EP1836313A2 - Methods for assessing patients with acute myeloid leukemia - Google Patents

Abstract

Methods for treating cancer, and preferably hematological malignancy, patients include analyzing gene expression profiles and/or molecular markers of a patient to determine status and/or prognosis of the patient. The invention also provides methods of analyzing whether a non-relapsed or non-refractory patient is likely to respond to treatment with farnesyl transferase inhibitors (FTIs) and, optionally, other therapeutics. The methods are also useful for monitoring patient therapy and for selecting a course of therapy. Genes modulated in response to FTI treatment are provided and are used in formulating the profiles.

Description

TITLE OF THE INVENTION

METHODS FOR ASSESSING PATIENTS WITH ACUTE MYELOID LEUKEMIA

A "Sequence Listing" listing appendix is hereby incorporated by reference herein. BACKGROUND OF THE INVENTION

This invention relates to diagnostics, prognostics, and treatments for acute myeloid leukemia (AML) based on the detection of molecular markers and/or gene expression analysis.

Karyotyping is currently effective in providing prognostic value while it also serves to identify biologically distinct subtypes of AML. In addition, mutations in genes such as FLT3, c-KIT, AMLl, GATAl, CEBPA and N-RAS are implicated in the pathogenesis of the disease. It is clear, that screening for FLT3 and CEBPA mutations can stratify groups that have different risks of relapse. Effective risk stratification can allow for the appropriate use of allogeneic stem cell transplantation or other adjuvant therapies. Two papers published recently describe gene expression profiling of newly diagnosed adult AML patients and its use in predicting clinical outcome. Bullinger et al. (2004); and VaIk et al. (2004). These studies show how gene-expression profiling can further refine clinical outcome prediction.

VaIk et al. (2004) evaluated 285 patients (bone marrow or peripheral blood) on the Affymetrix U133A chip. The patient samples encompassed a wide range of cytogenetic and molecular abnormalities. Only 16 clusters were identified indicating AML may not be as heterogeneous as previously thought. Several of the clusters corresponded well with the cytogenetically and molecularly defined subtypes of AML thus supporting their use in the WHO classification system. These clusters were also seen by Bullinger et al. (2004) and other previously published smaller studies. Schoch et al. (2002); Debernardi et al. (2003); and Kohlmann et al. (2003). These clusters, not surprisingly, correlated with prognosis since they were associated with well known prognostic karyotypes.

Bullinger et al. (2004) investigated expression profiles from 116 adult patients (65 peripheral blood and 54 bone marrow) using cDNA arrays. In addition to the work done by VaIk et al. (2004) they also developed al33 gene classifier for predicting clinical outcome across all cytogenetic risk groups. Using a training set of 59 samples and a testing set of 57 samples they showed that the 133 genes clustered patients into poor and good outcome groups (p = 0.006 log rank; odds ratio, 10, 95% CI, 2.6-29.3).

Notably, the genes identified in both these studies overlap, only in part, to predictor genes previously identified in childhood leukemia. Yagi et al. (2003). Also, there is no overlap between the prognostic gene set identified by Bullinger et al. (2004) and the 3 genes recently identified that predict response to tipifarnib. US patent application serial no. 10/883,436.

The farnesyl transferase (FTase) enzyme mediates the covalent attachment of a carbon farnesyl moiety to the C-terminal CAAX (C, cysteine; A, aliphatic residue; X, any amino acid) recognition motif. Reiss et al. (1990). This farnesylation is further processed by cleavage of the 3 terminal amino acids (AAX) and methylation of the C-terminal isoprenyl-cysteine. The inhibition of protein farnesylation abrogates the correct subcellular localization required for protein function. Originally, the oncogenic Ras protein was thought to be the target for the antiproliferative effects of FTIs in cancer biology. Reuter et al. (2000). However, it has since been shown that inhibition of Ras farnesylation does not account for all of actions of tipifarnib. For example, FTIs do not always require the presence of mutant Ras protein to produce antitumor effects. Karp et al. (2001). Indeed, while early clinical studies were designed around populations with a high frequency of ras mutations, such as advanced colorectal and pancreatic cancer, no significant difference in response rates were seen when compared to placebo. Van Cutsem et al. (2004); and Rao et al. (2004).

Several other farnesylated proteins have been implicated as candidate targets that may mediate the antitumorigenic effects of FTIs including the small GTPase proteins Rho B, the centromere proteins CENP-E and CENP-F, the protein tyrosine phosphatase PTP-CAAX, and the nuclear membrane structural lamins A and B. The inhibition of farnesylation of these proteins may lead to the antiproliferative effect of FTIs and also indirectly modulate several important signaling molecules including TGFβRII, MAPK/ERK, PI3K/AKT2, Fas (CD95), NF-κB, and VEGF. Adnane et al. (2000); Morgan et al. (2001); Jiang et al. (2000); Na et al. (2004); Takada et al. (2004); and Zhang et al. (2002). Regulation of these signaling pathways leads to the modulation of cell growth, proliferation, and apoptosis. Thus, FTIs may have complex inhibitory effects on several cellular events and pathways. There are currently no methods for determining status or predicting overall survival of these patients. BRIEF SUMMARY OF THE INVENTION

The invention is a method of using one or more gene signatures for predicting prognosis in patients with acute myeloid leukemia (AML). These signatures can be used alone or in combination depending upon the type of drug treatment.

The present invention provides a method of assessing acute myeloid leukemia (AML) status by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 where the expression levels of the Marker genes above or below pre-determined cut-off levels are indicative of AML status.

The present invention provides a method of staging acute myeloid leukemia (AML) patients by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 where the expression levels of the Marker genes above or below pre-determined cut-off levels are indicative of AML survival.

The present invention provides a method of determining acute myeloid leukemia (AML) patient treatment protocol by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 where the expression levels of the Marker genes above or below pre-determined cut-off levels are sufficiently indicative of response to therapy to enable a physician to determine the degree and type of therapy recommended to provide appropriate therapy.

The present invention provides a method of treating a acute myeloid leukemia (AML) patient by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 where the expression levels of the Marker genes above or below pre-determined cut-off levels are indicate a response to therapy and; treating the patient with adjuvant therapy if they have a responder profile.

The present invention provides a method of determining whether a acute myeloid leukemia (AML) patient is high or low risk of mortality by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3 where the expression levels of the Marker genes above or below pre-determined cut-off levels are sufficiently indicative of risk of mortality to enable a physician to determine the degree and type of therapy recommended.

The present invention provides a method of generating an acute myeloid leukemia (AML) prognostic patient report by determining the results of any one of the above-described methods; and preparing a report displaying the results and reports generated thereby. The present invention provides a kit for conducting an assay to determine acute myeloid leukemia (AML) prognosis in a biological sample comprising: materials for detecting isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9.

The present invention provides articles for assessing acute myeloid leukemia (AML) status comprising: materials for detecting isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9.

The present invention provides a microarray or gene chip for performing the above-described methods.

The present invention provides a diagnostic/prognostic portfolio comprising isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or

Table 9.

BRIEF DESCRIPTON OF DRAWINGS

Figure 1. Unsupervised clustering of relapsed and refractory AML patients. The dendogram shows the unsupervised k-means clustering of 58 relapsed or refractory AML patients, where each column represents a patient and each row represents a gene. The expression ratio for each gene was calculated by dividing the expression level of that gene in a patient by the mean of all other patients. The color bar indicates the fold-change (log₂). Red is upregulated, blue is down-regulated. White indicates no change. The presence of 6 main clusters is shown.

Figure 2. Real-time RT-PCR of 2 genes. AHR and AKAP13 were measured by real-time RT-PCR. The HPRT or PBGD control genes were used to normalize gene expression values. Error bars are standard deviations. The resulting values were plotted against the corresponding microarray data and linear regression analysis was performed.

Figure 3 depicts the predictive value of the AKAP 13 gene. Panel A shows a 2x2 table generated from a LOOCV performed using AKAP 13 expression as a classifier on the responders (R) and non-responders (NR). Panel B shows the AKAP 13 expression values for the same 58 patients. The P value indicates a significant difference in the gene expression between the mean values of each response group. Panel C shows the Kaplan-Meier curves generated from patients classified by the AKAP 13 gene as being responders and non-responders. Figure 4 provides identification of a minimal set of predictive markers. In Panel

A, a LOOCV was performed using a sensitivity of 100%. Independent classifiers were tested that contained from 1 to 19 genes. The resulting error rate is plotted. Panel B shows a 2x2 table generated from a LOOCV performed using the 3-gene signature as a classifier on the responders (R) and non-responders (NR). Panel C shows the scores generated from the 3-gene classifier. The P value indicates a significant difference in the gene expression between the response groups. Panel D is the Kaplan-Meier curves generated from patients classified by the 3-gene signature as being responders and non-responders. Median survival times are also indicated. Figure 5. A Kaplan-Meier analysis was performed on patients classified by the

3-gene signature as being predicted responders and non-responders. The survival curve of patients who were clinically defined as non-responders but classified as responders using the 3-gene signature is shown. Median survival times are also indicated. Figure 6 depicts over-expression of AKAPl 3 in an AML cell line. Cell counts were normalized to cultures with no drug (indicated at -12 log units) to give a percentage of control. Error bars indicate standard errors of the mean. Open data points indicate results from a second experiment exploring higher concentrations of drug. Figure 7 provides a model of FTI action in relapsed or refractory AML. A. In responders the IL3RA and AKAPl 3 genes are lowly expressed allowing for down-regulation of the ras, and RhoA, and lamin B pathways, respectively. Up-regulation of RhoH leads to increased inhibition of cellular transformation pathways. Together this allows for greater efficacy in FTI antitumorigenicity. B. The opposite expression profile is seen in non-responders allowing for the expression of compensatory pathways.

Figure 8. The Zarnestra predictive gene signature has superior utility to an independent prognostic gene signature. In panel A columns represent AML samples from relapsed or refractory patients and rows represent 167 probe sets that correspond to 103 of the 133 prognostic genes identified by Bullinger et al. ', ordered according to hierarchical clustering. Panel B shows Kaplan-Meier survival estimates of the cluster-defined groups of patients. In panel C the 3-gene classifier has been used to identify responders of tipifarnib in the good and poor prognostic groups defined by the Bullinger signature. Kaplan-Meier survival curves are shown for patients identified as being responders to tipifarnib in the good (Zn+.clusterl) and poor (Zn+.cluster2) prognostic groups. The median survival times for each group are indicated.

Figure 9 is a flow chart depicting how the genes from Bullinger et al. (2004) were matched to 167 probe sets (103 unique genes) on the Affymetrix Ul 33 A chip.

Figure 10 shows the utility of the 167 probe set signature in relapsed or refractory AML patients. In panel A columns represent AML samples from relapsed or refractory patients and rows represent 167 probe sets that correspond to 103 of the 133 prognostic genes identified by Bullinger et al. (2004), ordered according to hierarchical clustering. Panel B shows Kaplan-Meier survival estimates of the cluster-defined groups of patients.

Figure 11 provides comparisons of prognostic and Zarnestra predictive gene signatures. Panel A shows the Kaplan-Meier survival curves for the good and poor prognostic clusters as defined by the subset of 103 Bullinger et al. (2004) genes. Panel B shows the Kaplan-Meier survival curves for the good and poor prognostic clusters as defined by the 3-gene signature that predicts response to Zarnestra. Panel C shows the Kaplan-Meier survival curves for the good and poor prognostic clusters from Panel A further stratified by the 3-gene Zarnestra signature. Panel D shows the Kaplan-Meier survival curves for patients who are predicted to have a poor prognosis and not respond to Zarnestra versus the remainder of patients.

Figure 12 Identification of a minimal set of predictive markers, a) A LOOCV was performed selecting for genes with a sensitivity of 100%, specificity of 40% and fold-change > 2. Independent classifiers were tested that contained from 1 to 8 genes ranked by the AUC . The resulting error rate is plotted, b) A 2x2 table generated from a LOOCV performed using AKAP 13 as a classifier on the responders (R) and non-responders (NR). c) The gene-expression values of AKAPl 3. The P value indicates a significant difference in the gene expression between the response groups, d) The Kaplan-Meier curves generated from patients classified by AKAPl 3 as being responders and non-responders. Median survival times are also indicated.

Figure 13 depicts an overview of gene expression analysis. Figure 14 depicts AML samples maintain FTI-mediated global gene expression changes following termination of tipifarnib treatment.

Figure 15 depicts predictive expression profiles and testing of predictive classifiers in newly diagnosed AML.

Figure 16 depicts the 6-gene classifier stratifies newly diagnosed AML. DETAILED DESCRIPTION OF THE INVENTION A subset of genes previously described to have prognostic value in newly diagnosed AML is shown here to have utility in relapsed and refractory AML patients treated with a molecularly targeted therapy (Zarnestra). Currently there is no method for predicting response to farnesyl transferase inhibitors (such as Zarnestra). Also, current methods for understanding the prognosis of patients with AML is limited to histological subtype and karyotyping, both of which are not ideal markers for determining clinical outcome. The current signatures expand upon these traditional technologies by providing better stratification of prognostic high risk and low risk patients.

US Patent Application Serial No. 10/883,436 demonstrates that a 3-gene classifier (including AHR, AKAPl 3 and MINA53) predicts relapsed, refractory

AML patient response to tipifarnib (Zarnestra^®, Rl 15777) with the lowest error rate. This was also seen when a leave-five-out cross validation was performed. When more genes were added the error rate increased indicating that additional genes introduced noise to the classifier. For the 3 -gene classifier the LOOCV demonstrated a sensitivity of 86% and specificity of 70% with an overall diagnostic accuracy of 74%. Kaplan-Meier analysis again showed a significant difference in survival between the predicted responder group and the non-responder group. Moreover, comparing the incorrectly classified non-responders to the correctly classified non-responders, the misclassified non-responders showed a better overall survival.

