether the patient has CRC that includes at least a subset of cells having one or more mutations indicative of EGFR-therapy resistance, and if the patient has CRC characterized by one or more mutations indicative of EGFR-therapy resistance, administering to the patient an effective amount of a non-EGFR anti-cancer therapeutic. In one embodiment, CRC is characterized by one or more low frequency mutations indicative of EGFR- therapy resistance.

Mutations in the RAS/MAPK pathway

In one embodiment, the method includes a step of determining whether one or more mutations are present in a tissue sample from a patient. In one embodiment, the method includes a step of determining whether one or more mutations are present in at least a subset of cells in a tissue sample from a patient. In one embodiment, the method includes a step of determining whether one or more low frequency mutations are present in a tissue sample from a patient. In one embodiment, the method includes a step of determining whether one or more mutations in a gene associated with cell cycle regulation, differentiation, growth and senescence are present in a tissue sample. In one embodiment, the method includes a step of determining whether one or more mutations indicative of resistance to EGFR-therapy are present in at least a subset of tumor cells in a patient.

As described in more detail below, mutations in genes associated with the RAS/MAPK pathway are often associated with cancer and often result in a poor prognosis or therapy resistance. The RAS/MAPK pathway plays a key role in cell cycle regulation, differentiation, growth and senescence, each of which are important for normal development. Not surprisingly, dysregulation of this pathway has profound effects on development and is one of the primary causes of cancer. The pathway involves many proteins, including epidermal growth factor receptor (EGFR), and Kirsten rat sarcoma viral oncogene (KRAS), which interact with many other proteins associated with tumor development, such as anaplastic lymphoma receptor tyrosine kinase (ALK) and hepatocyte growth factor receptor (HGFR).

In one embodiment, the method includes determining whether one or more mutations in genes associated with cell cycle regulation, differentiation, growth and senescence are present in at least a subset of cells in a tissue sample from a patient. In one embodiment, the method includes determining whether one or more mutations in genes associated with the RAS/MAPK pathway are present in a tissue sample from a patient. In one embodiment, the method includes determining whether one or more mutations included in Tables 1-4 are present in a tissue sample from a patient. In one embodiment, one or more mutations in the RAS/MAPK pathway include low frequency mutations. In a more particular embodiment, one or more mutations are indicative of EGFR-therapy resistance. The method can include detecting mutations in one or more of the genes or gene products described herein, alone or in combination. In one embodiment, NGS technologies are used to detect low frequency mutations in one or more of these genes. The tissue sample can be obtained from either treatment naive or treatment experienced patients.

EGFR

Epidermal growth factor receptor is an oncogene (SEQ ID NO:3) that encodes a single-pass type I membrane protein (EGFR/ERBB-1) (SEQ ID NO: 4) which functions as a cell surface tyrosine kinase receptor for members of the epidermal growth factor family (EGF family). Amino acid residues 1-24 make up the signal peptide, leaving amino acids residues 25- 1210 in the full protein. The mature protein includes an extracellular domain (aa 25-645), a transmembrane domain (aa 646-668), and a cytoplasmic domain (aa 669-1210). Located within the cytoplasmic domain is a kinase domain (aa 712-979).

EGFR regulates multiple cellular processes, including proliferation, differentiation, survival, motility, and blood vessel formation. Genetic alterations in the sequence of the EGFR gene or aberrations in the levels of protein expression may result in deregulation of EGFR function. Often abnormal EGFR activity is associated with cancer. In particular, EGFR is often overexpressed or demonstrates increased activity in many tumor types, including colorectal cancer (CRC), squamous cell carcinoma of the head and neck (SCCHN), and non-small-cell lung cancer (NSCLC). Overexpression of EGFR has been correlated with poor outcome and impacts all aspects of carcinogenesis, including cell growth and invasion, angiogenesis, and metastasis. In addition, the presence of activating mutations of EGFR is common in malignant cells and correlates with neoplastic progression.

The identification of EGFR as an oncogene has resulted in the development of targeted anticancer therapeutics directed against EGFR specific antigens. Targeted therapeutics directed towards EGFR are referred to herein as "EGFR therapeutics" or "EGFR inhibitors." In one embodiment, EGFR therapeutics include, but are not limited to small molecule tyrosine kinase inhibitors such as gefitnib and erlotinib for the treatment of non-small cell lung cancer (NSCLC); and cetuximab and panitumumab, monoclonal antibody inhibitors ("anti-EGFR therapeutics") that can block the extracellular ligand binding domain of EGFR, for treatment of colorectal cancer (CRC), including metastatic colorectal cancer (mCRC). Other monoclonal antibody EGFR inhibitors in clinical development include zalutumumab, nimotuzumab, and matuzumab.

In one embodiment, the method includes detecting the presence or absence of one or more EGFR mutations in a tissue sample. In one embodiment, the tissue sample includes a heterogeneous mixture of genetically diverse cells. In a more particular embodiment, the method includes detecting the presence or absence of one or more low frequency EGFR mutations. In one embodiment, one or more mutations are indicative of EGFR-therapy resistance. In one embodiment, the method includes detecting the presence or absence of one or more low frequency mutations that are indicative of EGFR-therapy resistance.

In one embodiment, one or more mutations include one or more of the 11 non- synonymous single nucleotide changes for EGFR listed in Table 2. In one embodiment, one or more mutations in EGFR include: L76P, V374A, N413T, R494G, S498G, N603T, V651 A,

K714R, G719A, Y1016S, L1034R, and combinations thereof. In one embodiment, one or more mutation is located in the extracellular domain (aa 25-645) of EGFR. In a more particular embodiment, one or more mutations located in the extracellular domain of EGFR are selected from L76P, V374A, N413T, R494G, S498G, N603T, and combinations thereof. In one embodiment, one or more mutations located in the extracellular domain of EGFR are indicative of EGFR-therapy resistance. In one embodiment, one or more mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more mutations have a frequency of at least about 0.1%, 0.5%> or 1%.

In another embodiment, one or more mutations are located in the transmembrane domain (aa 646-668) of EGFR. In a more particular embodiment, one or more mutations located in the transmembrane domain of EGFR include V651 A. In one embodiment, one or more mutations located in the transmembrane domain of EGFR are indicative of EGFR-therapy resistance. In one embodiment, one or more mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more mutations have a frequency of at least about 0.1%, 0.5% or 1%.

In another embodiment, one or more mutations are located in the cytoplasmic domain (aa 669-1210) of EGFR. In a more particular embodiment, one or more mutations located in the cytoplasmic domain of EGFR are selected from K714R, G719A, Y1016S, L1034R, and combinations thereof. In one embodiment, one or more mutations located in the cytoplasmic domain of EGFR are indicative of EGFR-therapy resistance. In one embodiment, one or more mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more mutations have a frequency of at least about 0.1%, 0.5% or 1%.

In another embodiment, one or more mutations are located in the kinase domain of EGFR

(aa 712-979). In a more particular embodiment, one or more mutations located in the kinase domain of EGFR are selected from K714R, G719A, and combinations thereof. In one embodiment, one or more mutations located in the kinase domain of EGFR are indicative of EGFR-therapy resistance. In one embodiment, the mutation in the kinase domain of EGFR includes G719A. In one embodiment, one or more mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more mutations have a frequency of at least about 0.1 %, 0.5%> or 1%.

In one embodiment, one or more mutations include one or more of the 8 non- synonymous mutations identified in EGFR listed in Table 3. In one embodiment, one or more EGFR mutations include E66*, S170R, L173P, N182S, C236R, F420L, I780V, R832C, and combinations thereof. In one embodiment, one or more mutations are located in an N-terminal region of EGFR. In one embodiment, one or more mutations located in an N-terminal region of EGFR include a heterozygous stop codon mutation. In one embodiment, one or more mutations located in an N-terminal region of EGFR include E66*. In one embodiment, one or more mutations located in the N-terminal region of EGFR are indicative of EGFR therapy resistance. In one embodiment, one or more mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more mutations have a frequency of at least about 0.1%, 0.5% or 1%.

In another embodiment, one or more mutations are located in the cytoplasmic domain (aa

669-1210) of EGFR. In a more particular embodiment, one or more mutations located in the cytoplasmic domain of EGFR are selected from I780V, R832C, and combinations thereof. In one embodiment, one or more mutations located in the cytoplasmic domain of EGFR are indicative of EGFR-therapy resistance. In one embodiment, one or more mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more mutations have a frequency of at least about 0.1%, 0.5% or 1%.

In another embodiment, one or more EGFR mutations are located in a kinase domain of EGFR (aa 712-979). In one embodiment, one or more mutations located in a kinase domain of EGFR include a homozygous deletion mutation. In one embodiment, one or more mutations located in a kinase domain of EGFR include I780V, R832C, and combinations thereof. In one embodiment, one or more mutations located in a kinase domain of EGFR are indicative of EGFR therapy resistance. In one embodiment, one or more mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more mutations have a frequency of at least about 0.1 %, 0.5%> or 1%.

In another embodiment, one or more mutations include one or more of the 10 non- synonymous mutations located in EGFR shown in Table 4. In one embodiment, one or more EGFR mutations include T43A, N94D, V336M, N413T, L443P, V461A, C628R, A840T, L1034R, Al 158P, and combinations thereof. In one embodiment, one or more mutations are located in the extracellular domain (aa 25-645) of EGFR. In one embodiment, one or more mutations located in the extracellular domain of EGFR are selected from T43A, N94D, V336M, N413T, L443P, V461 A, C628R, and combinations thereof. In one embodiment, one or more mutations in the extracellular domain are indicative of EGFR-therapy resistance. In one embodiment, one or more mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more mutations have a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, one or more mutations are located in the cytoplasmic domain (aa 669-1210) of EGFR. In one embodiment, one or more mutations located in the cytoplasmic domain of EGFR are selected from A840T, L1034R, Al 158P, and combinations thereof. In one embodiment, one or more mutations are located in the kinase domain (aa 712-979) of EGFR. In one embodiment, one or more mutations located in the kinase domain include A840T. In one embodiment, one or more mutations in the cytoplasmic domain and/or kinase domain are indicative of EGFR-therapy resistance. In one embodiment, one or more mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more mutations have a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, one or more mutations in EGFR are selected from: L76P, V374A, N413T, R494G, S498G, N603T, V651A, K714R, G719A, Y1016S, L1034R, E66*, S170R, L173P, N182S, C236R, F420L, I780V, R832C, T43A, N94D, V336M, N413T, L443P, V461A, C628R, A840T, L1034R, Al 158P and combinations thereof. In another embodiment, the tissue sample includes one or more one or more mutations in EGFR are selected from: L76P, V374A, N413T, R494G, S498G, N603T, V651A, K714R, G719A, Y1016S, L1034R, E66*, S170R, L173P, N182S, C236R, F420L, I780V, R832C, N94D, V336M, N413T, L443P, V461A, C628R, A840T, L1034R, Al 158P in combination with one or more mutations in KRAS, ALK, MET. In one embodiment, the method includes detecting the presence or absence of one or more EGFR kinase domain mutations in combination with one or more low frequency K-RAS mutations. In one embodiment, the method includes detecting the presence or absence of one or more low frequency EGFR kinase domain mutations in combination with one or more low frequency K- RAS active domain mutations. In one embodiment, the EGFR kinase domain mutation includes G719A. In one embodiment, the K-RAS active domain mutation includes a point mutation at G12 and/or G13. In one embodiment, the K-RAS active domain mutation includes G12R. In a more particular embodiment, the K-RAS active domain mutation includes G12R and the EGFR kinase domain mutation includes G719A. In another embodiment, the method includes detecting the presence or absence of one or more low frequency mutations in K-RAS and EGFR, in combination with MET and ALK gene copy number increases. "Gene copy number increases" refers to an alteration in the DNA of a genome which results in the cell having an increased number of one or more genes, for example, wherein a region of the genome has been duplicated. In one embodiment, the term "gene copy number increases" refers to an average copy number of greater than 2, 3, 4, 5 or more.

KRAS

Kirsten rat sarcoma viral oncogene (KRAS2) (SEQ ID NO: 1) is a member of the mammalian RAS gene family. Alternative splicing of the oncogene leads to variants encoding two isoforms (KRASA and KRASB) that differ in the C-terminal region. The major splice variant (KRAS4B) a 188 amino acid GTPase protein (SEQ ID NO: 2) involved in many signal transduction pathways. The protein includes multiple coding exons (four coding exons and a 5' non-coding exon) and three GTP nucleotide binding domains (aa 10-17; aa 57-61; and aa 116- 119). KRASl is a pseudogene derived from processed KRAS2 mRNA.

KRAS is an important driver in the RAS/MAPK pathway that links EGFR activation to cell proliferation and survival and KRAS mutations have been implicated in the development of many cancers. Certain KRAS mutations can impact efficacy of EGFR therapy in patients with colorectal cancer (CRC) (Van Cutsem, et al. (2007) Open-label phase III trial of panitumumab plus best supportive care compared with best supportive care in patients with chemotherapy- refractory metastatic colorectal cancer. J. Clin. Oncol. 25: 1658-64; Jackman et al. (2009) Impact of epidermal growth factor receptor and KRAS mutations on clinical outcomes in previously untreated non- small cell lung cancer patients: Results of an online tumor registry of clinical trials. Clin. Cancer. Res. 15:5267-5273; Pao et al. (2005) KRAS mutations and primary resistance of lung adenocarcinomas to gefitinib or erlotinib. PLoS. Med. 2(l):el7; Eberhard et al. (2005) Mutations in the epidermal growth factor receptor and in KRAS are predictive and prognostic indicators in patients with non-small-cell lung cancer treated with chemotherapy alone and in combination with erlotinib. J. Clin. Oncol. 23:5900-5909; and Massarelli et al. (2007) KRAS mutation is an important predictor of resistance to therapy with epidermal growth factor receptor tyrosine kinase inhibitors in non-small cell lung cancer. Clin. Cancer. Res.

13:2890-2896). Mutant KRAS is found in about 35%-45% of CRCs, with frequent mutations that result in over-activation of the RAS/MAPK pathway that can render upstream inhibition by EGFR-therapeutics ineffective. G12 and G13 mutations account for about 95% of all mutation types, with approximately 80% occurring in codon 12 and 15% in codon 13. (Tan and Du.

(2012) KRAS mutation testing in metastatic colorectal cancer. World. J. Gastroenterol.

18(37):5171-5180). Other mutations in KRAS codons 61, 146 and 154 occur less frequently in CRC, accounting for 5% of all mutation type. Many reported KRAS mutations are single nucleotide point mutations, and often include G12D, G12A, G12R, G12C, G12S, G12V and G13D. KRAS mutations are also associated with reduced efficacy of EGFR-TKIs in non-small cell lung cancer (NSCLC) and CRC (Mao et al. (2010) KRAS mutations and resistance to EGFR-TKIs treatment in patients with non-small cell lung cancer: a meta-analysis of 22 studies. Lung Cancer. 69(3):272-8). KRAS mutations may be present in tissue samples from treatment naive patients. (Diaz et al. (2012) The molecular evolution of acquired resistance to targeted EGFR blockade in colorectal cancers. Nature. 486:537-40).

In one embodiment, one or more mutations are located in KRAS. In one embodiment, one or more mutations include one or more of the mutations within a first nucleotide binding domain of KRAS (amino acid residues 10-17) listed in Table 1. In one embodiment, one or more mutations in KRAS include: Al IV, G12D, G12S, G12R, G12V, G12C, G13D, V14I, and combinations thereof. In one embodiment, one or more mutations in the first nucleotide binding domain of KRAS are indicative of EGFR-therapy resistance. In one embodiment, one or more KRAS mutations are detected in a tissue sample from a treatment naive patient. In one embodiment, one or more KRAS mutations have a frequency of less than about 10%, 5% or 1%. In another embodiment, one or more KRAS mutations have a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, one or more mutations include the KRAS mutation shown in Table 2. In one embodiment, the mutation in KRAS includes G12R. In one embodiment, the G12R mutation is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5%> or 1%.

In one embodiment, one or more mutations include one or more KRAS mutations shown in Table 4. In one embodiment, one or more mutations include a mutation in a C-terminal region of KRAS. In one embodiment, the mutation in a C-terminal region of KRAS includes R164*, E153G, or combinations thereof. In one embodiment, R164* is a stop codon mutation that results in a truncated KRAS protein. In one embodiment, the mutation in a C-terminal region of KRAS is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, the tissue sample includes one or more mutations in KRAS in combination with one of more mutations in EGFR, ALK, MET or combinations thereof. In one embodiment, the tissue sample includes one or more mutations in a KRAS nucleotide binding domain in combination with one of more mutations in EGFR, ALK, MET or combinations thereof. In one embodiment, the tissue sample includes one or more mutations in a KRAS selected from Al l V, G12D, G12S, G12R, G12V, G12C, G13D, V14I, R164*, E153G, and combinations thereof, in combination with one or more mutations in EGFR, ALK, MET or combinations thereof. In one embodiment, one or more KRAS mutations selected from Al 1 V, G12D, G12S, G12R, G12V, G12C, G13D, V14I, R164*, E153G, or combinations thereof is indicative of EGFR-therapy resistance. In one embodiment, one or more KRAS mutations selected from Al l V, G12D, G12S, G12R, G12V, G12C, G13D, V14I, R164*, E153G, and combinations thereof, in combination with one or more mutations in EGFR, ALK and/or MET is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

ALK

Anaplastic lymphoma receptor tyrosine kinase is an oncogene (SEQ ID NO: 5) that encodes a 1620 amino acid transmembrane protein (SEQ ID NO:6) which belongs to the insulin receptor superfamily. Amino acid residues 1-18 make up the signal peptide, leaving amino acid residues 19-1620 in the mature protein, which includes an extracellular domain (aa 19-1038), an hydrophobic single pass transmembrane domain (aa 1039-1059), and a cytoplasmic domain (aa 1060-1620), which includes an intracellular kinase domain (aa 1116-1392). ALK plays an important role in the development of the brain and exerts its effects on specific neurons in the nervous system. ALK has been found to be rearranged, mutated, or amplified in a series of tumors including anaplastic large cell lymphomas, neuroblastoma, and non-small cell lung cancer (NSCLC).