Zarnestra^® is an orally available non-peptidomimetic competitive farnesyl transferase inhibitor (FTI) that has been shown to inhibit the proliferation of a variety of human tumor cell lines both in vitro and in vivo. End et al. (2001); and Cox et al. (2002). A phase I clinical trial of tipifarnib demonstrated a 32% response rate in patients with refractory or relapsed acute myeloid leukemia. Karp et al. (2001). Activity has also been seen in early clinical trials for myelodysplastic syndrome (MDS) (Kurzrock et al. (2004)), multiple myeloma (MM) (Alsina et al. (2003)) and chronic myeloid leukemia (CML). Cortes et al. (2003). Complete remission was defined as less than 5% bone marrow blasts with a neutrophil count greater than 1000/μL, a platelet count less than 100,000/μL, and no extramedullary disease. While it is clear that FTIs function by inhibiting protein farnesylation, it is still not known what genes are implicated in the antitumor effects of tipifarnib in hematopoietic malignancies. Microarray technology allows for the measurement of the steady-state mRNA level of thousands of genes simultaneously, thereby representing a powerful tool for identifying genes and gene pathways that correlate with FTI action. Global gene expression monitoring was therefore employed in a phase 2 clinical study of tipifarnib in relapsed and refractory AML to identify genes that predict response to this FTI in hematologic malignancies.

The mere presence of nucleic acid sequences having the potential to express proteins or peptides ("genes") within the genome is not determinative of whether a protein or peptide is expressed in a given cell. Whether or not a given gene capable of expressing proteins or peptides or transcribing RNA does so and to what extent such expression or transcription occurs, if at all, is determined by a variety of complex factors. Nevertheless, assaying gene expression can provide useful information about the cellular response to a given stimulus such as the introduction of a drug or other therapeutic agent. Relative indications of the degree to which genes are active or inactive can be found in such gene expression profiles. In some instances, the presence of a molecular marker can, by itself or with the use of gene expression information, provide useful information about treatment efficacy too. The gene expression profiles and molecular markers of this invention are used to identify and treat AML patients.

Cancers, including hematological malignancies, typically arise from mutations in a variety of genes. The same type of cancer may arise as a result of, or coincident with, one or more mutations that differ from those of another patient having the same type of cancer. The fact that there are often multiple molecular bases underlying the same cancers is consistent with the observation that some therapies that affect one patient do not necessarily equally affect another patient with the same type of cancer. Further, from a diagnostic point of view, the presence of particular mutations such as translocations, deletions, or SNPs can have powerful implications. In some instances, such molecular markers are themselves useful indicators for diagnosis, prognosis, or treatment response determinations. This is particularly true where the molecular mutations can be associated with response to particular treatments.

A Biomarker is any indicia of the level of expression of an indicated Marker gene. The indicia can be direct or indirect and measure over- or under-expression of the gene given the physiologic parameters and in comparison to an internal control, normal tissue or another carcinoma. Biomarkers include, without limitation, nucleic acids (both over and under-expression and direct and indirect). Using nucleic acids as Biomarkers can include any method known in the art including, without limitation, measuring DNA amplification, RNA, micro RNA, loss of heterozygosity (LOH), single nucleotide polymorphisms (SNPs, Brookes (1999)), microsatellite DNA, DNA hypo- or hyper-methylation. Using proteins as Biomarkers can include any method known in the art including, without limitation, measuring amount, activity, modifications such as glycosylation, phosphorylation, ADP-ribosylation, ubiquitination, etc., imunohistochemistry (IHC). Other Biomarkers include imaging, cell count and apoptosis markers.

The indicated genes provided herein are those associated with a particular tumor or tissue type. A Marker gene may be associated with numerous cancer types but provided that the expression of the gene is sufficiently associated with one tumor or tissue type to be identified using the algorithm described herein to be specific for a lung cancer cell, the gene can be used in the claimed invention to determine cancer status and prognosis. Numerous genes associated with one or more cancers are known in the art. The present invention provides preferred Marker genes and even more preferred Marker gene combinations. These are described herein in detail. A Marker gene corresponds to the sequence designated by a SEQ ID NO when it contains that sequence. A gene segment or fragment corresponds to the sequence of such gene when it contains a portion of the referenced sequence or its complement sufficient to distinguish it as being the sequence of the gene. A gene expression product corresponds to such sequence when its RNA, mRNA, or cDNA hybridizes to the composition having such sequence (e.g. a probe) or, in the case of a peptide or protein, it is encoded by such mRNA. A segment or fragment of a gene expression product corresponds to the sequence of such gene or gene expression product when it contains a portion of the referenced gene expression product or its complement sufficient to distinguish it as being the sequence of the gene or gene expression product.

The inventive methods, compositions, articles, and kits of described and claimed in this specification include one or more Marker genes. "Marker" or "Marker gene" is used throughout this specification to refer to genes and gene expression products that correspond with any gene the over- or under-expression of which is associated with a tumor or tissue type. The preferred Marker genes are described in more detail in Table 8.

The present invention provides a method of assessing acute myeloid leukemia (AML) status by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 where the expression levels of the Marker genes above or below pre-determined cut-off levels are indicative of AML status.

The present invention provides a method of staging acute myeloid leukemia (AML) patients by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 where the expression levels of the Marker genes above or below pre-determined cut-off levels are indicative of AML survival. The present invention provides a method of determining acute myeloid leukemia (AML) patient treatment protocol by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 where the expression levels of the Marker genes above or below pre-determined cut-off levels are sufficiently indicative of response to therapy to enable a physician to determine the degree and type of therapy recommended to provide appropriate therapy.

The present invention provides a method of treating a acute myeloid leukemia (AML) patient by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 where the expression levels of the Marker genes above or below pre-determined cut-off levels are indicate a response to therapy and; treating the patient with adjuvant therapy if they have a responder profile. The present invention provides a method of determining whether a acute myeloid leukemia (AML) patient is high or low risk of mortality by obtaining a biological sample from an AML patient; and measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3 where the expression levels of the Marker genes above or below pre-determined cut-off levels are sufficiently indicative of risk of mortality to enable a physician to determine the degree and type of therapy recommended.

The method provided herein may further include, contain or utilize measuring the expression level of at least one gene constitutively expressed in the sample. Preferably, the method provided herein results in a specificity of at least about 40%. Preferably, the method provided herein results in a sensitivity of at least at least about 80%. Preferably, the method provided herein results in a p-value of less than 0.05.

The method provided herein may be performed by measuring gene expression on a microarray or gene chip. The microarray can be a cDNA array or an oligonucleotide array and may further contain one or more internal control reagents. The method provided herein may be performed by determining gene expression by nucleic acid amplification conducted by polymerase chain reaction (PCR) of RNA extracted from the sample. The PCR can be reverse transcription polymerase chain reaction (RT-PCR) and can further contain one or more internal control reagent.

The method provided herein may be performed by measuring or detecting a protein encoded by the gene. Te protein can be detected by an antibody specific to the protein.

The method provided herein may be performed by measuring a characteristic of the gene. Characteristics include, without limitation, DNA amplification, methylation, mutation and allelic variation.

The present invention provides a method of generating an acute myeloid leukemia (AML) prognostic patient report by determining the results of any one of the above-described methods; and preparing a report displaying the results and reports generated thereby. The report may contain an assessment of patient outcome and/or probability of risk relative to the patient population and/or likelihood or response to chemotherapy. The present invention provides a kit for conducting an assay to determine acute myeloid leukemia (AML) prognosis in a biological sample comprising: materials for detecting isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9. The kit can further contain reagents for conducting a microarray analysis and/or a medium through which said nucleic acid sequences, their complements, or portions thereof are assayed.

The present invention provides articles for assessing acute myeloid leukemia (AML) status containing materials for detecting isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9. The articles can further contain reagents for conducting a microarray analysis and/or a medium through which said nucleic acid sequences, their complements, or portions thereof are assayed. The present invention provides a microarray or gene chip for performing the above-described methods. The microarray may contain isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9. Preferably, the microarray provides a measurement or characterization at least 1.5-fold over- or under-expression. Preferably, the microarray provides a measurement with a statistically significant p-value over- or under-expression. More preferably, the p-value is less than 0.05. The microarray can be any known in the art including, without limitation, cDNA array or an oligonucleotide array and can further contain internal control reagents.

The present invention provides a diagnostic/prognostic portfolio comprising isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9. Preferably, the measurement or characterization is at least 1.5-fold over- or under-expression. Preferably, the measurement provides a statistically significant p-value over- or under-expression. More preferably, the p-value is less than 0.05. Preferred methods for establishing gene expression profiles include determining the amount of RNA that is produced by a gene that can code for a protein or peptide. This is accomplished by reverse transcriptase PCR (RT-PCR), competitive RT-PCR, real time RT-PCR, differential display RT-PCR, Northern Blot analysis and other related tests. While it is possible to conduct these techniques using individual PCR reactions, it is best to amplify complementary DNA (cDNA) or complementary RNA (cRNA) produced from mRNA and analyze it via microarray. A number of different array configurations and methods for their production are known to those of skill in the art and are described in U.S. Patents such as: 5,445,934; 5,532,128; 5,556,752; 5,242,974; 5,384,261 ; 5,405,783; 5,412,087; 5,424,186; 5,429,807; 5,436,327; 5,472,672; 5,527,681; 5,529,756; 5,545,531; 5,554,501 ; 5,561,071; 5,571,639; 5,593,839; 5,599,695; 5,624,711; 5,658,734; and 5,700,637.

Microarray technology allows for the measurement of the steady-state mRNA level of thousands of genes simultaneously thereby presenting a powerful tool for identifying effects such as the onset, arrest, or modulation of uncontrolled cell proliferation. Two microarray technologies are currently in wide use. The first are cDNA arrays and the second are oligonucleotide arrays. Although differences exist in the construction of these chips, essentially all downstream data analysis and output are the same. The product of these analyses are typically measurements of the intensity of the signal received from a labeled probe used to detect a cDNA sequence from the sample that hybridizes to a nucleic acid sequence at a known location on the microarray. Typically, the intensity of the signal is proportional to the quantity of cDNA, and thus mRNA, expressed in the sample cells. A large number of such techniques are available and useful. Preferred methods for determining gene expression can be found in US Patents 6,271,002; 6,218,122; 6,218,1 14; and 6,004,755.

Analysis of the expression levels is conducted by comparing such signal intensities. This is best done by generating a ratio matrix of the expression intensities of genes in a test sample versus those in a control sample. For instance, the gene expression intensities from a diseased tissue can be compared with the expression intensities generated from benign or normal tissue of the same type. A ratio of these expression intensities indicates the fold-change in gene expression between the test and control samples.

Gene expression profiles can also be displayed in a number of ways. The most common method is to arrange raw fluorescence intensities or ratio matrix into a graphical dendogram where columns indicate test samples and rows indicate genes. The data are arranged so genes that have similar expression profiles are proximal to each other. The expression ratio for each gene is visualized as a color. For example, a ratio less than one (down-regulation) may appear in the blue portion of the spectrum while a ratio greater than one (indicating up-regulation) may appear as a color in the red portion of the spectrum. Commercially available computer software programs are available to display such data including "GENESPRING" from Silicon Genetics, Inc. and "DISCOVERY" and "INFER" software from Partek, Inc. In the case of measuring protein levels to determine gene expression, any method known in the art is suitable provided it results in adequate specificity and sensitivity. For example, protein levels can be measured by binding to an antibody or antibody fragment specific for the protein and measuring the amount of antibody-bound protein. Antibodies can be labeled by radioactive, fluorescent or other detectable reagents to facilitate detection. Methods of detection include, without limitation, enzyme-linked immunosorbent assay (ELISA) and immunoblot techniques.

Modulated Markers used in the methods of the invention are described in the Examples. The genes that are differentially expressed are either up regulated or down regulated in patients with various lung cancer prognostics. Up regulation and down regulation are relative terms meaning that a detectable difference (beyond the contribution of noise in the system used to measure it) is found in the amount of expression of the genes relative to some baseline. In this case, the baseline is determined based on the algorithm. The genes of interest in the diseased cells are then either up- or down-regulated relative to the baseline level using the same measurement method.

Assays for the gene expression status of a cell also can determine normal/abnormal tissue distribution for diagnostic purposes using techniques such as immunohistochemical analysis (IHC). Any method known in the art can be used, for example in the case of the LBC oncogene, the antibodies to LBC protein may be used in conjunction with both fresh-frozen and formalin-fixed, paraffin-embedded tissue blocks prepared for study by IHC. Each tissue block may consist of 50 mg of residual "pulverized" tumor.

Briefly, frozen-sections may be prepared by rehydrating 50 ng of frozen pulverized tumor at room temperature in phosphate buffered saline (PBS) in small plastic capsules; pelleting the particles by centrifugation; resuspending them in a viscous embedding medium (OCT); inverting the capsule and pelleting again by centrifugation; snap-freezing in -70°C isopentane; cutting the plastic capsule and removing the frozen cylinder of tissue; securing the tissue cylinder on a cryostat microtome chuck; and cutting 25-50 serial sections containing intact tumor cells. Permanent-sections may be prepared by a similar method involving rehydration of the 50 mg sample in a plastic microfuge tube; pelleting; resuspending in 10% formalin for 4 hr fixation; washing/pelleting; resuspending in warm 2.5% agar; pelleting; cooling in ice water to harden the agar; removing the tissue/agar block from the tube; infiltrating and embedding the block in paraffin; and cutting up to 50 serial permanent sections.