In one embodiment, a tissue sample obtained from a patient includes one or more mutations located in ALK. In one embodiment, one or more mutations include one or more ALK mutations shown in Table 2. In one embodiment, the tissue sample includes one or more mutations in a cytoplasmic domain of ALK, including, but not limited to G1548E. In one embodiment, the ALK mutation is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, one or more mutations include one or more ALK mutations shown in Table 3. In one embodiment, the tissue sample includes one or more mutations in an

extracellular domain (aa 19-1038) of ALK. In a more particular embodiment, the mutation in an extracellular domain of ALK includes H368P, F241L, V198A, L170P, or a combination thereof. In one embodiment, the mutation in an extracellular domain of ALK is indicative of EGFR- therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, one or more mutations include one or more ALK mutations shown in Table 4. In one embodiment, the tissue sample includes one or more mutations in ALK selected from S1611G, Rl 181C, D460G, S329F, and combinations thereof. In one embodiment, the mutation in ALK is indicative of EGFR-therapy resistance. In one embodiment, one or more ALK mutations are in a kinase domain of ALK (aa 1116-1392). In one embodiment, one or more ALK mutations include Rl 181C. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, the tissue sample includes one or more mutations in a cytoplasmic domain of ALK selected from S1611G, R1181C, and combinations thereof. In one embodiment, the mutation in the cytoplasmic domain of ALK is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, the tissue sample includes one or more mutations in an extracellular domain of ALK selected from D460G, S329F, and combinations thereof. In one embodiment, the mutation in the extracellular domain of ALK is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, a tissue sample includes one or more mutations located in ALK selected from: G1548E, H368P, F241L, V198A, L170P, S1611G, R1181C, D460G, S329F, or combinations thereof. In one embodiment, a tissue sample includes an ALK mutation that includes an ALK gene copy number increase. In one embodiment, the tissue sample includes at least 2, 3, 4, 5 or more copies of ALK. In one embodiment, the ALK mutation is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, a tissue sample includes one or more mutations located in ALK selected from G1548E, H368P, F241L, V198A, L170P, S1611G, R1181C, D460G, S329F, or combinations thereof, and/or an ALK gene copy number increase, in combination with one or more mutations in EGFR, KRAS, MET or combinations thereof. In one embodiment, the presence of one or more mutations located in ALK selected from G1548E, H368P, F241L, V198A, L170P, S161 IG, Rl 181C, D460G, S329F, or combinations thereof, in combination with one or more mutations in EGFR, KRAS, MET, or combinations thereof and/or an ALK gene copy number increase is indicative of EGFR-therapy resistance. In one embodiment, the presence of one or more low frequency mutations located in ALK selected from G1548E, H368P, F241L, V198A, L170P, S1611G, R1181C, D460G, S329F, or combinations thereof, and/or an ALK gene copy number increase in combination with one or more low frequency mutations in EGFR, KRAS, MET, or combinations thereof is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%. ALK/EML4 fusions

Chromosomal rearrangements are common with ALK and result in multiple fusion genes involved in tumorigenesis, including ALK/EML4 (echinoderm microtubule-associated protein-like 4-anaplastic lymphoma kinase) fusion proteins. Two variants for ALK/EML4 fusions are known: variant a (SEQ ID NO: 9) and variant b (SEQ ID NO: 10). The presence of the EML4-ALK fusion gene (SEQ ID NOS:7 and 8) has been identified as a driver mutation in a subgroup of NSCLC patients (Soda et al. (2007) Identification of the transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature. 448:561-566). A standard test used to detect the EML4-ALK gene fusion in tumor samples is fluorescence in situ hybridization (FISH), although other techniques such as immunohistochemistry (IHC) and reverse-transcriptase PCR (RT-PCR) can also be used.

The clinical features of lung cancer patients that harbor EML4-ALK include light- or never-smokers, younger age, adenocarcinomas with acinar pattern or signet ring

adenocarcinoma, and a lack of EGFR or KRAS mutations (Shaw et al. (2009) Clinical features and outcome of patients with non-small-cell lung cancer who harbor EML4-ALK. J. Clin. Oncol. 27:4247-4253). Crizotinib (XALKORI ®) has been approved for use in patients harboring this variant (Kwak et al. (2010) Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N. Engl. J. Med. 363: 1693-1703). The average frequency of this gene fusion across different ethnic population is reported to be approximately 3.4% in unselected NSCLC patients and 4.5% in adeno carcinoma-enriched NSCLC patients (Bang YJ. (2011) The potential for crizotinib in non-small cell lung cancer: a perspective review. Ther Adv. Med. Oncol. 3:279-91) with a higher frequency in the Chinese population (4.9%>~11.7%>) in unselected NSCLC patients and 5.3%~16.1% in adenocarcinoma-enriched NSCLC patients (Wong et al. (2009) The EML4- ALK fusion gene is involved in various histologic types of lung cancers from nonsmokers with wild-type EGFR and KRAS. Cancer. 115: 1723-33; and Zhang et al. (2010) Fusion of EML4 and ALK is associated with development of lung adenocarcinomas lacking EGFR and KRAS mutations and is correlated with ALK expression. Mol. Cancer. 9: 188).

Approximately 1.0% of NSCLC patients have concomitant EGFR mutations and EML4-ALK fusions. Concurrent KRAS mutations and EML4 ALK fusion have been observed in 12.5% of Asian NSCLC patients using Sanger sequencing and RT-PCR, respectively

(Shaozhang et al. (2012) Detection of EML4-ALK fusion genes in non-small cell lung cancer patients with clinical features associated with EGFR mutations. Genes Chromosomes Cancer. 51 :925-32).

In one embodiment, the tissue sample includes an EML4-ALK fusion. In one embodiment, the tissue sample includes an EML4 ALK fusion in combination with one or more mutations in ALK, EGFR, and MET, including, but not limited to one or more mutations described herein. In another embodiment, the tissue sample includes one or more low frequency mutations in KRAS, EGFR, and MET, including but not limited to, one or more mutations described herein, that co-exist with an ALK copy number increase. In one embodiment, the sample includes a mutation in a kinase domain of ALK (aa 1116-1392). In a more particular embodiment, the mutation in a kinase domain of ALK includes G1548E, Rl 181C, or a combination thereof. In one embodiment, the sample includes a mutation in a cytoplasmic domain of ALK (aa 1116-1392). In a more particular embodiment, the mutation in a

cytoplasmic domain of ALK includes G1548E, Rl 181C, or a combination thereof. In another embodiment, the tissue sample includes one or more mutations in an extracellular domain (aa 19- 1038) of ALK selected from H368P, F241L, V198A, L170P, and combinations thereof. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

MET

Met proto-oncogene (MET) (SEQ ID NO: l 1) encodes a 1,390 amino acid receptor tyrosine kinase known as hepatocyte growth factor receptor (HGFR) (SEQ ID NO: 12). Amino acids 1-24 make up the signal peptide, leaving amino acids 25-1390 in the mature protein, which includes an extracellular domain (aa 25-932), a transmembrane domain (aa 933-955) and a cytoplasmic domain (aa 956-1390). The extracellular domain also includes a sema domain (aa 27-515). The primary single chain precursor protein is post-translationally cleaved to produce alpha and beta subunits, which are disulfide linked to form a mature receptor. HGFR transduces signals from the extracellular matrix into the cytoplasm by binding to hepatocyte growth factor (HGF) and activating ras signaling. HGFR regulates physiological process such as proliferation, scattering, morphogenesis and survival. MET is deregulated in many types of human

malignancies, including cancers of kidney, liver, stomach, breast, and brain. Amplification of MET can bypass EGFR to activate downstream signaling in the cell.

In one embodiment, a tissue sample from a patient includes one or more mutations in MET. In one embodiment, one or more mutations in MET include one or more of the mutations listed in Table 2. In one embodiment, a tissue sample from a patient includes one or more mutations in MET selected from a mutation in a cytoplasmic domain (aa 956-1390) of MET. In one embodiment, the mutation in a cytoplasmic domain of MET includes L1053P. In another embodiment, the tissue sample includes one or more mutations in an extracellular domain (aa 25- 932) of MET. In one embodiment, the mutation in an extracellular domain of MET includes K864R. In one embodiment, the mutation in MET is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, one or more mutations in MET includes the mutation listed in Table 3. In one embodiment, a tissue sample from a patient includes one or more mutations in an extracellular domain (aa 25-932) of MET. In one embodiment, one or more mutations in an extracellular domain of MET include C541G. In one embodiment, the mutation in an

extracellular domain of MET is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, the tissue sample from the patient includes one or more MET mutations listed in Table 4. In one embodiment, one or more MET mutations include: A320V, D340G, R739C, K864R, I868V, K1262R, or a combination thereof. In one embodiment, one or more mutations are located in the extracellular domain (aa 25-932) of MET. In one embodiment, one or more mutations located in the extracellular domain of MET are selected from A320V, D340G, R739C, K864R, I868V, or a combination thereof. In one embodiment, one or more mutations are located in the cytoplasmic domain (aa 956-1390) of MET. In one embodiment one or more mutations located in the cytoplasmic domain of MET include K1262R. In one embodiment, the mutation in MET is indicative of EGFR-therapy resistance. In one

embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, a tissue sample from a patient includes one or more mutations in MET selected from: K864R, L1053P, C541G, A320V, D340G, R739C, K864R, I868V,

K1262R, or combinations thereof. In one embodiment, a tissue sample includes an MET mutation that includes an MET gene copy number increase. In one embodiment, the tissue sample includes a mutation that includes at least 2, 3, 4, 5 or more copies of MET. In one embodiment, the MET mutation is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%.

In one embodiment, a tissue sample from a patient includes one or more mutations in

MET selected from K864R, L1053P, C541G, A320V, D340G, R739C, K864R, I868V, K1262R, or combinations thereof, and/or a MET gene copy number increase, in combination with one or more mutations in EGFR, KRAS, ALK, or combinations thereof. In one embodiment, a tissue sample from a patient includes one or more low frequency mutations in MET selected from K864R, L1053P, C541G, A320V, D340G, R739C, K864R, I868V, K1262R, or combinations thereof, and/or a MET gene copy number increase, in combination with one or more low frequency mutations in EGFR, KRAS, ALK, or combinations thereof. In one embodiment, the presence of one or more mutations in MET selected from K864R, L1053P, C541G, A320V, D340G, R739C, K864R, I868V, K1262R, or combinations thereof, and/or a MET gene copy number increase, in combination with one or more mutations in EGFR, KRAS, ALK, or combinations thereof in a tissue sample from a patient is indicative of EGFR-therapy resistance. In one embodiment, the mutation is detected in a tissue sample from a treatment naive patient. In one embodiment, the mutation has a frequency of less than about 10%, 5% or 1%. In another embodiment, the mutation has a frequency of at least about 0.1%, 0.5% or 1%. Pharmaceutical Compositions

Once it is determined whether or not one or more mutations described herein are present in at least a subset of cells in a tumor or tissue sample from a patient, an appropriate therapeutic regimen can be developed for the patient. In one embodiment, the presence or absence of one or more mutations in at least a subset of cells in a tumor or tissue sample from a patient can be used to predict whether or not a particular therapeutic regimen will achieve a desired clinical outcome. In one embodiment, an anti-cancer therapy can be specifically tailored to treat cancer cells having one or more of the mutations described herein. For example, if one or more mutations indicative of EGFR-therapy resistance are present in at least a subset of cells in a tissue sample, a therapeutic regimen that includes one or more non-EGFR therapy anti-cancer agents can be implemented. For example, a tumor that includes at least a subset of cells in which one or more mutations indicative of EGFR-therapy resistance is detected can be treated with a variety of known antineoplastic agents. Since the mutations indicative of EGFR-therapy resistance may only be present in a low frequency (i.e., less than about 10%>, 5%, or 1%), it may be desirable to treat a tumor in which one or more low frequency mutations indicative of EGFR-therapy resistance are detected with a combination therapy that includes both EGFR-therapy and non- EGFR therapy. In one embodiment, the pharmaceutical composition includes a therapeutically effective amount of a non-EGFR anti-cancer therapeutic agent, and a pharmaceutically acceptable carrier. In another embodiment, the pharmaceutical composition includes a therapeutically effective amount of an EGFR-therapeutic and a non-EGFR anti-cancer therapeutic, in combination with a pharmaceutically acceptable carrier.

The term "effective amount" or "therapeutically effective amount" refers to an amount that will elicit a biological or medical response of a tissue, system, animal or human that is being sought by the researcher, veterinarian, medical doctor or other clinician. In the case of cancer, the effective amount of the drug may reduce the number of cancer cells; reduce the tumor size; inhibit (i.e., slow to some extent and typically stop) cancer cell infiltration into peripheral organs; inhibit (i.e., slow to some extent and typically stop) tumor metastasis; inhibit, to some extent, tumor growth; allow for treatment of the resistant tumor, and/or relieve to some extent one or more of the symptoms associated with the disorder. To the extent the drug may prevent growth and/or kill existing cancer cells, it may be cytostatic and/or cytotoxic.

The term "pharmaceutically acceptable carriers" refers to carriers, excipients, or stabilizers which are nontoxic to the cell or mammal being exposed thereto at the dosages and concentrations employed. Often the physiologically acceptable carrier is an aqueous pH buffered solution. Examples of physiologically acceptable carriers include buffers such as phosphate, citrate, and other organic acids; antioxidants including ascorbic acid; low molecular weight (less than about 10 residues) polypeptide; proteins, such as serum albumin, gelatin, or

immunoglobulins; hydrophilic polymers such as polyvinylpyrrolidone; amino acids such as glycine, glutamine, asparagine, arginine or lysine; monosaccharides, disaccharides, and other carbohydrates including glucose, mannose, or dextrins; chelating agents such as EDTA; sugar alcohols such as mannitol or sorbitol; salt-forming counterions such as sodium; and/or nonionic surfactants such as TWEEN®, polyethylene glycol (PEG), and PLURONICS®.

The terms "anti-cancer therapeutic" or "antineoplastic agent" are used interchangeably herein to refer to a composition useful in treating cancer. Antineoplastic agents include, but are not limited to, chemotherapeutic agents, growth inhibitory agents, cytotoxic agents, radiation therapy, anti-angiogenic agents, apoptotic agents, anti-tubulin agents, neutralizing antibodies, tyrosine kinase inhibitors, antimetabolites; hormones; platelet derived growth factor inhibitors, COX-2 inhibitors, interferons, cytokines, organic chemical agents, antisense reagents, and combinations thereof.

Chemotherapeutic agents include, but are not limited to: alkylating agents,

antimetabolites, antitumor antibiotics, antimitotics, topoisomerase inhibitors, proteasome inhibitors, tyrosine kinase inhibitors, and combinations thereof. More particularly,

chemotherapeutic agents can include: alkylating agents such as cisplatin, carboplatin, oxaliplatin, cyclophosphamide, nitrogen mustard, melphalan, chlorambucil, busulphan and nitrosourea; antimetabolites such as antifolates, raltitrexed, gemcitabine, capecitabine, methotrexate, pemetrexed (Alimta), cytosine arabinoside and hydroxyurea; antitumor antibiotics such as anthracyclines, bleomycin, doxorubicin, daunomycin, epirubicin, idarubicin, mitomycin-C, dactinomycin and mithramycin; antimitotic agents such as vinca alkaloids and taxoids;

topoisomerase inhibitors such as epipodophyllotoxins, irinotecan, amsacrine, topotecan and camptothecin; proteasome inhibitors; tyrosine kinase inhibitors such as Axitinib, Bosutinib, Cediranib, Dasatinib, Erlotinib, Gefitinib, Imatinib, Lapatinib, Lestaurtinib, Nilotinib,

Semaxanib, Sunitinib, and Vandetanib; antimetabolites such as pyrimidine and purine analogs, radioactive isotopes such as ²¹¹At, ¹³¹I, ¹²⁵I, ⁹⁰Y, ¹⁸⁶Re, ¹⁸⁸Re, ¹⁵³Sm, ²¹²Bi, ³²P and radioactive isotopes of Lu, and combinations thereof.

In one embodiment, the antineoplastic agent includes 5-Fluoruracil, 6-mercatopurine, Actinomycin, Adriamycin®, Admen®, Aminoglutethimide, Anastrozole, Aredia®, Arimidex®, Aromasin®, Bonefos®, Bleomycin, carboplatin, Cactinomycin, Capecitabine, Cisplatin, Clodronate, Cyclophosphamide, Cytadren®, Cytoxan®, Dactinomycin, Docetaxel, Doxyl®, Daunorubicin, Doxorubicin, Epirubicin, Etoposide (VP- 16), Exemestane, Femora®, Fluorouracil (5-FU), Fluoxymesterone, Halotestin®, Herceptin®, Letrozole, Leucovorin calcium, Megace®, Megestrol acetate, Methotrexate, Mitomycin, Mitoxantrone, Mutamycin®, Navelbine®,

Nolvadex®, Novantrone®, Oncovin®, Ostac®, Paclitaxel, Pamidronate, Pharmorubicin®, Platinol®, prednisone, Procytox®, Tamofen®, Tamone®, Tamoplex®, Tamoxifen, Taxol®, Taxotere®, Trastuzumab, Thiotepa, Velbe®, Vepesid®, Vinblastine, Vincristine, Vinorelbine, Xeloda®, or combinations thereof.

In one embodiment, anti-cancer therapy can include combination therapy. In one embodiment, an EGFR inhibitor is administered in combination with a non-EGFR anti-cancer therapeutic agent. In one embodiment, the EGFR inhibitor and the non-EGFR anti-cancer therapeutic agent are administered simultaneously. In one embodiment, the EGFR inhibitor and the non-EGFR anti-cancer therapeutic agent are administered sequentially. The EGFR inbhitor and non-EGFR anti-cancer therapeutic agent can be administered using any suitable

administration method, including, but not limited to, parenteral (e.g., intradermal, intramuscular, intraperitoneal, intravenous and subcutaneous), epidural, and mucosal (e.g., oral), intramusclarly, intravaneously or subcutaneously. The phrase administration "in combination with" includes simultaneous (concurrent) and consecutive administration in any order.

EGFR inhibitors include but are not limited to small molecule inhibitors, antibodies or antibody fragments, peptide or RNA aptamers, antisense constructs, small inhibitory RNAs (i.e. RNA interference by dsRNA; RNAi), and ribozymes. In one embodiment, the EGFR kinase inhibitor is a small organic molecule or an antibody that binds specifically to the human EGFR. EGFR kinase inhibitors include, for example quinazoline EGFR kinase inhibitors, pyrido- pyrimidine EGFR kinase inhibitors, pyrimido-pyrimidine EGFR kinase inhibitors, pyrrolo- pyrimidine EGFR kinase inhibitors, pyrazolo-pyrimidine EGFR kinase inhibitors, phenylamino- pyrimidine EGFR kinase inhibitors, oxindole EGFR kinase inhibitors, indolocarbazole EGFR kinase inhibitors, phthalazine EGFR kinase inhibitors, isoflavone EGFR kinase inhibitors, quinalone EGFR kinase inhibitors, and tyrphostin EGFR kinase inhibitors.