For the IHC assay, the sections are overlaid with a blocking solution containing: 3% bovine serum albumin (BSA) in PBS or other blocking reagents. The blocking reagents include non-specific serum or dry milk. Blocking is allowed to proceed for 1 hr at room temperature. Anti-LBC protein antibody is diluted with PBS buffer containing 3% BSA, 0.1% TritonX™-100 and t-octylphenoxypolyethoxyethanol, at a ratio of 1 : 100. The sample sections are generally overlaid with the antibody solution for 16 hr at 4°C. The duration and temperature conditions may be varied according to the antibody selected and the material tested. The optimal conditions are determined empirically. The antibody treated sections are then washed three times in PBS for 15 min. each to remove unbound antibody and then overlaid with PBS containing 3% BSA and a secondary antibody at a dilution of 1 :2000. The secondary antibodies may be coupled to a chromogenic enzyme such as: horseradish peroxidase, alkaline phosphatase, fluorescein isothiocyanate, or other suitable enzymes. Alternatively, the secondary antibody may be conjugated to biotin and used in conjunction with chromophore-labeled avidin.

Another exemplary method for detecting the presence of a gene is via in situ hybridization. Generally, in situ hybridization comprises the following major steps: (1) fixation of tissue or biological structure to be analyzed; (2) prehybridization treatment of the biological structure to increase accessibility of target DNA, and to reduce nonspecific binding; (3) hybridization of the mixture of nucleic acids to the nucleic acid in the biological structure or tissue; (4) post-hybridization washes to remove nucleic acid fragments not bound in the hybridization and (5) detection of the hybridized nucleic acid fragments. The reagent used in each of these steps and the conditions for use vary depending on the particular application.

In this case, a hybridization solution comprising at least one detectable nucleic acid probe capable of hybridizing to a gene (at its chromosomal locus) is contacted with the cell under hybridization conditions. Any hybridization is then detected and compared to a predetermined hybridization pattern from normal or control cells. Preferably, the probes are alpha-centromeric probes. Such probes can be made commercially available from a number of sources (e.g., from Visys Inc., Downers Grove, IL). In a preferred embodiment, the hybridization solution contains a multiplicity of probes, specific for an area on the chromosome that corresponds to the translocation of the sequences that make up the chimera (e.g., 15q24-25). Hybridization protocols suitable for use with the methods of the invention are described, e.g., in Albertson (1984); Pinkel (1988); EP No. 430,402; and Methods in Molecular Biology, Vol. 33: In Situ Hybridization Protocols, Choo, ed., Humana Press, Totowa, NJ (1994), etc. In one particularly preferred embodiment, the hybridization protocol of Pinkel et al. (1998) or of Kallioniemi (1992) is used. Methods of optimizing hybridization conditions are well known (see, e.g., Tijssen (1993) Laboratory Techniques in Biochemistry and Molecular Biology, Vol. 24: Hybridization With Nucleic Acid Probes, Elsevier, NY).

In a preferred embodiment, background signal is reduced by the use of a detergent (e.g., C-TAB) or a blocking reagent (e.g., sperm DNA, cot-1 DNA, etc.) during the hybridization to reduce non-specific binding. Preferably, the hybridization is performed in the presence of about 0.1 to about 0.5 mg/ml DNA (e.g., cot-1 DNA).

The probes may be prepared by any method known in the art, including synthetically or grown in a biological host. Synthetic methods include but are not limited to oligonucleotide synthesis, riboprobes, and PCR.

The probe may be labeled with a detectable marker by any method known in the art. Methods for labeling probes include random priming, end labeling, PCR and nick translation. Enzymatic labeling is conducted in the presence of nucleic acid polymerase, three unlabeled nucleotides, and a fourth nucleotide which is either directly labeled, contains a linker arm for attaching a label, or is attached to a hapten or other molecule to which a labeled binding molecule may bind. Suitable direct labels include radioactive labels such as ³²P, ³H, and ³⁵S and non-radioactive labels such as fluorescent markers, such as fluorescein, Texas Red, AMCA blue, lucifer yellow, rhodamine, and the like; cyanin dyes which are detectable with visible light; enzymes and the like. Labels may also be incorporated chemically into DNA probes by bisulfite-mediated transamination or directly during oligonucleotide synthesis.

Fluorescent markers can readily be attached to nucleotides with activated linker arms incorporated into the probe. Probes may be indirectly labeled by the methods disclosed above, by incorporating a nucleotide covalently linked to a hapten or other molecule such as biotin or digoxygenin, and performing a sandwich hybridization with a labeled antibody directed to that hapten or other molecule, or in the case of biotin, with avidin conjugated to a detectable label. Antibodies and avidin may be conjugated with a fluorescent marker, or with an enzymatic marker such as alkaline phosphatase or horseradish peroxidase to render them detectable. Conjugated avidin and antibodies are commercially available from companies such as Vector

Laboratories (Burlingame, CA) and Boehringer Mannheim (Indianapolis, IN). The enzyme can be detected through a colorimetric reaction by providing a substrate for the enzyme. In the presence of various substrates, different colors are produced by the reaction, and these colors can be visualized to separately detect multiple probes. Any substrate known in the art may be used. Preferred substrates for alkaline phosphatase include 5-bromo-4-chloro-3-indolylphosphate (BCIP) and nitro blue tetrazolium (NBT). The preferred substrate for horseradish peroxidase is diaminobenzoate (DAB). Fluorescently labeled probes suitable for use in the in situ hybridization methods of the invention are preferably in the range of 150-500 nucleotides long. Probes may be DNA or RNA, preferably DNA.

Hybridization of the detectable probes to the cells is conducted with a probe concentration of 0.1-500 ng/μL, preferably 5-250 ng/μL. The hybridization mixture will preferably contain a denaturing agent such as formamide. In general, hybridization is carried out at 25°C-45°C, more preferably at 32°C-40°C, and most preferably at 37°C-38°C. The time required for hybridization is about 0.25-96 hours, more preferably 1 -72 hours, and most preferably for 4-24 hours. Hybridization time will vary based on probe concentration and hybridization solution content which may contain accelerators such as hnRNP binding protein, trialkyl ammonium salts, lactams, and the like. Slides are then washed with solutions containing a denaturing agent, such as formamide, and decreasing concentrations of sodium chloride or in any solution that removes unbound and mismatched probe. The temperature and concentration of salt will vary depending on the stringency of hybridization desired. For example, high stringency washes may be carried out at 42°C-68°C, while intermediate stringency may be in the range of 37°C-55°C, and low stringency may be in the range of 30°C-37°C. Salt concentration for a high stringency wash may be 0.5-1 times SSC (0.15M NaCl, 0.015M Na citrate), while medium stringency may be 1-4 times, and low stringency may be 2-6 times SSC.

The detection incubation steps, if required, should preferably be carried out in a moist chamber at 23°C-42°C, more preferably at 25°C-38°C and most preferably at 37-38°C. Labeled reagents should preferably be diluted in a solution containing a blocking reagent, such as BSA, non-fat dry milk, or the like. Dilutions may range from 1 :10-1 :10,000, more preferably 1 :50-1 :5,000, and most preferably at 1 :100-

1 : 1 ,000. The slides or other solid support should be washed between each incubation step to remove excess reagent.

Slides may then be mounted and analyzed by microscopy in the case of a visible detectable marker, or by exposure to autoradiographic film in the case of a radioactive marker. In the case of a fluorescent marker, slides are preferably mounted in a solution that contains an antifade reagent, and analyzed using a fluorescence microscope. Multiple nuclei may be examined for increased accuracy of detection.

Additionally, assays for the expression product of the LBC oncogene can also be used to determine whether the LBC oncogene mutation has occurred. Most preferably, such assays are immunoassays. Immunoassays, in their most simple and direct sense, are binding assays. Certain preferred immunoassays are the various types of enzyme linked immunosorbent assays (ELISAs) and radioimmunoassays (RIA) known in the art. IHC detection using tissue sections is also particularly useful as are in situ hybridization and enzyme immunoassay.

In one exemplary ELISA, protein-specific antibodies are immobilized onto a selected surface exhibiting protein affinity, such as a well in a polystyrene microtiter plate. Then, a test composition containing the desired antigen, such as a clinical sample, is added to the wells. After binding and washing to remove non-specifϊcally bound immune complexes, the bound antigen may be detected. Detection is generally achieved by the addition of another antibody, specific for the desired antigen, that is linked to a detectable label. This type of ELISA is a simple "sandwich ELISA." Detection may also be achieved by the addition of a second antibody specific for the desired antigen, followed by the addition of a third antibody that has binding affinity for the second antibody, with the third antibody being linked to a detectable label.

Variations of ELISA techniques are well known. Pn one such variation, the samples containing the desired antigen are immobilized onto the well surface and then contacted with the antibodies of the invention. After binding and appropriate washing, the bound immune complexes are detected. Where the initial antigen specific antibodies are linked to a detectable label, the immune complexes may be detected directly. Again, the immune complexes may be detected using a second antibody that has binding affinity for the first antigen specific antibody, with the second antibody being linked to a detectable label.

In embodiments of the invention in which gene expression is detected for determining AML prognosis or status, the use of gene expression portfolios is most preferred. A portfolio of genes is a set of genes grouped so that expression information obtained about them provides the basis for making a clinically relevant judgment such as a diagnosis, prognosis, or treatment choice. In this case, gene expression portfolios can be fashioned to help make therapeutic decisions regarding AML patients.

Diseased, in this context, refers to an alteration of the state of a body that interrupts or disturbs, or has the potential to disturb, proper performance of bodily functions as occurs with the uncontrolled proliferation of cells. Someone is diagnosed with a disease when some aspect of that person's genotype or phenotype is consistent with the presence of the disease. However, the act of conducting a diagnosis or prognosis may include the determination of disease/status issues such as determining the likelihood of relapse, type of therapy and therapy monitoring. In therapy monitoring, clinical judgments are made regarding the effect of a given course of therapy by comparing the expression of genes over time to determine whether the gene expression profiles have changed or are changing to patterns more consistent with normal tissue.

Genes can be grouped so that information obtained about the set of genes in the group provides a sound basis for making a clinically relevant judgment such as a diagnosis, prognosis, or treatment choice. These sets of genes make up the portfolios of the invention. As with most diagnostic markers, it is often desirable to use the fewest number of markers sufficient to make a correct medical judgment. This prevents a delay in treatment pending further analysis as well unproductive use of time and resources. One method of establishing gene expression portfolios is through the use of optimization algorithms such as the mean variance algorithm widely used in establishing stock portfolios. This method is described in detail in US patent publication number 20030194734. Essentially, the method calls for the establishment of a set of inputs (stocks in financial applications, expression as measured by intensity here) that will optimize the return (e.g., signal that is generated) one receives for using it while minimizing the variability of the return. Many commercial software programs are available to conduct such operations. "Wagner Associates Mean- Variance Optimization Application," referred to as "Wagner Software" throughout this specification, is preferred. This software uses functions from the "Wagner Associates Mean- Variance Optimization Library" to determine an efficient frontier and optimal portfolios in the Markowitz sense is one option. Use of this type of software requires that microarray data be transformed so that it can be treated as an input in the way stock return and risk measurements are used when the software is used for its intended financial analysis purposes. The process of selecting a portfolio can also include the application of heuristic rules. Preferably, such rules are formulated based on biology and an understanding of the technology used to produce clinical results. More preferably, they are applied to output from the optimization method. For example, the mean variance method of portfolio selection can be applied to microarray data for a number of genes differentially expressed in subjects with cancer. Output from the method would be an optimized set of genes that could include some genes that are expressed in peripheral blood as well as in diseased tissue. If samples used in the testing method are obtained from peripheral blood and certain genes differentially expressed in instances of cancer could also be differentially expressed in peripheral blood, then a heuristic rule can be applied in which a portfolio is selected from the efficient frontier excluding those that are differentially expressed in peripheral blood. Of course, the rule can be applied prior to the formation of the efficient frontier by, for example, applying the rule during data pre-selection. Other heuristic rules can be applied that are not necessarily related to the biology in question. For example, one can apply a rule that only a prescribed percentage of the portfolio can be represented by a particular gene or group of genes. Commercially available software such as the Wagner Software readily accommodates these types of heuristics. This can be useful, for example, when factors other than accuracy and precision (e.g., anticipated licensing fees) have an impact on the desirability of including one or more genes.

The gene expression profiles of this invention can also be used in conjunction with other non-genetic diagnostic methods useful in cancer diagnosis, prognosis, or treatment monitoring. For example, in some circumstances it is beneficial to combine the diagnostic power of the gene expression based methods described above with data from conventional markers such as serum protein markers (e.g., Cancer Antigen 27.29 ("CA 27.29"))- A range of such markers exists including such analytes as CA 27.29. In one such method, blood is periodically taken from a treated patient and then subjected to an enzyme immunoassay for one of the serum markers described above. When the concentration of the marker suggests the return of tumors or failure of therapy, a sample source amenable to gene expression analysis is taken. Where a suspicious mass exists, a fine needle aspirate (FNA) is taken and gene expression profiles of cells taken from the mass are then analyzed as described above. Alternatively, tissue samples may be taken from areas adjacent to the tissue from which a tumor was previously removed. This approach can be particularly useful when other testing produces ambiguous results.