Other small molecule EGFR kinase inhibitors include [6,7-bis(2-methoxyethoxy)-4- quinazolin-4-yl]-(3-ethynylphenyl) amine (also known as OSI-774, erlotinib, or TARCEVA® (erlotinib HC1); OSI Pharmaceuticals/Genentech/Roche); CI-1033 (formerly known as

PD183805; Pfizer); PD-158780 (Pfizer); AG-1478 (University of California); CGP-59326 (Novartis); PKI-166 (Novartis); EKB-569 (Wyeth); GW-2016 (also known as GW-572016 or lapatinib ditosylate; GSK); and gefitinib (also known as ZD 1839 or IRESSA™ Astrazeneca)

EGFR inhibitors can also include multi-kinase inhibitors, i.e. inhibitors that inhibit EGFR kinase and one or more additional kinases, including, but not limited to CI-1033 (formerly known as PD183805; Pfizer); GW-2016 (also known as GW-572016 or lapatinib ditosylate; GSK); AG490 (a tyrphostin); ARRY334543 (Array BioPharma); BIBW-2992 (Boehringer Ingelheim Corp.); EKB-569 (Wyeth); ZD6474 (also known as ZACTIMA™; AstraZeneca Pharmaceuticals), and BMS-599626 (Bristol-Myers Squibb).

EGFR inhibitors also include anti-EGFR antibodies or fragments thereof that can partially or completely block EGFR activation by its natural ligand. Non-limiting examples of antibody-based EGFR kinase inhibitors include, but are not limited to, IMC-C225 (also known as cetuximab or ERBITUX™.; Imclone Systems), ABX-EGF (Abgenix), EMD 72000 (Merck KgaA, Darmstadt), RH3 (York Medical Bioscience Inc.), and MDX-447 (Medarex/Merck KgaA).

In one embodiment, the EGFR inhibitor is selected from cetuximab, panitumumab, erlotinib, gefitinib, and combinations thereof.

Kits

In one embodiment, a kit for identifying a patient with an EGFR-therapy resistant tumor is provided. In one embodiment, the kit includes: one or more reagents for determining a presence of one or more mutations indicative of EGFR-therapy resistance; and instructions for performing the assay. In other embodiments, the kit can also include a container for the reagents.

All publications, patents and patent applications mentioned in this specification are incorporated by reference into the specification to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated herein by reference.

EXAMPLES

The examples below are provided to illustrate the practice of this invention. They are not intended to limit or define the entire scope of this invention. The reagents employed in the examples are commercially available or can be prepared using commercially available instrumentation, methods, or reagents known in the art. Thus, although the invention has been described in some detail by way of illustration and examples for purposes of clarity of understanding, those of ordinary skill in the art will realize readily that many changes and modifications can be made thereto without departing from the spirit or scope of the appended claims.

CRC and NSCLC samples

Sixty-one fresh frozen colon rectum tumors were purchased from ILSbio (Chestertown, MD) and Asterand (Detroit, MI). All the tumors are from treatment naive patients diagnosed with stage I to C3 adenocarcinoma of colon or rectum. All the patients were Caucasian with 28 females and 33 males, from 32-84 years old. Six specimens were sequenced in at least two independent sequencing runs, with two replicated in three (i.e. technical replicates).

NSCLC samples were collected in 3 treatment naive patients from Shanghai Chest Hospital. For Patients A and C, two independent biological specimens were procured and each was replicated in two independent sequencing runs (i.e. two technical replicates for each biological specimen). For Patient B, there was only a single biological specimen procured and this was not replicated by sequencing. The study protocol and all NSCLC specimens were approved by the ethics committee at Shanghai Chest Hospital.

DNA extraction, library preparation, and sequencing

DNA extractions from fresh frozen tissue were performed using the Qiagen DNeasy

Blood and Tissue Kit (Qiagen, Germantown, MD) using the manufacturer's suggested protocol. DNA extractions from FFPE tissue were performed using the Ambion RecoverAU Total Nucleic Acid Isolation Kit (Life technologies, Grand Island, NY) following the manufacturer's instructions. Target sequencing libraries were prepared using Fluidigm Access Array System (Fluidigm, CA) using multiplexed sample barcoding and amplicon tagging following

manufacture's instruction.

The sequencing libraries were normalized, and run on an Illumina Hi Seq 2000

Instrument (San Diego, CA) using a 100 bp paired end sequencing read protocol.

FISH

EML4-ALK fusion was detected by fluorescent in situ hybridization (FISH) assays on 4 micron formalin-fixed paraffin-embedded (FFPE) section. FISH probes were generated internally by directly labeling BAC (RP 1 1 - 100C 1 , Invitrogen, USA) DNA with Spectrum Red (ENZO, USA, cat no. 02N34-050) for ALK gene and BAC (RP1 1-142M12, Invitrogen, USA) with Spectrum Green (ENZO, USA, cat no. 02N32-050 ) for EML4 gene using a nick translation (Abbott, USA, cat no. 07J00-001) based method according to manufacturer's instructions. In brief, TMA sections were deparaffmized and pretreated using the SpotLight Tissue Kit

(Invitrogen, USA, cat no.00-8401) according to manufacturer's instructions. Tissue sections and ALK/EML4 probes were then co-denaturated at 80°C for 5 minutes and hybridized at 37°C for 48 hours. Excess probe was removed with post-hybridization wash buffer (0.3% NP40/lxSSC) by washing the slides at 75°C for 5 minutes, followed by 2><SSC at room temperature for 2 minutes. Sections were then counterstained with 0.3 μg/ml DAPI (4', 6 diamidino-2- phenylindole; Vector, H-1200), coverslipped and stored at 4°C until signal observation.

ALK and EML4 signals were observed under a fluorescence microscope (Olympus, Japan, BX61) using a lOOx objective and proper filters. Co-localization of red ALK signal and green EML4 signal was defined as EML4-ALK fusion positive. ALK gene copy number increase was defined as the average gene copy was greater than four.

Sanger sequencing for KRAS exon 2

Sanger sequencing of KRAS exon 2 was performed to assess mutations in amino acids 12/13 in 61 CRC samples and 2 cell line controls (Calu-1 and HCT1 16). Thirty ng of genomic DNA from each sample was amplified in duplicate 20 reactions containing 0.5μΜ forward and reverse primers (Life technologies, Grand Island, NY 14072) and Ι ΟμΙ, HotStar Taq Plus 2x Master Mix (Qiagen). PCR program: initial denaturation at 95°C (5 min); 35 amplification cycles of 94°C (30 sec), 59°C (30 sec), 72°C (1 min); and a final extension at 72°C (10 min). Amplification products were purified using the QIAquick PCR purification kit (Qiagen) and eluted in 30 μί. Sequencing reactions were performed using 2 μΐ, of template with 0.3 μΜ forward or reverse primer and 2 μΐ_^ of BigDye Terminator v3.1 Ready Reaction Mix in a 10 μΐ_^ reaction (Invitrogen). Sequencing program: 25 cycles of 96°C (10 sec), 50°C (5 sec), 60°C (4 min). Extension products were purified using the BigDye XTerminator purification protocol (Invitrogen) and run on an Applied Biosystems 3730 DNA Analyzer. All samples were sequenced in the forward and reverse directions.

Forward primer: 5' CGATACACGTCTGCAGTCAAC 3' (SEQ ID NO: 13)

Reverse primer: 5' CCTGACATACTCCCAAGGAAAG 3'.(SEQ ID NO: 14)

Data analysis

Following demultiplexing, FASTQ files generated from the Illumina HiSeq whole genome sequencing run were aligned to the human genome (hgl9) using Bowtie2 (v2.0.0- beta7). SAMtools (vO.1.18) was used to convert the aligned sequence data from a SAM to BAM format. The data was subsequently sorted and indexed with SAMtools. The SAMtools mpileup function was used to summarize base calls at each locus. SNP and INDEL calls were made with VarScan (v2.3.2). SNPs and INDELs were characterized as being significantly different from the reference sequence if the variant to reference base frequency was >1%, the calculated VarScan p-value for variant calls (based on Fisher's exact test) was <0.001, total minimum depth>500, variant minimum depth>5, and base quality value greater than 30. All identified variants within a particular sample were exported as a VCF (v4.1) file. VCFs were examined with snpEff (v3.0) for variant annotation and prediction of variant effects on genes. Further data interrogation and graphs were conducted/created in R (v2.15.1).

For variants that matched hits in dBSNP, at least 2 biological or technical replicates were required to confirm the genotype call. For novel variants not catalogued in dBSNP, at least 3 biological or technical replicates were required to confirm the genotype call. For specimens with at least two technical replicates (Patients A and C), the variant frequency threshold for calling a mutation was set to at least 5%, while for the specimen with only a single run (Patient B), the variant frequency threshold was set to at least 10%. Variant calling threshold determination

To determine the level of sensitivity and establish a threshold for identifying rare somatic mutations with targeted ultra-deep sequencing, an empirical distribution was constructed using all coding KRAS loci for the 61 CRC patient specimens. This strategy was developed to assign a variant frequency threshold that exceeds the noise level associated with sequencing error, such as that of Illumina sequencing instruments ranging from 0.05%-l% (Quail MA, et al. (2008) A large genome center's improvements to the Illumina sequencing system. Nat. Methods. 5: 1005- 1010; Nazarian R, et al. (2010) Melanomas acquire resistance to B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature. 468:973-977; He Y, et al. (2010) Heteroplasmic mitochondrial DNA mutations in normal and tumour cells. Nature. 464:610-614 ; and Gore A, et al. (2011) Somatic coding mutations in human induced pluripotent stem cells. Nature. 471 : 63- 67.). With sequencing depth >100,000X for each targeted gene, this allowed a characterization of the boundary between a true variant call and sequencing error with confidence. For each coding locus across the KRAS gene in each of the CRC patients, the frequency of the most prevalent nucleotide (among reads mapped to the hgl9 reference), alternative to the reference genome (MPNAR) called at that locus was determined. These MPNAR frequencies were calculated and a cumulative convergence curve (i.e. scree plot) was constructed for each patient indicating the cumulative summation of the number of MPNARs at frequency thresholds ranging from 0.05% to 5% in increments of 0.05%. To identify the points of convergence in each curve, change-point regression was used and the breakpoints identified in each patient's curve were retained. Each changepoint in a patient curve indicates the threshold (greater than) at which a true variant call would be considered. Then, for each range of breakpoint values (e.g. 0.25%>- 0.5%), 0.5%)-0.75%>, 0.75%)-l%>, etc.), the proportion of patients with a breakpoint occurring within the corresponding breakpoint range was calculated. The results from this analysis are provided in Figure 1A, where each patient's convergence curve (blue line) is plotted with the corresponding proportion of patients (and 95% confidence intervals) having a breakpoint at the MPNAR frequency value (red points) on the x-axis.

From the distribution observed in Figure 1A, the vast majority of patients have breakpoints at MPNAR frequencies up to 0.75% (97% [88.8%, 99.6%] at 0%-0.25%, 97%

[88.8%, 99.6%] at 0.25%-0.5%, and 74% [61.5%, 84.5%] at 0.5%-0.75%), which means that at the variant frequency threshold of >0.75%, 74% of the CRC patients would have a similar number of variants called. Then at a threshold frequency of 0.75%- 1% and greater, the proportion of patients with a similar number of variants called is significantly reduced (3%

[0.4%, 11.2%]. Based on this distribution difference between a MPNAR frequency threshold of 0.5%-0.75% and 0.75%- 1%, and assuming that a similar number of rare variant counts are more likely to exist in a minority of CRC patients (rather than majority), the threshold of >1% was identified for calling true variants. A representative histogram of MPNAR frequencies for a single CRC patient is provided in Figure IB to demonstrate this distribution and determined threshold. Low frequency KRAS mutation was observed in tumors from treatment naive CRC patients

A total of 39 (63%>) CRC patients harbored 26 unique non-synonymous coding mutations with 28 patients (46%) carrying the known active mutations in the first nucleotide binding domain (residues 10-17; Table 1). Eleven CRC patients (18%>) harbored mutations in the GTPase nucleotide binding sites of KRAS which were not detected using Sanger sequencing and five among the eleven have low variant frequencies less than 10%. Twelve (20%) CRC patients harbored more than one mutation in the coding region of KRAS. This information is illustrated in Figure 2. Co-existence of low frequency active KRAS and EGFR kinase domain mutations in a treatment naive Chinese NSCLC patient

The coding exons within KRAS and EGFR were sequenced in a tumor biopsy from a treatment naive Chinese NSCLC patient (Patient A). After excluding germline polymorphisms, 11 non-synonymous single nucleotide changes were identified in EGFR (Table 2). Six of these mutations were located in the extracellular domain, one in the transmembrane domain, two in the C-terminal cytoplasmic domain, and two in the kinase domain. Among these two kinase domain mutations, a well characterized active mutation of EGFR was identified (G719A), which has been associated with sensitivity to EGFR-TKI drugs such as gefitinib and erlotinib in NSCLC patients. Eight (of 11) of these mutations were predicted to impact protein function by SIFT (Ng PC and Henikoff S. Predicting Deleterious Amino Acid Substitutions. Genome

Res 2001 ;11 :863-74) and the variant frequencies ranged from 1.0%-4.3%, with exception to the active mutation which had a variant frequency of 9.9% (thereby on the cusp of Sanger sequencing detection range). For this same NSCLC patient, we also observed a single G12R mutation in KRAS with a low variant frequency of 2.8%. The entire coding exons of ALK and MET were also sequenced, revealing one novel non-synonymous mutation in the cytoplasmic domain of ALK, one novel non-synonymous mutation in the extracellular domain of MET, and one novel non-synonymous mutation in the cytoplasmic domain of MET. The ALK and MET mutations had low variant frequencies ranging from 1.2%~4.4% while no mutations were detected in the MET gene (Figure 3).

Co-existence of EML4-ALK fusion or ALK copy number increase and EGFR, MET and KRAS mutations in tumor tissue from two treatment naive Chinese NSCLC patients

Two treatment naive Chinese NSCLC patients were tested for an EML4-ALK fusion using fluorescent in situ hybridization (FISH). Patient B was positive for the EML4-ALK fusion (Figure 4A) and Patient C was identified with an ALK gene copy number increase with the average copy number around 5 (Figure 4B).

To evaluate the mutations related to crizotinib resistance, all coding exons of KRAS, EGFR, ALK, and MET were sequenced in the tumors from these two NSCLC patients. The summary of the mutational status for Patient B is summarized in Table 3. Four novel non- synonymous mutations in ALK were identified in this patient with all of them located in the extracellular domain of this gene. The mutation P170L, was predicted to have damaging effects on protein function from the SIFT database.

Eight novel non-synonymous mutations were identified in EGFR in this patient with three of them predicted to have damaging effects on protein function from the SIFT database. Among these 8, the heterozygous stop code mutation E66* was located in the N-terminus of EGFR and the homozygous deleterious mutations R832C was found in the kinase domain of EGFR. One novel heterozygous mutation was also identified in the extracellular domain of MET. The variant frequencies of the mutations identified in Patient B ranged from 10.4%- 99.7%.

The summary of the mutational status for Patient C is provided in Table 4. Four non- synonymous mutations were identified in ALK with all the mutations predicted to have damaging effects on protein function (SIFT database). The mutation Rl 181C is located in the kinases domain of ALK. Ten non-synonymous mutations were detected in EGFR. Among them, six are located in the extra-cellular domain and of three are predicted to have damaging effects on protein function (SIFT database). Four of the non- synonymous mutations are located in the cytoplasmic domain and among these four, one mutation (A840T) is in the kinase domain of EGFR. Two non-synonymous mutations were found in the C terminal region of KRAS. Among them, R 164* is a stop codon mutation and results in a truncated KRAS protein. Six non- synonymous mutations were found in MET, with 5 of them located in the extracellular domain and one in the cytoplasmic domain in the region involved in interactions with RANBP9.

The variant frequencies of 24 non-synonymous mutations in Patient C ranged from

1.0%~3.9%. Although 15 of these mutations had an entry in dbSNP, we classified these SNPs as mutations with the following considerations: 1) the variant frequencies of these mutations in the tumor specimens differed by at least 10-fold from the true germline SNPs in these same tumor specimens that had allele frequencies in the range of 32.37%-99.98%, 2) the allele frequencies of these SNPs are unknown in the dbSNP database, and most of these SNPs were generated from the cancer genome project, while the true germline SNPs in these tumor specimens have population data reported in dbSNP.

Discussion

To further explore the presence and potential impact that low frequency somatic mutations in oncodriver genes may have in the resistance to targeted therapies, we sequenced the coding exons of KRAS in 61 treatment naive colorectal cancer tumor (CRC) patients using targeted ultra- deep resequencing. With the extraordinary depth used here, we found that 5 patients (8%) harbored low frequency mutations in GTPase nucleotide biding sites which were not detected using Sanger sequencing. We also sequenced the coding exons of EGFR, KRAS, ALK, and MET in tumor biopsies from 3 Chinese NSCLC patients. We found the co-existence of a low frequency KRAS active mutation (G12R) and EGFR kinase domain mutation (G719A) in tumor specimens in one NSCLC patient, multiple co-existing low frequency mutations in KRAS, EGFR and MET with an ALK gene copy number increase in the second patient, and an EML4-ALK fusion with ALK, EGFR, and MET mutations in the third patient. We also found that multiple low frequency mutations presented in a single gene from the same CRC of NSCLC patients. These data suggest that intrinsic low frequency driver mutations, sometimes on different oncodrivers, in cancer tissues may exist prior to treatment, providing direct evidence for the cause of unsustainable clinical improvement in cancer patients undergoing monospecific targeted therapies.

Using ultra-deep NGS, we were able to detect low frequency mutations in key oncodriver genes associated with treatment response to EGFR therapy in patients with CRC or NSCLC. The sensitivity of NGS, extreme depth and coverage allows such low frequency mutations to become detectable, with >10,000X read depth supporting the variant calls (De Grassi et al.

(2010) Ultradeep sequencing of a human ultraconserved region reveals somatic and

constitutional genomic instability. PLoS Biol. 8(l):el000275).