Kits made according to the invention include formatted assays for determining the gene expression profiles. These can include all or some of the materials needed to conduct the assays such as reagents and instructions and a medium through which Biomarkers are assayed.

Preferred methods for establishing gene expression profiles (including those used to arrive at the explication of the relevant biological pathways) include determining the amount of RNA that is produced by a gene that can code for a protein or peptide or transcribe RNA. This is best accomplished by reverse transcription PCR (RT- PCR), competitive RT-PCR, real time RT-PCR, differential display RT-PCR, Northern Blot analysis and other related tests. While it is possible to conduct these techniques using individual PCR reactions, it is often desirable to amplify copy DNA (cDNA) or copy RNA (cRNA) produced from mRNA and analyze it via microarray. A number of different array configurations and production methods are known to those of skill in the art and are described in US Patents such as: 5,445,934; 5,532,128; 5,556,752; 5,242,974; 5,384,261; 5,405,783; 5,412,087; 5,424,186; 5,429,807; 5,436,327; 5,472,672; 5,527,681 ; 5,529,756; 5,545,531 ; 5,554,501; 5,561,071; 5,571,639; 5,593,839; 5,599,695; 5,624,711 ; 5,658,734; and 5,700,637.

Microarray technology measures steady-state mRNA levels of thousands of genes simultaneously thereby presenting a powerful tool for identifying AML patient gene expression profiles. Two microarray technologies are currently in wide use. The first are cDNA arrays and the second are oligonucleotide arrays. Although differences exist in the construction of these chips, essentially all downstream data analysis and output are the same. The products of these analyses are typically measurements of the intensity of the signal received from a labeled probe used to detect a cDNA sequence from the sample that hybridizes to a nucleic acid sequence at a known location on the microarray. Typically, the signal intensity is proportional to the cDNA quantity, and thus mRNA, expressed in the sample cells. A large number of such techniques are available and useful. Preferred methods can be found in US Patents 6,271,002; 6,218,122; 6,218,1 14; and 6,004,755.

Analysis of the expression levels is conducted by comparing such intensities. This is best done by generating a ratio matrix of the expression intensities of genes in a test sample versus those in a control sample. For instance, the gene expression intensities from a tissue that has been treated with a drug can be compared with the expression intensities generated from the same tissue that has not been treated with the drug. A ratio of these expression intensities indicates the fold-change in gene expression between the test and control samples. Gene expression profiles can be displayed in a number of ways. A common method is to arrange a ratio matrix into a graphical dendogram where columns indicate test samples and rows indicate genes. The data are arranged so genes that have similar expression profiles are proximal to each other. The expression ratio for each gene is visualized as a color. For example, a ratio less than one (indicating down-regulation) may appear in the blue portion of the spectrum while a ratio greater than one (indicating up-regulation) may appear as a color in the red portion of the spectrum. Commercially available computer software programs are available to display such data including "GENESPR1NT" from Silicon Genetics, Inc. and "DISCOVERY" and "INFER" software from Partek, Inc.

The differentially expressed genes are either up regulated or down regulated in diseased cells, as deduced by an assessment of gene expression as described above. Up regulation and down regulation are relative terms meaning that a detectable difference (beyond the contribution of noise in the system used to measure it) is found in the amount of expression of the genes relative to some baseline. In this case, the baseline is the measured gene expression of a normal cell. The genes of interest in the diseased cells are then either up regulated or down regulated relative to the baseline level using the same measurement method. Preferably, levels of up and down regulation are distinguished based on fold changes of the intensity measurements of hybridized microarray probes. A 1.5 fold difference is preferred for making such distinctions. That is, before a gene is said to be differentially expressed in treated versus untreated diseased cells, the treated cell is found to yield at least 1.5 times more, or 1.5 times less intensity than the untreated cells. A 1.7 fold difference is more preferred and a 2 or more fold difference in gene expression measurement is most preferred.

One method of the invention involves comparing gene expression profiles for various genes to determine whether a person is likely to respond to the use of a therapeutic agent. Having established the gene expression profiles that distinguish responder from non-responder, the gene expression profiles of each are fixed in a medium such as a computer readable medium as described below. A patient sample is obtained that contains diseased cells (such as hematopoietic blast cells in the case of AML) is then obtained. Most preferably, the samples are of bone marrow and are extracted from the patient's sternum or iliac crest according to routine methods. Preferably the bone marrow aspirate is processed to enrich for leukemic blast cells using routine methods. Sample RNA is then obtained and amplified from the diseased patient cells and a gene expression profile is obtained, preferably (in the case of a large gene portfolio) via micro-array, for genes in the appropriate portfolios. The expression profiles of the samples are then compared to those previously analyzed for prognostic outcome. When a small number of genes are used in the portfolio such as when the three gene profile is used, a simple nucleic acid amplification and detection scheme is the most preferred method of measuring gene modulation. In such a case, PCR, NASBA, rolling circle, LCR, and other amplification schemes known to skilled artisans can be used with PCR being most preferred. Where the portfolios include a large number of genes or it is desirable to measure the expression of numerous other genes then it is preferred to assess the expression patterns based on intensity measurements of microarrays as described above.

In similar fashion, gene expression profile analysis can be conducted to monitor treatment response. In one aspect of this method, gene expression analysis as described above is conducted on a patient treated with any suitable treatment at various periods throughout the course of treatment. If the gene expression patterns are consistent with a positive outcome the patient's therapy is continued. If it is not, the patient's therapy is altered as with additional therapeutics, changes to the dosage, or elimination of the current treatment. Such analysis permits intervention and therapy adjustment prior to detectable clinical indicia or in the face of otherwise ambiguous clinical indicia.

With respect to the molecular markers of the invention, a number of other formats and approaches are available for diagnostic use. Methylation of genomic regions can affect gene expression levels. For example, hypermethylation of gene promoter regions can constitutively down-regulate gene expression whereas hypomethylation can lead to an increase in steady-state mRNA levels. As such, detection of methylated regions associated with genes predictive of drug response, prognosis or status can be used as an alternative method for diagnosing gene expression levels. Such methods are known to those skilled in the art. Alternatively, single nucleotide polymorphisms (SNPs) that are present in promoter regions can also affect transcriptional activity of a gene. Therefore, detection of these SNPs by methods known to those skilled in the art can also be used as a diagnostic for detecting genes that are differentially expressed in different prognostic outcomes. Articles of this invention are representations of the gene expression profiles useful for treating, diagnosing, prognosticating, staging, and otherwise assessing diseases. Preferably they are reduced to a medium that can be automatically read such as computer readable media (magnetic, optical, and the like). The articles can also include instructions for assessing the gene expression profiles in such media. For example, the articles may comprise a CD ROM having computer instructions for comparing gene expression profiles of the portfolios of genes described above. The articles may also have gene expression profiles digitally recorded therein so that they may be compared with gene expression data from patient samples. Alternatively, the profiles can be recorded in different representational format. Clustering algorithms such as those incorporated in "DISCOVERY" and "INFER" software from Partek, Inc. mentioned above can best assist in the visualization of such data.

Additional articles according to the invention are kits for conducting the assays described above. Each such kit would preferably include instructions in human or machine readable form as well as the reagents typical for the type of assay described. These can include, for example, nucleic acid arrays (e.g. cDNA or oligonucleotide arrays), as described above, configured to discern the gene expression profiles of the invention. They can also contain reagents used to conduct nucleic acid amplification and detection including, for example, reverse transcriptase, reverse transcriptase primer, a corresponding PCR primer set, a thermostable DNA polymerase, such as Taq polymerase, and a suitable detection reagent(s), such as, without limitation, a scorpion probe, a probe for a fluorescent probe assay, a molecular beacon probe, a single dye primer or a fluorescent dye specific to double-stranded DNA, such as ethidium bromide. Kits for detecting surface antigens contain staining materials or are antibody based including components such as buffer, anti-antigenic antibody, detection enzyme and substrate such as Horse Radish Peroxidase or biotin-avidin based reagents. Kit components for detecting blast cells generally include reagents for conducting flow cytometry, blast cell adhesion assays, and other common blast cell assays. Conventional anti-cancer agents include, without limitation, tyrosine kinase inhibitors, MEK kinase inhibitors, P13K kinase inhibitors, MAP kinase inhibitors, apoptosis modulators and combinations thereof. Exemplary drugs that are most preferred among these are the "GLEEVEC" tyrosine kinase inhibitor of Novartis, U-0126 MAP kinase inhibitor, PD-098059 MAP kinase inhibitor, SB-203580 MAP kinase inhibitor, and antisense, ribozyme, and DNAzyme, BcI-XL, and anti- apoptotics. Examples of other useful drugs include, without limitation, the calanolides of US Patent 6,306,897; the substituted bicyclics of US Patent 6,284,764; the indolines of US Patent 6,133,305; and the antisense oligonucleotides of US Patent 6,271,210; platinum coordination compounds for example cisplatin or carboplatin, taxane compounds for example paclitaxel or docetaxel, camptothecin compounds for example irinotecan or topotecan, anti-tumor vinca alkaloids for example vinblastine, vincristine or vinorelbine, anti-tumor nucleoside derivatives for example 5-fluorouracil, gemcitabine or capecitabine, nitrogen mustard or nitrosourea alkylating agents for example cyclophosphamide, chlorambucil, carmustine or lomustine, anti-tumor anthracycline derivatives for example daunorubicin, doxorubicin or idarubicin; HER2 antibodies for example trastzumab; and anti-tumor podophyllotoxin derivatives for example etoposide or teniposide; and antiestrogen agents including estrogen receptor antagonists or selective estrogen receptor modulators preferably tamoxifen, or alternatively toremifene, droloxifene, faslodex and raloxifene, or aromatase inhibitors such as exemestane, anastrozole, letrazole and vorozole.

Anti-cancer agents can also include therapeutics directed to gene therapy or antisense therapy or RNA interference. These include, without limitation, oligonucleotides with sequences complementary to an mRNA sequence can be introduced into cells to block the translation of the mRNA, thus blocking the function of the gene encoding the mRNA. The use of oligonucleotides to block gene expression is described, for example, in, Strachan and Read, Human Molecular Genetics, 1996. These antisense molecules may be DNA, stable derivatives of DNA such as phosphorothioates or methylphosphonates, RNA, stable derivatives of RNA such as 2'-O-alkylRNA, or other antisense oligonucleotide mimetics. Antisense molecules may be introduced into cells by microinjection, liposome encapsulation or by expression from vectors harboring the antisense sequence.

In gene therapy, the gene of interest can be ligated into viral vectors that mediate transfer of the therapeutic DNA by infection of recipient host cells. Suitable viral vectors include retrovirus, adenovirus, adeno-associated virus, herpes virus, vaccinia virus, polio virus and the like. Alternatively, therapeutic DNA can be transferred into cells for gene therapy by non-viral techniques including receptor-mediated targeted DNA transfer using ligand-DNA conjugates or adenovirus-ligand-DNA conjugates, lipofection membrane fusion or direct microinjection. These procedures and variations thereof are suitable for ex vivo as well as in vivo gene therapy. Protocols for molecular methodology of gene therapy suitable for use with the gene is described in Gene Therapy Protocols, edited by Paul D. Robbins, Human press, Totowa NJ, 1996.

Compounds identified according to the methods disclosed herein may be used alone at appropriate dosages defined by routine testing in order to obtain optimal inhibition or activity while minimizing any potential toxicity. In addition, coadministration or sequential administration of other agents may be desirable. The invention is further illustrated by the following nonlimiting examples. All references cited herein are hereby incorporated herein by reference.

Example 1

Clinical Evaluation and Response Definitions

The current study was part of an open label, multicenter, non-comparative phase 2 clinical study in which patients with relapsed or refractory AML (Harousseau et al. (2003)) were treated with tipifarnib at a starting oral dose of 600 mg bid for the first 21 consecutive days of each 28-day cycle. Patients were enrolled into 2 cohorts, those with relapsed AML and those with refractory AML. A total of 252 patients (135 relapsed and 1 17 refractory) were treated. Eighty patients chose to provide bone marrow samples for RNA microarray analysis, for which a separate informed consent was required. The overall response rate was relatively low in this study.

Therefore, for the purposes of the gene expression profiling, response to tipifarnib was defined as patients who had an objective response (complete remission [CR], complete remission with incomplete platelet recovery [CRp] or partial remission [PR]), a hematological response (decrease of >50% of leukemic blast cells in bone marrow) as determined by either central review or by the clinical site, or stable disease (no hematological response but no progression of the disease) as determined by both central review and the clinical site. Complete remission with incomplete platelet recovery was defined similarly, except for a platelet count less than 100,000/μL sufficient to ensure transfusion independence. Partial remission was defined as at least a 50% decrease in bone marrow blasts with partial neutrophil (>500/μL) and platelet count (>50,000/μL) recovery. Response had to be confirmed at least 4 weeks after first documentation. Sample Collection and Microarrav Processing

Bone marrow samples were collected from patients before treatment with tipifarnib, diluted with phosphate buffered saline (PBS) and centrifuged with Ficoll- diatrizoate (1.077 g/mL). White blood cells were washed twice with PBS, resuspended in fetal bovine serum (FBS) with 10% dimethyl sulfoxide (DMSO) and immediately stored at -80⁰C. Cells were thawed and total RNA was extracted from cell samples using the RNeasy Kit (Qiagen, Valencia, CA). RNA quality was checked using the Agilent Bioanalyzer. Synthesis of cDNA and cRNA was performed according to Affymetrix (Santa Clara, CA) protocols. Microarrav Processing

Two rounds of linear amplification were performed because the RNA yield for several samples was too low to obtain enough labeled cRNA for chip hybridization using one round of amplification. For hybridization, 11 μg ofcRNA were fragmented randomly by incubation at 94°C for 35 minutes in 40 mM Tris-acetate, pH 8.1 , 100 mM potassium acetate, and 30 mM magnesium acetate. Fragmented cRNA was hybridized to U133A arrays at 45°C for 16 hours in a rotisserie oven set at 60 rpm. Following hybridization, arrays were washed (with 6x SSPE and 0.5x SSPE containing Triton X-100 [0.005%]), and stained with streptavidin- phycoerythrin (SAPE; Molecular Probes, Eugene, OR). Quantification of bound labeled probe was conducted using the Agilent G2500A GeneArray scanner (Agilent Technologies, Palo Alto, CA).

The total fluorescence intensity for each array was scaled to the uniform value of 600. Chip performance was quantitated by calculating a signal to noise ratio (raw average signal/noise). Chips were removed from further analysis if their signal-to- noise ratio was less than 5. Genes were only included in further analysis if they were called "present" in at least 10% of the chips. Eleven thousand seven hundred twenty three Affymetrix probe sets remained following this cut-off. The quality of the gene expression data were further controlled by identifying outliers based on principal components analysis and by analyzing the normal distributions of the gene intensities (Partek Pro V5.1). Statistical Analysis

To identify genes that predict response with high sensitivity, a percentile analysis was employed. Genes that were up- or down-regulated in 100% of responders compared to at least 40% of non-responders were identified. The chi- squared test and Student's t-test were then used to test the significance of the correlations between patient response and patient co-variates, including ras mutation status, and gene expression. Unsupervised k-means and hierarchical clustering were performed in Omniviz. The predictive value of the selected genes was then analyzed by leave-one-out and leave-five-out cross validation methods. Here, one (or five) sample(s) was (were) removed from the data set and the marker was reselected from 1 1,723 genes. The predictive value of this gene was then tested on the left-out sample(s) using a linear discriminant analysis. Sensitivity was calculated as the number of true positives detected by the test divided by the sum of true positives plus false negatives. Specificity was calculated as the number of true negatives detected by the test divided by the sum of true negatives and false positives. Positive predictive value was calculated as the number of true positives divided by the number of true positives and false positives. Negative predictive value was calculated as the number of true negatives divided by the number of true negatives and false negatives. The positive likelihood ratio of a patient responding to treatment is sensitivity divided by 1 minus specificity. Receiver operator curves (ROC) were used to choose appropriate thresholds for each classifier, requiring a sensitivity of 100%. The ROC diagnostic calculates the sensitivity and specificity for each parameter. Real-Time RT-PCR Validation

TaqMan^® real-time RT-PCR was employed to verify the microarray results of the AHR and AKAP 13 genes. For each 1 μg sample of amplified RNA, cDNA was produced using T7 oligo(dT) primer and Superscript II reverse transcriptase according to the manufacturer's instructions (Invitrogen). Primers and MGB-probes for AKAP 13 gene and control gene PBGD were designed using Primer Express (Applied Biosystems), while those for AHR gene and control gene HPRT were available as Assays-on-Demand from ABI. Primer/probe sequences for AKAP 13 were as follows: AKAPl 3 forward, 5'ggtcagatgtttgccaaggaa3' (SEQ ID NO: 1); AKAP 13 reverse, 5'tcttcagaaacacactcccatc-3' (SEQ ID NO: 2); AKAP 13 probe, 6FAM-tgaaacggaagaagcttgtA-3' (SEQ ID NO: 3).

All primers and probes were tested for optimal amplification efficiency above 90%. The relative standard curve was composed of 5 dilutions (10-fold each) of HeLa cDNA (in most cases, ranging from 25 ng to 2.5 pg). RT-PCR amplification mixtures (25 μL) contained 100 ng template cDNA, 2x TaqMan^® universal PCR master mix (12.5 μL; Applied Biosystems), 500 nM forward and reverse primers, and 250 nM probe. Reactions were run on an ABI PRISM 7900HT Sequence Detector (Applied Biosystems). The cycling conditions were: 2 min of AmpErase UNG activation at 5O⁰C, 10 min of polymerase activation at 95⁰C and 50 cycles at 95⁰C for 15 sec and annealing temperature (59⁰C or 6O⁰C) for 60 sec. In each assay, a standard curve and a no-template control along with template cDNA were included in triplicates for the gene of interest and control gene. The relative quantity of each gene was calculated based on the standard curve, and was normalized with the quantity of the control gene. The median coefficient of variation (based on calculated quantities) of triplicate samples was 8%. The correlation between repeated runs using independently diluted templates from the stock was above 0.95. Samples were only compared with microarray data if duplicate TaqMan^® experiments showed reproducible results. Cell Line Culture and AKAP 13 Over-expression Assay

The AKAPl 3 vectors, oncoLBC and protoLBC, and vector control (pSRalpha- neo) were obtained from Dr. Deniz Toksoz. Zheng et al. (1995). The HL60 cell line was obtained from the American Tissue Culture Collection and grown in RPMI 1640 with 10% FBS. Cells were transiently transfected with each vector using the Effectene reagent (Qiagen) according to the manufacturer's instructions and kept under G418 (600 μg/mL) for 7 days. Tipifarnib was then added in various concentrations (0, 1.5, 3.1, 6.3, 13, 25, 50, 100, 200, 1000, and 10,000 nM) to duplicate cultures (1.5x10⁵ cells/mL). Cells were counted at Day 6. Cell counts were normalized to cultures with no drug to give a percent of viable control cells. Results

Expression Profiling of Relapsed and Refractory AML

FTIs were originally designed to specifically inhibit FTase activity, thereby blocking the oncogenic ras pathways. Therefore, we initially analyzed DNA from the bone marrow of 80 patients with relapsed or refractory AML for activating ras mutations and investigated the possible correlation between ras mutation and the response to tipifarnib. While 26% of the AML samples harbored N-ras mutations, mutation status did not correlate with objective response or overall survival. Harousseau et al. (2003). We therefore performed gene expression profiling to identify novel signatures that could be used to predict response to the FTI tipifarnib. Bone marrow samples were obtained for gene expression analysis from 80 patients prior to treatment with tipifarnib. Table 1 shows the patient information. Table 1. Patient Information

A30007 REFRACTORY FEMALE 54 NR 106 no

A30008 REFRACTORY MALE 52 NR 27 no

A30053 RELAPSED MALE 51 NR 48 no

A30057 REFRACTORY MALE 74 NR 102 no

A30060 REFRACTORY FEMALE 33 NR 175 no

A30096 REFRACTORY MALE 69 NR 182 no

A30179 REFRACTORY FEMALE 70 NR 148 no

A30182 REFRACTORY MALE 70 NR 92 no

A30190 RELAPSED FEMALE 54 NR 51 no

A30191 RELAPSED FEMALE 67 NR 78 no

A30245 RELAPSED MALE 63 NR 366 no

A30300 RELAPSED MALE 47 NR 414 no

A30302 RELAPSED MALE 62 NR 234 no

A3O3O8 RELAPSED MALE 66 NR 71 no

A3031 1 RELAPSED FEMALE 61 NR 1 15 no

A30377 RELAPSED MALE 68 NR 364 no

A30047 RELAPSED FEMALE 63 NR 94 no

A30055 RELAPSED FEMALE 71 NR 56 no

A30063 RELAPSED MALE 46 NR 220 no

A30090 REFRACTORY FEMALE 85 NR 56 no

A30091 REFRACTORY FEMALE 67 NR 56 no

A30092 REFRACTORY FEMALE 54 NR 40 no

A301 11 REFRACTORY FEMALE 71 NR 38 no

A301 12 RELAPSED FEMALE 61 NR 12 no

A301 13 REFRACTORY MALE 75 NR 177 no

A301 19 REFRACTORY MALE 19 NR 36 no

A30153 RELAPSED FEMALE 68 NR 105 no

A30176 REFRACTORY MALE 75 NR 54 no

A30178 RELAPSED FEMALE 70 NR 39 no

A30180 REFRACTORY MALE 62 NR 72 no 3:

A30183 REFRACTORY MALE 63 NR 64 no

A30244 RELAPSED FEMALE 34 NR 35 no

A30247 REFRACTORY FEMALE 72 NR 35 no

A30248 RELAPSED MALE 46 NR 61 no

A30304 RELAPSED MALE 65 NR 44 no

A30306 RELAPSED FEMALE 28 NR 74 no

A30349 REFRACTORY MALE 58 NR 22 no

A30354 REFRACTORY FEMALE 31 NR 103 no

A30359 RELAPSED MALE 65 NR 8 no

A30363 RELAPSED MALE 64 NR 37 no

A30376 RELAPSED FEMALE 24 NR 383 no

A30378 RELAPSED FEMALE 76 NR 184 no

A30381 REFRACTORY FEMALE 70 NR 128 no

A30395 REFRACTORY MALE 61 NR 83 no

* Stable disease (SD) only included if confirmed by independent investigators HI = hematological improvement CR = complete response PR = partial response NR = no response

Fifty-eight of the 80 samples passed quality control measures including RNA quality and chip performance. There were no significant differences in age, sex, AML class (relapsed or refractory), cytogenic risk factors, baseline blast counts, response, and overall survival between these 58 patients and the remainder of the clinical study population (N = 194; Table 2).

Table 2 co-variate 58 subset 194 remainder p-value response 14 28 0.1237 male 28 119 0.1055 average age 60 56 0.1046 relapsed 31 104 0.8977 cytogenetic risk 34 5 0.1503 average blasts 55% 50% 0.1629

The gene expression data were integrated with the clinical information and retrospective analyses were performed to identify genes that could separate responders from non-responders with a high level of sensitivity. The data went through several filtering steps before identification of differentially expressed genes. First, genes that were not expressed in at least 10% of the samples were removed. This reduced the number of genes from approximately 22,000 to 1 1,723 genes. For unsupervised analyses genes that showed little variation in expression across the dataset (coefficient of variance of <45% across all the samples) were also excluded and quantile normalization was applied to the remaining 5,728 genes. At this stage an unsupervised k-means clustering analysis was performed to identify any differences between patients based on their global gene expression profiles. Six main clusters of patients were identified using this technique. No separation between responders and non-responders was observed (Figure 1 ). Only a handful of genes may be associated with the anti-tumor effect of FTIs, for example, it is possible that the differential expression of a single gene that is involved in FTI biology impacts clinical response and this would be masked by the noise introduced from the other 1 1,722 genes.

Example 2

Identification of Genes that Are Differentially Expressed Between Responders and Non-responders

We next performed supervised analysis using the gene expression data to identify genes that were differentially expressed between all responders and at least 40% of non-responders. These criteria were chosen to identify genes that could predict response to tipifarnib with the highest level of sensitivity possible. From 1 1 ,723 genes, a total of 19 genes were identified that could stratify responders and non-responders (Table 3 and Table 10 for more detail) and that gave significant P values in a t-test (P <0.05). The genes included those involved in signal transduction, apoptosis, cell proliferation, oncogenesis, and potentially, FTI biology (ARHH, AKAP 13, IL3RA).

Table 3 Top 19 Genes that Predict Response to tipifarnib and Results of Analysis

SEQ ID NO Symbol Specificity P value Functional description

151 AHR 0.52 0.00000255 Apoptosis, cell cycle, signal transduction

309 AKAP 13 0.63 0.00006133 Small GTPase mediated signal transduction, oncogenesis

488 MINA53 0.50 0.00006934 Cell proliferation

411 IDS 0.50 0.00023964 Glycosaminoglycan degradation

632 OPN3 0.40 0.00064297 G-protein coupled receptor protein signaling

280 GPRl 05 0.43 0.00087608 G-protein coupled receptor protein signaling

582 TENCl 0.43 0.0010309 Signal transduction

376 TNFSF 13 0.40 0.00104219 Cell proliferation

134 SVIL 0.45 0.00145723 Cytoskeletal anchoring

272 IL3RA 0.40 0.00198392 Receptor signaling

209 C6orf56 0.40 0.00261553

697 FRAGl 0.45 0.00298989 Tumor suppressor

476 GOSRl 0.45 0.01201057 Intra-Golgi transport

204 KIAAl 036 0.43 0.01262079 483 BTG3 0.47 0.01659402 Regulation of cell cycle

487 MAPK8IP3 0.40 0.01817428 Regulation of JNK cascade

419 L1LRB3 0.41 0.02374898 Immune response

242 ARHH 0.40 0.02721922 Small GTPase mediated signal transduction

496 NPTX2 0.45 0.03346833 Heterophilic cell adhesion

Real Time RT-PCR Validation of Gene Markers

To verify the microarray gene expression data, TaqMan real time RT-PCR was performed on cDNA that was used for generating the labeled target cRNA for microarray hybridization. Two genes were selected to verify the gene expression data. The AHR and AKAP 13 genes were chosen because the use of these genes resulted in the highest level of specificity for responders. The correlation coefficient was 0.74 for AHR and 0.94 for AKAP 13 indicating that the microarray gene expression data could be validated by PCR (Figure 2). Identification of the AKAP 13 Gene as a Marker of Resistance AKAP 13 was over-expressed in patients who were resistant to tipifarnib. The predictive value of this gene was calculated for the 58 samples using a leave-one-out cross validation (LOOCV; Figure 3A). AKAP 13 gene expression predicted non- response with a negative predictive value (NPV) of 96%, and low expression levels mediated response with a positive predictive value (PPV) of 43% (χ2 = 13.7; P = 0.0022). The overall diagnostic accuracy was 69% and positive likelihood ratio of responding was 2.4. Therefore, stratification of this patient population based on AKAP 13 gene expression increased the response rate from 24% (14/58) in the entire group to 43% (13/30) among those patients with low expression of the gene. Expression values for the AKAPl 3 gene in each patient are shown in Figure 3B. When survival was analyzed by Kaplan-Meier analysis, the median survival of patients with low expression of this gene was 90 days longer than those patients who had high expression levels (P = 0.008; Figure 3C). Identification of a Minimal Set of 3 Gene Markers

LOOCV was used to identify a candidate set of gene markers that could predict response to tipifarnib with an improved accuracy compared to AKAPl 3 alone.