Table 1. Summary of mutations identified within in the first nucleotide binding domain (residues 10-17) in KRAS in 28 CRC patient tumor specimens

Physical AA Sanger Variant

Patient Chr Genotype Codon dbSNP ID position change result frequency

Ref Reads Var Reads

1 12 25398284 C/T 12 G/D G/D 225479 265428 54.05 rsl21913529

2 12 25398281 C/T 13 G/D G/D 294912 137593 31.72 rsl 12445441

6 12 25398284 C/T 12 G/D G/D 280184 112988 28.73 rsl21913529

8 12 25398284 C/T 12 G/D G/D 268793 210304 43.88 rsl21913529

10 12 25398279 C/T 14 v/i negative 466547 43473 8.52 rsl04894365

11 12 25398287 G/A 11 A/V negative 489126 9436 1.89 NA

13 12 25398285 C/T 12 G/S G/S 185069 82814 30.89 rsl21913530

14 12 25398281 C/T 13 G/D G/D 303499 154506 33.72 rsll2445441

16 12 25398284 C/T 12 G/D negative 330901 10386 3.04 rsl21913529

17 12 25398284 C/T 12 G/D G/D 387693 139363 26.44 rsl21913529

18 12 25398284 C/T 12 G/D G/D 249584 176771 41.45 rsl21913529

19 12 25398284 C/A 12 G/V G/V 142621 317999 68.79 rsl21913529

20 12 25398284 C/A 12 G/V negative 286307 89578 23.79 rsl21913529

21 12 25398284 C/T 12 G/D negative 267279 116448 30.33 rsl21913529

22 12 25398284 C/T 12 G/D G/D 381445 10224 2.61 rsl21913529

23 12 25398284 C/T 12 G/D G/D 269119 214402 44.33 rsl21913529

24 12 25398284 C/A 12 G/V G/V 312311 232101 42.51 rsl21913529

25 12 25398285 C/A 12 G/C G/C 187782 234737 55.06 rsl21913530

30 12 25398284 C/A 12 G/V G/V 283896 141116 33.13 rsl21913529

31 12 25398284 C/A 12 G/V G/V 167032 146537 46.62 rsl21913529

32 12 25398284 C/A 12 G/V negative 412155 64179 13.45 rsl21913529

33 12 25398285 C/T 12 G/S negative 510115 5286 1.02 rsl21913530

33 12 25398281 C/T 13 G/D negative 461791 4727 1.01 rsll2445441

34 12 25398284 C/A 12 G/V negative 323608 107178 24.85 rsl21913529

35 12 25398281 C/T 13 G/D G/D 168767 344810 67.09 rsll2445441

36 12 25398284 C/A 12 G/V negative 276673 100422 26.56 rsl21913529

37 12 25398284 C/A 12 G/V negative 378849 76508 16.78 rsl21913529

38 12 25398284 C/T 12 G/D G/D 180748 189179 51.12 rsl21913529 Table 2: Mutation Summary for Patient A

Physical AA Variant SIFT

Patient Gene Chromosome position Genotype Codon change frequency Type prediction dbSNP ID

A ALK 2 29416310 C/T 1548 G/E 4.41% unknown TOLERATED rs78868998

A EGFR 7 55210117 T/C 76 L/P 1.03% Novel DAMAGING

A EGFR 7 55224340 T/C 374 V/A 1.14% Novel TOLERATED

A EGFR 7 55225386 A/C 413 N/T 1.36% Novel DAMAGING

A EGFR 7 55228013 A/G 494 R/G 1.30% Novel DAMAGING

A EGFR 7 55228025 A/G 498 S/G 1.02% Novel TOLERATED

A EGFR 7 55233058 A/C 603 N/T 3.23% Novel TOLERATED

A EGFR 7 55240708 T/C 651 V/A 1.22% Novel DAMAGING

A EGFR 7 55241693 A/G 714 K/R 1.12% Novel TOLERATED

A EGFR 7 55241708 G/C 719 G/A 9.89% cosmic DAMAGING rsl21913428

A EGFR 7 55268981 A/C 1016 Y/S 3.63% Novel DAMAGING

A EGFR 7 55269035 T/G 1034 L/R 4.34% germline DAMAGING rs34352568

A KRAS 12 25398285 C/G 12 G/R 2.28% cosmic DAMAGING rsl2191353C

A MET 7 116403276 A/G 864 K/R 1.23% Novel TOLERATED

A MET 7 116415010 T/C 1053 L/P 2.07% Novel TOLERATED

Table 3: Mutation summary for Patient B

Physical AA Variant

Patient Gene Chromosome position Genotype Codon change frequency Type SIFT prediction

B ALK 2 29754832 T/G 368 H/P 10.35% Novel TOLERATED

B ALK 2 29940510 AJG 241 F/L 26.50% Novel TOLERATED

B ALK 2 30142933 AJG 198 V/A 44.02% Novel TOLERATED

B ALK 2 30143017 AJG 170 L/P 21.70% Novel DAMAGING

B EGFR 7 55210086 G/T 66 E/* 99.69% Novel STOP

B EGFR 7 55214384 T/G 170 S/R 34.34% Novel TOLERATED

B EGFR 7 55214392 T/C 173 L/P 31.39% Novel TOLERATED

B EGFR 7 55214419 AJG 182 N/S 37.63% Novel TOLERATED

B EGFR 7 55220316 T/C 236 C/R 99.85% Novel DAMAGING

B EGFR 7 55225406 T/C 420 F/L 40.32% Novel DAMAGING

B EGFR 7 55249040 A/G 780 I/V 12.65% Novel TOLERATED

B EGFR 7 55259436 C/T 832 R C 99.65% Novel DAMAGING

B MET 7 1 16380999 T/G 541 C/G 23.55% Novel DAMAGING

Table 4: Mutation summary for Patient C

Physical AA Variant SIFT

C ALK 2 29416122 T/C 1611 S/G 1.06% Novel DAMAGING

C ALK 2 29443676 G/A 1181 R/C 1.66% cosmic DAMAGING rs56315533 unknown

(not in 4522

c ALK 2 29551251 T/C 460 D/G 1.29% chromosome) DAMAGING rsl49198543 c ALK 2 29754949 G/A 329 S/F 2.23% somatic DAMAGING rsl44135148 c EGFR 7 55210017 A/G 43 T/A 1.03% unknown DAMAGING rsl47740818 c EGFR 7 55211037 AJG 94 N/D 2.43% Novel DAMAGING c EGFR 7 55223639 G/A 336 V/M 1.88% cosmic DAMAGING rs 144997037 c EGFR 7 55225386 A/C 413 N/T 1.12% Novel DAMAGING c EGFR 7 55227861 T/C 443 L/P 1.19% Novel TOLERATED c EGFR 7 55227915 T/C 461 V/A 1.21% uknown TOLERATED rs200592648 c EGFR 7 55238869 T/C 628 C/R 1.42% Novel DAMAGING c EGFR 7 55259460 G/A 840 A T 1.04% unknown DAMAGING rsl43884981 unknown

(cancer

genome

c EGFR 7 55269035 T/G 1034 L/R 3.03% project) DAMAGING rs34352568 c EGFR 7 55273149 G/C 1158 A P 1.97% unknown TOLERATED rs 140028234 c MET 7 116340097 C/T 320 A V 1.96% unknown DAMAGING rs35776110 c MET 7 116340157 A/G 340 D/G 3.84% unknown DAMAGING rs200690492 c MET 7 116398625 C/T 739 R/C 1.21% unknown DAMAGING rs45587940 c MET 7 116403276 A/G 864 K/R 1.20% Novel TOLERATED c MET 7 116403287 A/G 868 I/V 1.12% uknown TOLERATED rs200524064 c MET 7 116423456 A/G 1262 K/R 1.25% uknown TOLERATED rsl21913677 c KRAS 12 25368455 G/A 164 R/* 2.08% uknown DAMAGING rs200186815 c KRAS 12 25368487 T/C 153 E/G 1.18% Novel DAMAGING

Table 5:

SEO ID NO: Description

SEQ ID NO: 1 Kirsten rat sarcoma viral oncogene homolog (KRAS) (nucleotide)

SEQ ID NO: 2 KRAS precursor

(Protein)

SEQ ID NO: 3 epidermal growth factor receptor (EGFR) nucleotide)

SEQ ID NO: 4 EGFR (protein)

SEQ ID NO: 5 anaplastic lymphoma receptor tyrosine kinase (ALK) nucleotide

SEQ ID NO: 6 ALK tyrosine kinase receptor precursor

SEQ ID NO: 7 ALK/EML4 fusion (nucleotide)

SEQ ID NO: 8 ALK/EML4 fusion (nucleotide)

SEQ ID NO: 9 ALK/EML4 fusion (protein)

SEQ ID NO: 10 ALK/EML4 fusion (protein)

SEQ ID NO: 11 met proto-oncogene (hepatocyte growth factor receptor) (MET)

(nucleotide)

SEQ ID NO: 12 hepatocyte growth factor receptor (MET) precursor

(Protein)

SEQ ID NO: 13 Artificial forward primer sequence (KRAS)

nucleotide

SEQ ID NO: 14 Artificial reverse primer sequence (KRAS)

nucleotide

Claims

1. A method for selecting a treatment for cancer in a patient in need thereof, the method comprising:

(i) determining whether one or more mutations are present in a tissue sample obtained from the patient, wherein one or more mutations are selected from:

(a) one or more mutations selected from a mutation in a cytoplasmic domain of MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof;

(b) one or more mutations in an extracellular domain of ALK selected from H368P, F241L, V198A, L170P, and combinations thereof;

(c) a mutation in a kinase domain of ALK including Rl 181C;

(d) one or more mutations selected from a mutation in an extracellular domain of EGFR selected from E66*, S170R, L173P, N182S, C236R, F420L, and combinations thereof; and

(e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof; and

(ii) administering to the patient an effective amount of an anti-cancer therapy tailored to treat cancer cells having one or more of the mutations.

2. A method of detecting the presence of one or more mutations in a tissue sample, the method comprising:

(i) amplifying nucleic acid from the tissue sample;

(ii) sequencing the amplified nucleic acid; and

(iii) determining whether one or more mutations are present, wherein the one or more mutations are selected from:

(a) one or more mutations selected from a mutation in a cytoplasmic domain of MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof; (b) one or more mutations in an extracellular domain of ALK selected from H368P, F241L, V198A, L170P, and combinations thereof;

(c) a mutation in a kinase domain of ALK including Rl 181C;

(e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof.