Classifiers were built with an increasing number of genes based on t-test P values, and the error rate of these classifiers was calculated using LOOCV while keeping the sensitivity of predicting response at 100% (Figure 4A). The 3-gene classifier could predict response with the lowest error rate (Figure 4A). This was also observed when a leave-five-out cross validation was performed. When more genes were added the error rate increased, indicating that additional genes were introducing noise to the classifier. For the 3-gene classifier, the LOOCV demonstrated a NPV of 94% and a PPV of 48%, with an overall diagnostic accuracy of 74% and positive likelihood ratio of 2.9 (Figure 4B). The combined expression values for the 3 genes in each patient are shown in Figure 4C. Therefore, for the group of patients with this gene signature, the response rate to tipifarnib was 48% (12/25) compared to 24% (14/58) in this patient population. Using the 3-gene signature (AHR, AKAP 13 and M1NA53), Kaplan-Meier analysis again showed a significant difference in survival between the predicted responder group and the non-responder group (Figure 4D). The 13 patients who were incorrectly classified as responders had a better overall survival compared to the 31 patients correctly classified as non-responders (Figure 5). Interestingly, the 2 patients that were misclassified as non-responders only demonstrated hematological improvements with relatively short overall survival times (71, 87 days). Over-expression of AKAP 13 Increases Resistance to tipifarnib in AML

The AKAP 13 gene was the most robust marker of resistance to tipifarnib. We therefore investigated its involvement in FTI biology by over-expressing the oncoLBC and protoLBC variants of this gene in the HL60 cell line. Transient transfectants were then tested for sensitivity to tipifarnib. Over-expression of both AKAP 13 variants in this AML cell line model led to an approximate 20-fold increase in resistance to tipifarnib compared to control cells (Figure 3). Both the LBC oncogene and proto-oncogene increased the resistance to tipifarnib to the same extent, as seen by a parallel rightward shift of the kill curves by more than one log- unit compared to control. Discussion

Two groups recently identified gene expression profiles of newly diagnosed adult AML patients that are useful for predicting clinical outcome. Bullinger et al. (2004); and VaIk et al. (2004). These profiles seem to be more powerful than currently used prognostic markers such as karyotyping. Moreover, expression profiles have been found that predict response to anticancer compounds including standard chemotherapeutics (Chang et al. (2003); Okutsu et al. (2002); and Cheok et al. (2003)) and novel selective anticancer agents. Hofmann et al. (2002); and McLean et al. (2004). Similarly, pharmacogenetic profiles have recently been found that correlate with patient response to the tyrosine kinase inhibitor gefitinib. Paez et al. (2004) and Lynch et al. (2004). In that study, a subgroup of non-small cell lung cancer patients had activating mutations within the target epidermal growth factor receptor that correlated with clinical response to the targeted therapy.

In a phase 2 study of relapsed and refractory AML patients, we have identified gene expression profiles that predict response to tipifarnib, a novel farnesyl transferase inhibitor. This class of compounds is showing promise in the treatment of hematological malignancies (Karp et al. (2001); Kurzrock et al. (2004); Alsina et al. (2003); Cortes et al. (2003); and Thomas et al. (2001)) and solid tumors such as breast cancer (Johnston et al. (2003)) and recurrent glioma. Brunner et al. (2003). However, while clinical responses are being demonstrated, there is a growing need to tailor therapy by identifying patients who are most likely to respond to the drug and are, therefore, the best candidates for treatment. Furthermore, while ras was considered to be a primary target of this class of drugs, several clinical studies have shown that they are not necessarily effective in populations with a high frequency of ras mutations. Van Cutsem et al. (2004); and Rao et al. (2004).

Several gene markers were identified that have the potential to predict response to tipifarnib. A subset of these markers was both predictive of drug response and also thought to have the potential to be involved in FTI biology. One of the top candidates discovered from the microarray studies was the lymphoid blast crisis oncogene (oncoLBC or AKAP 13). This gene functions as a guanine nucleotide exchange factor for the Rho proteins (Zheng et al. (1995)) and as a protein kinase A anchoring protein. Carr et al. (1991). AKAP 13 contains a region that is homologous to an α-helical domain that is known to interact with lamin B. Foisner et al. (1991). This association could lead to lamin B activation via protein kinase A, consequently increasing mitotic activity. Both RhoB and lamin B are farnesylated and are candidate targets of FTIs. AKAP 13 is also a proto-oncogene, because loss of its 3-prime end causes cellular transformation. Sterpetti et al. (1999). While it was originally identified from a patient with chronic myeloid leukemia, its expression has not been documented in AML.

The identification of several genes that are potentially involved in FTI biology (ARHH, AKAP13, IL3RA) support the idea that the interaction of multiple pathways can affect how FTIs function in this population of AML patients (Figure 7). These genes interact with several farnesylated proteins including ras, rho, and potentially lamin B. Rho proteins are potentially important antitumorigenic targets for FTIs. Sahai et al. (2002); and Lancet et al. (2003). RhoB, RhoA, and RhoC have been found to be over-expressed in multiple cancer types. Sahai et al. (2002). In addition, RhoH (ARHH) is frequently re-arranged in tumors of myeloid origin, and this may lead to its over-expression. Pasqualucci et al. (2001). While most of these Rho proteins are geranygeranylated, they interact closely with each other and the farnesylated ras, RhoE, and RhoB small GTPases. Sahai et al. (2002); and Li et al. (2002). Furthermore, it has been shown that RhoH, RhoB, and RhoE can act in an antagonistic fashion to the transforming abilities of RhoA and RhoG. Li et al. (2002). The activity of RhoA, and possibly other related small GTPases, is increased by the guanine nucleotide exchange factor lymphoid blast crisis oncogene (AKAP13). Sterpetti et al. (1999); and Toksoz et al. (1994). In addition, AKAP13 may increase mitotic activity by activating lamin B via protein kinase A. Foisner et al. (1991). It is also well known that the IL3 receptor activates ras pathways. Testa et al. (2004). Therefore, as indicated in Figure 7, the increased activity of IL3RA and AKAP 13, and the decrease in RhoH expression could lead to an increased cellular profile of transformation. This might allow for the leukemic blast cell to overcome the anti-tumorigenic effects of FTIs through compensatory pathways. In contrast, when IL3RA and AKAP 13 are under-expressed and there is an increase in RhoH activity, FTIs may be more effective in blocking these pathways.

Finally, we demonstrated that over-expression of AKAP 13 (both oncoLBC and protoLBC variants) increased the IC₅₀ of the HL60 AML cell line by approximately 20-fold. This indicates that over-expression of AKAP 13 is a relevant marker of resistance and that it may also be a useful alternative drug target for patients who are resistant to tipifarnib.

Overall, our findings allow development of a gene expression based diagnostic assay to identify patients likely respond to tipifarnib. This information could be used to better direct treatment to the appropriate patient population. Using survival as the gold standard, the gene signature predicts a level of response to therapy that cannot be predicted by using conventional clinical response criteria. Alternatively, this raises the question of whether the gene signature for predicting response to FTI treatment also has prognostic value irrespective of FTI therapy. We thus evaluated a prognostic signature previously identified in newly diagnosed AML patients who were treated with conventional chemotherapy. Bullinger et al. (2004). While this signature showed utility in the current patient population our 3-gene signature further stratified these poor and good prognostic groups showing that it is a predictor of response to FTIs.

Example 3

Analysis of an AML prognostic gene signature

The 3-gene signature can predict prognosis irrespective of the type of drug treatment. To determine this, we first evaluated a gene-expression signature recently identified in newly diagnosed AML patients who were treated with conventional chemotherapy. Bullinger et al. (2004). This signature was defined using a cDNA array and therefore we first matched these genes with the probes present on the Affymetrix gene chip. Of the 133 predictive genes identified by Bullinger et al., 167 probe sets (corresponding to 103 unique genes) were matched to the Affymetrix U133A chip. The 3 genes identified in our present analysis are not present in the

Bullinger et al. 133 gene list. SEQ ID NOs: shown in Table 4. Two main groups of patients were defined by hierarchical clustering using these 167 probe sets (Fig 8A). Kaplan-Meier analysis showed a clear stratification of these clusters into patients with good and poor prognosis (Fig 8B, p = 0.000003). Our data therefore show that a subset of the 133-gene prognostic signature identified by Bullinger et al. (2004) can also be used in a relapsed and refractory cohort of patients. Consequently, this indicates that the prognostic gene profile is surprisingly robust across different microarray platforms, and different classes of AML.

Neither of the clusters defined by the prognostic gene signature had significantly more responders. However, when the tipifarnib 3-gene signature was applied to the good and poor prognostic groups, patients who responded to tipifarnib were further stratified from both prognostic groups (Fig 8C). Therefore, the 3-gene signature has independent utility to the prognostic signature and that it is specific for FTI treatment in this population of patients. Example 4

Clinical Evaluation and Response Definitions

The current study was an open label, multicenter, non-comparative Phase 2 study investigating the efficacy and safety of farnesyl transferase inhibition with tipifarnib administered as a single agent, at a starting oral dose of 600 mg b.i.d. for the first 21 days of each 28 day cycle in AML. Subjects were enrolled into two cohorts, those with relapsed AML and those with refractory AML. A total of 252 patients (135 relapsed and 117 refractory) were treated. For the purposes of the gene expression profiling response to tipifarnib was defined as patients who had an objective response (CR, CRp, or PR) (as described above), or patients who demonstrated confirmed stable disease, or a hematological response (decrease of >50% of leukemic blast cells) as determined by either central review or by the clinical site at any time during follow up. Sample Collection and Microarray Processing All samples were obtained from patients who had consented to the described processing and analyses. Bone marrow samples were collected from patients before treatment with tipifarnib, diluted with PBS and centrifuged with Ficoll-diatrizoate (1.077g/ml). White blood cells were washed twice with PBS, resuspended in FBS with 10% DMSO and immediately stored at -8O⁰C. Cells were thawed and total RNA was extracted from cell samples using the RNeasy Kit (Qiagen, Valencia, CA). RNA quality was checked using the Agilent Bioanalyzer. Synthesis of cDNA and cRNA were performed according to Affymetrix (Santa Clara, CA) protocols. Two rounds of linear amplification were performed because the RNA yield for several samples was too low to obtain enough labeled cRNA for chip hybridization using one round of amplification.

For hybridization, 11 μg of cRNA were fragmented randomly by incubation at 94°C for 35 min in 40 mM Tris-acetate, pH 8.1, 100 mM potassium acetate, and 30 mM magnesium acetate. Fragmented cRNA was hybridized to Ul 33A arrays at 45°C for 16 h in a rotisserie oven set at 60 rpm. Following hybridization, arrays were washed (with 6x SSPE and 0.5x SSPE containing Triton X-100 (0.005%)), and stained with streptavidin-phycoerythrin (SAPE; Molecular Probes, Eugene, OR). Quantification of bound labeled probe was conducted using the Agilent G2500A GeneArray scanner (Agilent Technologies, Palo Alto, CA). The total fluorescence intensity for each array was scaled to the uniform value of 600. Chip performance was quantitated by calculating a signal to noise ratio (raw average signal/noise). Chips were removed from further analysis if their signal-to- noise ratio was less than 5. Genes were only included in further analysis if they were called "present" in at least 10% of the chips. Approximately 12,000

Affymetrix probe sets remained following this cut-off. Gene expression data quality was further controlled by identifying outliers based on principal components analysis and by analyzing the normal distributions of the gene intensities (Partek Pro V5.1). Statistical Analysis

Unsupervised hierarchical clustering and clustering was performed in Omniviz. Kaplan-Meier analysis was performed using S-Plus.

Example 5

A prognostic signature identified in de novo AML has utility in relapsed and refractory patients

Two papers were recently published describing gene-expression profiling of newly diagnosed adult AML patients and its use in predicting clinical outcome.

Bullinger et al. (2004); and VaIk et al. (2004). We have profiled 58 patients with relapsed and refractory AML using the Affymetrix U133A gene chip. Of the 133 predictive genes identified by Bullinger et al. 167 probe sets (corresponding to 103 unique genes) were identified on the U133A chip (Fig 9). Bullinger et al. (2004).

The 167 probe sets are listed in the Sequence Listing Table and designated the SEQ

ID NOs: shown in Table 4.

Two main groups of patients were defined by hierarchical clustering using these 167 probe sets (Fig 1 OA). Kaplan-Meier analysis showed a clear stratification of these clusters into patients with good and poor prognosis (Fig. 1OB, p = 0.0000219).

Our data therefore shows that a 103 gene subset of the 133-gene prognostic signature identified by Bullinger et al. (2004) can also be used in a relapsed and refractory cohort of patients. Table 4. Consequently, this indicates that the prognostic gene profile is surprisingly robust across different microarray platforms, different classes of AML, and for different treatment algorithms.

Table 4

The prognostic signature is independent of a 3-gene signature that predicts response to tipifarnib.

We identified a 3-gene signature (AHR, AKAP13, MINA53) that predicts response to tipifarnib in relapsed and refractory AML patients. These genes can stratify patients into good and poor prognostic outcome groups (Fig 1 IB, p = 0.002). The question arises as to whether this gene signature is predicting response to FTI treatment or merely identifying patients who have a generally good prognosis. When the 3-gene signature was applied to the good and poor prognostic groups, responders were further stratified from the prognostic groups (Fig 11C, p = 0.000003). Following the application of both gene signatures there is clear stratification of a group of patients that do not respond to tipifarnib and have a poor prognosis irrespective of treatment type (Fig 1 ID, p = 0.0000005). Therefore, the 3- gene signature seems to be independent of the prognostic signature that has been identified and it is specific for FTI treatment in this population of patients. As a result we suggest that the prognostic signature maybe used in conjunction with drug- specific signatures (such as the tipifarnib predictive profile) to better manage patient therapy. Example 6

Identification of genes that are differentially expressed between responders and non-responders (not including stable disease patients^')

Four patients were removed from the analysis since they were classified as having stable disease and these patients cannot be clearly defined as either responders or non-responders. Inclusion of stable disease patients may bias the analysis for selecting genes associated with prognosis irrespective of drug treatment. This resulted in comparing 10 responders with 44 non-responders. Selected genes were required to show a specificity of 40% and a minimum mean fold-change of 2.0. These criteria were chosen to identify genes that could predict response to tipifarnib with the highest level of sensitivity possible. From 11,723 genes, a total of 8 genes were identified that could stratify responders and non-responders (Table 5) and that gave significant P values in a t-test (P <0.05). The genes included those involved in signal transduction, apoptosis, cell proliferation, oncogenesis, and potentially, FTI biology. AKAP 13 is the most robust marker

We next aimed at identifying a minimal set of genes that would provide the best diagnostic accuracy from the 8 selected genes. Classifiers were built with an increasing number of genes based on the AUC values from receiver operator characteristic analysis, and the error rate of these classifiers was calculated using LOOCV while keeping the sensitivity of predicting response at 100% (Fig. 12a). The AKAP 13 gene could predict response with the lowest error rate of less than 40% (Fig. 12a). The error rate increased to more than 50% when more than 2 genes were used in the classifier. For the AKAPl 3 the LOOCV demonstrated a NPV of 93% and a PPV of 31%, with an overall diagnostic accuracy of 63% and positive likelihood ratio of 2.0 (Fig. 12b). The expression value for AKAP13 in each patient is shown in Fig. 12c. Therefore, for the group of patients with low expression of AKAP 13, the response rate to tipifarnib was 31% (8/26) compared to 18% (10/54) in the current patient population. Using the AKAPl 3 gene, Kaplan-Meier analysis showed a significant difference in survival between the predicted responder group and the non-responder group (Fig. 12d).

Table 5. List of Top 8 Genes that Predict Response to Tipifarnib

AUC = area under the curve from receiver operator characteristic analysis. This is an indication of the overall diagnostic accuracy.

Example 7

Gene Expression Profiling Predictive of Tipifarnib (ZARNESTRA , ® , Rl 15777) Response in Patients With Newly Diagnosed Acute Myeloid Leukemia

Tipifarnib (ZARNESTRA®, Rl 15777), has demonstrated clinical response in patients with hematological disease. While the inhibition of protein farnesylation is the primary mechanism of action (MOA), the level of farnesyl inhibition is not a reliable pharmacodynamic marker of response, nor is it clear what genetic markers can be employed to predict response. This prospectively designed study was conducted to identify potential genetic markers and expression signatures that may be surrogate predictors of response for tipifarnib in patients with acute myeloid leukemia (AML). Bone marrow samples were collected and gene expression profiles analyzed from a single arm phase 2 clinical study of tipifarnib in poor-risk patients with newly diagnosed AML. Lancet et al. (2004). In total, 79 samples were profiled before (n = 25), during (n = 30), and after (n = 24) tipifarnib treatment.

Bone marrow samples were analyzed using the Affymetrix U133A GeneChip® array. Global gene expression signatures revealed that tipifarnib treatment resulted in gene expression changes that were maintained for up to 120 days following treatment termination. Pretreatment vs post-treatment samples identified approximately 500 genes that had significant changes (False Discovery Rate [FDR] <0.005) in gene expression following farnesyl transferase inhibition, including several genes associated with farnesylation (eg K-ras, FNTA). Many of the modulated genes were identified as those significantly involved in protein biosynthesis, DNA replication, intracellular signaling, and cell cycle pathways, thus, reflective of inhibition of cellular proliferation. A subset of 27 genes (including genes associated with signal transduction and cell cycle) was also identified as being differentially modulated between responders and non-responders (P < 0.01). Gene expression signatures previously identified from a phase 2 clinical trial in relapsed and refractory AML were also tested in pretreatment samples to examine their ability to predict response. Raponi et al. (2004). A combination of 6 genes was found to have significant predictive accuracy in this independent set of samples (P = 0.05). The genes identified from these studies might be used as surrogate biomarkers of tipifarnib activity. Patient Samples In a phase 2 study, patients with newly diagnosed AML received

ZARNESTRA®, 600 mg bid21dQ4wks. Bone marrow samples were obtained before, during, and after treatment with ZARNESTRA ®. Mononuclear cells isolated by Ficoll-Hypaque density centrifugation and viably frozen. Microarray Analysis Message RNA was amplified from patient blast cells and hybridized to the

Affymetrix U133A chip, which can probe for approximately 22,000 genes (Fig. 13). Chip data were pre-filtered to remove poor quality data and genes that were not expressed in at least 10% of the patient samples. In addition genes that did not vary across the dataset were removed (CV < 40%). Approximately, 8000 genes remained for further analysis. A total of 79 chips passed quality control measures and also had associated clinical response data.

Statistical Analysis

Analysis of variance (ANOVA) and t-tests were used to investigate the effect of drug treatment and time and their interactions for each gene. Multiple hypotheses testing was controlled by applying the false discovery rate (FDR) algorithm. All statistical analyses were performed in S-Plus 6.1 (Insightful Corporation). Principal components analysis was performed in Partek Pro. Hierarchical clustering was performed using a correlation metric and complete linkage (OmniViz Pro™, OmniViz, Maynard, MA). Pathway analysis was performed using Gene Ontology functional classifications. Table 6 shows the results of the analysis.

Table 6. Functional Gene Classes significantly modulated by tipifarnib

Approximately 8000 genes were used for global unsupervised clustering. Pretreatment samples clustered distally from during- and after-treatment samples. After-treatment samples included those that were from patients up to 120 days following treatment termination. Figure 14 shows that AML samples maintain FTI-mediated global gene expression changes following termination of tipifarnib treatment.

502 genes were found to be differentially expressed after treatment with tipifarnib (p<0.005). These genes are listed in Tables 7A-7C and in more detail in Table 9.

620 [218605 [670 |221505

Included in this list were many genes involved in FTI biology including AKTl, CENPF, KRAS, RAFl , STATs and farnesyltransferase. Several gene functional categories were found to be significantly enriched in this gene list (Table 8).

Table 8.

Genes differentially expressed between responders and non-responders after treatment with tipifarnib

We have previously identified 8 genes that were predictive of resistance to tipifarnib in relapsed and refractory AML. The most predictive gene in that dataset was AKAP-13 (AUC = 0.83). The predictive value of these genes was tested in the current set of newly diagnosed AML samples (CTEP20). Samples from patients with complete response or progressive disease were used. The results are shown in Figure 15.

A predictive classifier was built from the top 6 genes since these gave less than a 50% error rate in the training set. This 6 gene classifier is shown in Table 9 and its ability to stratify newly diagnosed AML is shown in Figure 16. Table 9

Conclusions • 27% of the clinical bone marrow samples could be used indicating optimization of sample collection is required.

• Global gene expression signatures of AML cells indicated FTI-treatment resulted in stable gene expression changes following treatment termination.

• Approximately 500 genes (p < 0.005) were affected by tipifarnib reflecting the multiple pathways that are targeted by FTIs. These included many pathways previously associated with FTI biology.

• Differential expression of 27 genes was found between responders and non- responders following treatment with tipifarnib (Table 8). These are candidate PD markers. • A 6-gene classifier identified in relapsed and refractory AML was found to have predictive value in newly diagnosed AML Example 8

Antibodies (Prophetic)

An LBC oncogene-derived peptide is synthesized, coupled to keyhole limpet hemocyanin, and used to immunize rabbits for production of polyclonal antibodies. The sera are tested for reactivity against the corresponding peptide with ELlSA, and the positive batches are affinity-purified. The purified antibody specifically detects the peptide that has the epitope in tissue sections. This is verified by complete abolishment of the signal if the corresponding peptide is added simultaneously with the antibody. In addition to this polyclonal antibody, which works well in IHC, monoclonal antibodies able to detect the protein in its natural fold are produced. To produce monoclonal antibodies, a purified antigen, produced in mammalian cells to ensure natural fold and posttranslational modifications, is generated. The antigen, LBC onco protein-IgG constant part fusion protein, is expressed in mouse myeloma cells, and the protein is purified using the Fc part as bait. This purified antigen is recognized in Western blot by the C-terminal polyclonal antibody. The antigen is used to generate mouse monoclonal antibodies against LBC peptides by selecting out of the positive clones those that produce antibodies that react against LBC peptide instead of the IgG constant part. Kits for the clinical identification of LBC oncogene can be readily fashioned employing these and similar antibodies. Such kits would include antibodies directed to LBC peptide identification (and hence,

LBC oncogene), appropriate indicator reagents (e.g., enzymes, labels, and the like), and (optionally) other reagents useful in the clinical application of such a kit such as dilution buffers, stabilizers, and other materials typically used in such assays.

Example 9 Immunohistochemistry (Prophetic)

An affinity-purified polyclonal antibody against the C-terminal peptide of LBC oncogene is used for the IHC detection and localization of LBC oncogene. Four μm sections from formalin-fixed and paraffin embedded normal and tumor tissue is mounted on 3-aminopropyl-triethoxy-silane (APES, Sigma, St. Louis, MO) coated slides. The sections are deparaffinized and rehydrated in graded concentrations of ethanol and treated with methanolic peroxide (0.5% hydrogen peroxide in absolute methanol) for 30 minutes at room temperature to block the endogenous peroxidase activity. Antigen retrieval is done in a microwave oven twice for 5 minutes (650W). An Elite ABC Kit (Vectastain, Vector Laboratories, Burlingame, CA) is used for immunoperoxidase staining. The LBC peptide antibody is used at an optimal dilution of 1 :2000. Both the biotinylated second antibody and the peroxidase- labeled avidin-biotin complex are incubated on the sections for 30 minutes. The dilutions are made in PBS (pH 7.2), and all incubations are carried out in a moist chamber at room temperature. Between the different staining steps the slides are rinsed three times with PBS. The peroxidase staining is visualized with a 3-amino- 9-ethylcarbazole (Sigma) solution (0.2 mg/ml in 0.05 M acetate buffer containing 0.03% hydrogen peroxide, pH 5.0) at room temperature for 15 minutes. Finally, the sections are lightly counterstained with Mayer's haematoxylin and mounted with aqueous mounting media (Aquamount, BDH). In control experiments the primary antibodies are replaced with the IgG fraction of normal rabbit serum or the primary antibody was preabsorbed with the LBC peptide. These stainings indicate the presence of the LBC oncogene in a subset of cells.

Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, the descriptions and examples should not be construed as limiting the scope of the invention.

Table 10

Sequence Listing Description

Reference list

5,242,974 5,561 ,071

5,384,261 5,571 ,639

5,405,783 5,593,839

5,412,087 5,599,695

5,424,186 5,624,71 1

5,429,807 5,658,734

5,436,327 5,700,637

5,445,934 6,004,755

5,472,672 6,133,305

5,527,681 6,218,1 14

5,529,756 6,218,122

5,532,128 6,271 ,002

5,545,531 6,271 ,210

5,554,501 6,284,764

5,556,752 6,306,897

Adnane et al. (2000) "Inhibition of famesyltransferase increases TGFbeta type II receptor expression and enhances the responsiveness of human cancer cells to TGFbeta" Oncogene 19:5525- 5533

Albertson (1984) EMBO J. 3: 1227-1234

Alsina et al. (2003) "Famesyltransferase inhibitor FTI-Rl 15777 is well tolerated, induces stabilization of disease, and inhibits farnesylation and oncogenic/tumor survival pathways in patients with advanced multiple myeloma" Eur. J. Haematol. 70:269

Brunner et al. (2003) "Famesyltransferase inhibitors: an overview of the results of preclinical and clinical investigations" Cancer Res. 63:5656-5668

Bullinger et al. (2004) "Use of gene-expression profiling to identify prognostic subclasses in adult acute myeloid leukemia" N. Engl. J. Med. 350: 1605-1616

Carr et al. (1991) "Interaction of the regulatory subunit (RII) of cAMP-dependent protein kinase with Rll-anchoring proteins occurs through an amphipathic helix binding motif J. Biol. Chem. 266: 14188-14192

Chang et al. (2003) "Gene expression profiling for the prediction of therapeutic response to docetaxel in patients with breast cancer" Lancet 362:362-369

Cheok et al. (2003) "Treatment-specific changes in gene expression discriminate in vivo drug response in human leukemia cells" Nat. Genet. 34:85-90

Cortes et al. (2003) "Efficacy of the famesyl transferase inhibitor Rl 15777 in chronic myeloid leukemia and other hematologic malignancies" Blood 101 :1692-1697

Cox et al. (2002) "Famesyltransferase inhibitors: promises and realities" Curr. Opin. Pharmacol. 2:388-393

Debemardi et al. (2003) "Genome-wide analysis of acute myeloid leukemia with normal karyotype reveals a unique pattern of homeobox gene expression distinct from those with translocation-mediated fusion events" Genes Chrom. Cancer 37:149-158

End et al. (2001) "Characterization of the antitumor effects of the selective famesyl protein transferase inhibitor Rl 15777 in vivo and in vitro" Cancer Res. 61 : 131-137

EP 430,402

Esteva et al. (2003) "Phase III study of weekly docetaxel and trastuzumab for patients with HER-2- overexpressing metastatic breast cancer" J Clin Oncol. 20: 1800-1808.