3. A method for treating non-small cell lung cancer (NSCLC) in a patient, the method comprising:

(i) obtaining a tissue sample from the patient;

(ii) determining whether the tissue sample includes at least a subset of cells comprising one or more mutations selected from:

(c) a mutation in a kinase domain of ALK including Rl 181C;

(iii) if at least a subset of cells includes one or more of the mutations, administering to the patient an effective amount of an anti-cancer therapy tailored to treat cancer cells having one or more of the mutations.

4. A method for treating colorectal cancer (CRC) in a patient, the method comprising: (i) obtaining a tissue sample from the patient;

(c) a mutation in a kinase domain of ALK including Rl 181 C;

(iii) if the tissue sample includes at least a subset of cells comprising one or more of the mutations, administering to the patient an effective amount of an anti-cancer therapy tailored to treat cancer cells having one or more of the mutations.

5. A method for determining whether a tissue sample obtained from a patient includes at least a subset of EGFR anti-cancer therapy non-responsive cells, the method comprising determining whether one or more mutations are present in the tissue sample, wherein one or more mutations are selected from:

(b) one or more mutations in an extracellular domain of ALK selected from H368P, F241L, VI 98 A, L170P, and combinations thereof;

(c) a mutation in a kinase domain of ALK including Rl 181C;

(d) one or more mutations selected from a mutation in an extracellular domain of EGFR selected from E66*, S170R, L173P, N182S, C236R, F420L, and combinations thereof; and (e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof,

wherein the presence of one or more of the mutations in the tissue sample indicates that a least a subset of cells in the tissue sample may be resistant to EGFR anti-cancer therapy.

6. A kit for determining whether a tissue sample includes at least a subset of EGFR-therapy resistant cells, the kit comprising:

(i) one or more reagents for determining a presence of one or more mutations selected from:

(a) one or more mutations selected from a mutation in a cytoplasmic domain of

MET including L1053P; a mutation in an extracellular domain of MET including K864R, a mutation in a cytoplasmic domain of ALK including G1548E, and combinations thereof;

(c) a mutation in a kinase domain of ALK including Rl 181C;

(e) one or more mutations selected from a mutation in a cytoplasmic domain of

EGFR selected from I780V, R832C, and combinations thereof; and

(ii) instructions for performing the assay.

7. The method of claim 6, wherein the kit further comprises a container for the reagents.

8. A method of identifying a patient as a candidate for an anti-cancer therapy comprising: determining whether one or more mutations are present in a tissue sample obtained from the patient, wherein one or more mutations are selected from:

(c) a mutation in a kinase domain of ALK including Rl 181 C;

(e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof;

wherein presence of one or more mutations identifies the patient as the candidate for the anti-cancer therapy.

9. A pharmaceutical composition for treating a cancer patient in which one or more mutations are detected, wherein one or more mutations are selected from:

(b) one or more mutations in an extracellular domain of ALK selected from H368P, F241 L, V 198 A, L 170P, and combinations thereof;

(c) a mutation in a kinase domain of ALK including Rl 181C;

(e) one or more mutations selected from a mutation in a cytoplasmic domain of EGFR selected from I780V, R832C, and combinations thereof,

the pharmaceutical composition comprising: (i) a therapeutically effective amount of an anti-cancer therapeutic agent tailored to treat cancer cells having one or more of the mutations; and (ii) a pharmaceutically acceptable carrier.

10. The method of any of claims 1-8, wherein the tissue sample comprises a genetically heterogeneous population of cancer cells.

11. The method of claim 10, wherein one or more mutations are present in at least a subset of cells in the genetically heterogeneous tissue sample.

12. The method of any of the preceding claims, wherein one or more mutations comprise low frequency mutations

13. The method of any of the preceding claims, wherein one or more mutations have a frequency of less than about 10%.

14. The method of any of the preceding claims, wherein one or more mutations have a frequency of less than about 5%.

15. The method of any of the preceding claims, wherein one or more mutations have a frequency of at least about 1%.

16. The method of any of the preceding claims, wherein one or more mutations indicate that at least a subset of cells may be resistant to EGFR-therapy.

17. The method of any of the preceding claims, wherein one or more mutations comprise one or more mutations selected from a mutation in a cytoplasmic domain of MET comprising L1053P; a mutation in an extracellular domain of MET comprising K864R, a mutation in a kinase domain of ALK comprising G1548E, and combinations thereof.

18. The method of claim 17, wherein one or more mutations further comprises one or more low frequency K-RAS active domain mutations in combination with one or more low frequency EGFR kinase domain mutations.

19. The method of claim 17, wherein the K-RAS active domain mutation comprises one or more point substitutions in codons 12 or 13.

20. The method of claim 19, wherein the K-RAS active domain mutation comprises G12R.

21. The method of claim 18, wherein the EGFR kinase domain mutation comprises G719A.

22. The method of claim 18, wherein the K-RAS active domain mutation comprises G12R and the EGFR kinase domain mutation comprises G719A.

23. The method of any of claims 1-16, wherein one or more mutations comprise one or more mutations in an extracellular domain of ALK selected from H368P, F241L, V198A, L170P, and combinations thereof.

24. The method of claim 23, wherein one or more mutation further comprise more than one low frequency mutations in K-RAS and EGFR in combination with MET and ALK gene copy number increases.

25. The method of any of claims 1-16, wherein one or more mutations comprise a mutation in a kinase domain of ALK comprising Rl 181C.

26. The method of any of claims 1-16, wherein one or more mutations comprise one or more EGFR mutations selected from E66*, S170R, L173P, N182S, C236R, F420L, I780V, R832C, and combinations thereof.

27. The method of claim 26, wherein E66* is a heterozygous stop code mutation located in an N-terminus of EGFR.

28. The method of claim 26, wherein R832C is a homozygous deletion mutation in a kinase domain of EGFR.

29. The method of any of claims 1-16, wherein one or more mutations further comprise a mutation in an extracellular domain of MET.

30. The method of claim 29, wherein the mutation comprises C541G.

31. The method of any of claims 1, 3-4, 5 or 10-30, wherein the anti-cancer therapy includes at least one non-EGFR anti-cancer therapeutic.

32. The method of claim 31 , wherein the non-EGFR anti-cancer therapeutic is selected from: chemotherapeutic agents, growth inhibitory agents, cytotoxic agents, agents used in radiation therapy, anti-angiogenesis agents, apoptotic agents, anti-tubulin agents, neutralizing antibodies, tyrosine kinase inhibitors, platelet derived growth factor inhibitors, COX-2 inhibitors, interferons, cytokines, organic chemical agents, and combinations thereof.

33. The method of claim 31 , wherein the non-EGFR anti-cancer therapeutic includes one or more chemotherapeutic agents selected from: an alkylating agent, an antimetabolite, an antitumor antibiotic, an antimitotic, a topoisomerase inhibitor, a proteasome inhibitor, tyrosine kinase inhibitors, and combinations thereof.

34. The method of claim 31 , wherein the non-EGFR anti-cancer therapeutic agent comprises one or more chemotherapeutic agents selected from:

(a) an alkylating agent selected from cisplatin, carboplatin, oxaliplatin,

cyclophosphamide, nitrogen mustard, melphalan, chlorambucil, busulphan and nitrosourea;

(b) an antimetabolite selected from antifolates, raltitrexed, gemcitabine, capecitabine, methotrexate, pemetrexed (Alimta), cytosine arabinoside and hydroxyurea;

(c) antitumor antibiotics selected from anthracyclines, bleomycin, doxorubicin, daunomycin, epirubicin, idarubicin, mitomycin-C, dactinomycin and mithramycin;

(d) antimitotic agents selected from vinca alkaloids and taxoids;

(e) topoisomerase inhibitors selected from epipodophyllotoxins, irinotecan, amsacrine, topotecan and camptothecin;

(f) proteasome inhibitors; and

(g) tyrosine kinase inhibitors selected from Axitinib, Bosutinib, Cediranib, Dasatinib, Erlotinib, Gefitinib, Imatinib, Lapatinib, Lestaurtinib, Nilotinib, Semaxanib, Sunitinib, and

Vandetanib.

35. The method of any of claims 1,3-4 and 10-30, wherein the anti-cancer therapy comprises combination therapy.

36. The method of claim 35, wherein combination therapy comprises an EGFR inhibitor in combination with a non-EGFR anti-cancer therapeutic agent.

37. The method of claim 36, wherein the EGFR inhibitor is selected from cetuximab, panitumumab, erlotinib, gefitinib, and combinations thereof.

38. The method of claim 36, wherein the EGFR inhibitor and the non-EGFR

chemotherapeutic agent are administered simultaneously.

39. The method of claim 36, wherein the EGFR inhibitor and the non-EGFR

chemotherapeutic agent are administered sequentially.

40. The method of any of the preceding claims, wherein cancer includes non-small cell lung cancer (NSCLC).

41. The method of any of the preceding claims, wherein cancer includes colorectal cancer (CRC).

42. The method of any of the preceding claims, wherein the patient is human.

43. The method of any of the preceding claims, wherein the patient is treatment naive.

44. The method of any of the preceding claims, wherein the tissue sample comprises a solid or fluid tissue sample.

45. The method of claim 44, wherein the tissue sample is obtained from a cancerous tissue.

46. The method of any of claims 1-9, wherein the presence of the mutation is determined by amplifying nucleic acid from the tumor and sequencing the amplified nucleic acid.

47. The method of claim 46, wherein the presence of the low frequency mutation is determined using next generation sequencing (NGS) technology selected from 454 sequencing, Solexa, SOLiD, Polonator, and HeliScope Single Molecule Sequencer technologies.