Foisner et al. (1991) "Protein kinase A- and protein kinase C-regulated interaction of plectin with lamin B and vimentin" Proc. Natl. Acad. Sci. USA 88:3812-3816

Gene Therapy Protocols, edited by Paul D. Robbins, Human press, Totowa NJ, 1996 Harousseau et al. (2003) "Zarnestra™ (Rl 15777) in patients with relapsed or refractory acute myelogenous leukemia (AML): results of a multicenter phase 2 study" Blood 102: 176a Abstract 614 Hofrnann et al. (2002) "Relation between resistance of Philadelphia-chromosome-positive acute lymphoblastic leukaemia to the tyrosine kinase inhibitor STI571 and gene-expression profiles: a gene- expression study" Lancet 359:481-486

Jensen et al. (2001) "Estrogen receptors and proliferation markers in primary and recurrent breast cancer" Proc Natl Acad Sci U S A. 98: 15197-15202.

Jiang et al. (2000) "The phosphoinositide 3-OH kinase/ AKT2 pathway as a critical target for farnesyltransferase inhibitor-induced apoptosis" MoI. Cell. Biol. 20:139-148

Johnston et al. (2003) "Phase II study of the efficacy and tolerability of two dosing regimens of the farnesyl transferase inhibitor, Rl 15777, in advanced breast cancer" J. Clin. Oncol. 21 :2492-2499 Kallioniemi (1992) Proc. Natl. Acad. Sci. USA 89:5321-5325

Karp et al. (2001) "Clinical and biologic activity of the farnesyltransferase inhibitor Rl 15777 in adults with refractory and relapsed acute leukemias: a phase 1 clinical-laboratory correlative trial" Blood 97:3361-3369

Kohlmann et al. (2003) " Molecular characterization of acute leukemias by use of microarray technology" Genes Chrom. Cancer Genes Chrom. Cancer 37:396-405

Kurzrock et al. (2004) "Phase II study of Rl 15777, a farnesyl transferase inhibitor, in myelodysplastic syndrome" J. Clin. Oncol. 22: 1287-1292

Lancet et al. (2003) "Farnesyltransferase inhibitors in hematologic malignancies: new horizons in therapy" Blood 102:3880-3889

Lancet et al. (2004) "Tipifarnib (ZARNESTRA™ in Previously Untreated Poor-Risk AML of the Elderly: Updated Results of a Multicenter Phase 2 Trial" Blood 104:874a

Li et al. (2002) "The hematopoiesis-specifϊc GTP-binding protein RhoH is GTPase deficient and modulates activities of other Rho GTPases by an inhibitory function" MoI. Cell. Biol. 22: 1 158-1 171

Lynch et al. (2004) "Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to Gefitinib" N. Engl. J. Med. 350:2129-2139

McLean et al. (2004) "Pharmacogenomic analysis of cytogenetic response in chronic myeloid leukemia patients treated with imatinib" Clin. Cancer Res. 10:155-165

Methods in Molecular Biology, Vol. 33: In Situ Hybridization Protocols, Choo, ed., Humana Press, Totowa, NJ (1994)

Morgan et al. (2001) "Cell-cycle-dependent activation of mitogen-activated protein kinase kinase (MEK- 1/2) in myeloid leukemia cell lines and induction of growth inhibition and apoptosis by inhibitors of RAS signaling" Blood 97: 1823-1834

Na et al. (2004) "Inhibition of farnesyltransferase prevents collagen-induced arthritis by down- regulation of inflammatory gene expression through suppression of p21 (ras)-dependent NF-kappaB activation" J Immunol. 173:1276-1283.

Okutsu et al. (2002) "Prediction of chemosensitivity for patients with acute myeloid leukemia, according to expression levels of 28 genes selected by genome-wide complementary DNA microarray analysis" MoI. Cancer Ther. 1 : 1035-1042

Paez et al. (2004) "EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy activating mutations in the epidermal growth factor receptor underlying responsiveness of non- small-cell lung cancer to Gefitinib" Science 304:1497-1500

Pasqualucci et al. (2001) "Hypermutation of multiple proto-oncogenes in B-cell diffuse large-cell lymphomas" Nature 412:341-346

Pinkel (1988) " Fluorescence in situ Hybridization with Human Chromosome-Specific Libraries: Detection of Trisomy 21 and Translocations of Chromosome 4" Proc Natl Acad Sci USA 85:9138- 9142

Pinkel et al. (1998) "High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays" Nature Genetics 20:207-211 Rao et al. (2004) "Phase III Double-Blind Placebo-Controlled Study of Farnesyl Transferase Inhibitor Rl 15777 in Patients With Refractory Advanced Colorectal Cancer" J Clin Oncol 22 3950-3957

Raponi, et al (2004) "Identification of Molecular Predictors of Response to ZARNESTRA™ (Tipifarnib, Rl 15777) in Relapsed and Refractory Acute Myeloid Leukemia" Blood 104'861a

Reiss et al (1990) "Inhibition of purified p21ras farnesyl protein transferase by Cys'AAX tetrapeptides" Cell 62:81-88

Reuter et al. (2000) "Targeting the Ras signaling pathway a rational, mechanism-based treatment for hematologic malignancies?" Blood 96: 1655-1669

Sahai et al. (2002) "RHO-GTPases and cancer" Nat Rev. Cancer 2 133-142

Schoch et al. (2002) "Acute myeloid leukemias with reciprocal rearrangements can be distinguished by specific gene expression profiles" Proc. Natl. Acad. Sci USA 99.10008-10013

Sterpetti et al. (1999) "Activation of the Lbc Rho exchange factor proto-oncogene by truncation of an extended C terminus that regulates transformation and targeting" MoI Cell Biol 19 1334-1345

Strachan and Read, Human Molecular Genetics, 1996

Takada et al. (2004) "Protein farnesyltransferase inhibitor (SCH 66336) abolishes NF-kappaB activation induced by various carcinogens and inflammatory stimuli leading to suppression of NF- kappaB-regulated gene expression and up-regulation of apoptosis" J Biol Chem. 279:26287-26299.

Testa et al (2004) "Interleukin-3 receptor in acute leukemia" Leukemia 18 219-226

Thomas et al (2001) "Rl 15777, a farnesyl transferase inhibitor (FTI), has significant anti-leukemia activity in patients with chronic myeloid leukemia (CML)" Blood 98

Tijssen (1993) Laboratory Techniques in Biochemistry and Molecular Biology, VoI 24 Hybridization With Nucleic Acid Probes, Elsevier, NY

Toksoz et al (1994) "Novel human oncogene lbc detected by transfection with distinct homology regions to signal transduction products" Oncogene 9:621 -628

VaIk et al (2004) "Prognostically useful gene-expression profiles in acute myeloid leukemia" N. Engl. J Med. 350 1617-1628

Van Cutsem et al. (2004) "Phase III tπal of gemcitabine plus tipifarnib compared with gemcitabine plus placebo in advanced pancreatic cancer" J Clin Oncol 22 1430-1438

Yagi et al. (2003) "Identification of a gene expression signature associated with pediatnc AML prognosis" Blood 102 1849-1856

Zhang et al (2002) "Farnesyltransferase inhibitors reverse Ras-mediated inhibition of Fas gene expression" Cancer Res. 62 450-458

Zheng et al (1995) "Direct involvement of the small GTP -binding protein Rho in lbc oncogene function" J Biol. Chem. 270:9031-9034

Claims

1. A method of assessing acute myeloid leukemia (AML) status comprising the steps of a. obtaining a biological sample from an AML patient; and b. measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 wherein the expression levels of the Marker genes above or below pre-determined cut-off levels are indicative of AML status.

2. A method of staging acute myeloid leukemia (AML) patients comprising the steps of a. obtaining a biological sample from an AML patient; and b. measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 wherein the expression levels of the Marker genes above or below pre-determined cut-off levels are indicative of AML survival.

3. A method of determining acute myeloid leukemia (AML) patient treatment protocol comprising the steps of a. obtaining a biological sample from an AML patient; and b. measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 wherein the expression levels of the Marker genes above or below pre-determined cut-off levels are sufficiently indicative of response to therapy to enable a physician to determine the degree and type of therapy recommended to provide appropriate therapy.

4. A method of treating a acute myeloid leukemia (AML) patient comprising the steps of: a. obtaining a biological sample from an AML patient; and b. measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9 wherein the expression levels of the Marker genes above or below pre-determined cut-off levels are indicate a response to therapy and; c. treating the patient with adjuvant therapy if they have a responder profile.

5. A method of determining whether a acute myeloid leukemia (AML) patient is high or low risk of mortality comprising the steps of a. obtaining a biological sample from an AML patient; and b. measuring Biomarkers associated with Marker genes corresponding to those selected from Table 3 wherein the expression levels of the Marker genes above or below pre-determined cut-off levels are sufficiently indicative of risk of mortality to enable a physician to determine the degree and type of therapy recommended.

6. The method of one of claims 1 -5 further comprising measuring the expression level of at least one gene constitutively expressed in the sample.

7. The method of one of claims 1 -5 wherein the specificity is at least about 40%.

8. The method of one of claims 1-5 wherein the sensitivity is at least at least about 80%.

9. The method of claim 28 wherein the p-value is less than 0.05.

10. The method of one of claims 1-5 wherein gene expression is measured on a microarray or gene chip.

11. The method of claim 10 wherein the microarray is a cDNA array or an oligonucleotide array.

12. The method of claim 10 wherein the microarray or gene chip further comprises one or more internal control reagents.

13. The method of one of claims 1-5 wherein gene expression is determined by nucleic acid amplification conducted by polymerase chain reaction (PCR) of RNA extracted from the sample.

14. The method of claim 13 wherein said PCR is reverse transcription polymerase chain reaction (RT-PCR).

15. The method of claim 14, wherein the RT-PCR further comprises one or more internal control reagents.

16. The method of one of claims 1-5 wherein gene expression is detected by measuring or detecting a protein encoded by the gene.

17. The method of claim 16 wherein the protein is detected by an antibody specific to the protein.

18. The method of one of claims 1-5 wherein gene expression is detected by measuring a characteristic of the gene.

19. The method of claim 18 wherein the characteristic measured is selected from the group consisting of DNA amplification, methylation, mutation and allelic variation.

20. A method of generating an acute myeloid leukemia (AML) prognostic patient report comprising the steps of: determining the results of any one of claims 1-5; and preparing a report displaying the results.

21. The method of claim 20 wherein the report contains an assessment of patient outcome and/or probability of risk relative to the patient population and/or likelihood or response to chemotherapy.

22. A patient report generated by the method according to claim 21.

23. A kit for conducting an assay to determine acute myeloid leukemia (AML) prognosis in a biological sample comprising: materials for detecting isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9.

24. The kit of claim 24 further comprising reagents for conducting a microarray analysis.

25. The kit of claim 24 further comprising a medium through which said nucleic acid sequences, their complements, or portions thereof are assayed.

26. Articles for assessing acute myeloid leukemia (AML) status comprising: materials for detecting isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9.

27. The articles of claim 26 further comprising reagents for conducting a microarray analysis.

28. The articles of claim 27 further comprising a medium through which said nucleic acid sequences, their complements, or portions thereof are assayed.

29. A microarray or gene chip for performing the method of one of claims 1-5.

30. The microarray of claim 29 comprising isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9.

31. The microarray of claim 30 wherein the measurement or characterization is at least 1.5 -fold over- or under-expression.

32. The microarray of claim 30 wherein the measurement provides a statistically significant p- value over- or under-expression.

33. The microarray of claim 32 wherein the p-value is less than 0.05.

34. The microarray of claim 30 comprising a cDNA array or an oligonucleotide array.

35. The microarray of claim 30 further comprising or more internal control reagents.

36. A diagnostic/prognostic portfolio comprising isolated nucleic acid sequences, their complements, or portions thereof of a combination of genes selected from the group consisting of Marker genes corresponding to those selected from Table 3, Table 4, Table 5, Table 7, Table 8 or Table 9.

37. The portfolio of claim 36 wherein the measurement or characterization is at least 1.5 -fold over- or under-expression.

38. The portfolio of claim 37 wherein the measurement provides a statistically significant p-value over- or under-expression.

39. The portfolio of claim 37 wherein the p-value is less than 0.05.