[go: up one dir, main page]

WO2008125791A1 - Marqueurs de cancer - Google Patents

Marqueurs de cancer Download PDF

Info

Publication number
WO2008125791A1
WO2008125791A1 PCT/GB2007/001343 GB2007001343W WO2008125791A1 WO 2008125791 A1 WO2008125791 A1 WO 2008125791A1 GB 2007001343 W GB2007001343 W GB 2007001343W WO 2008125791 A1 WO2008125791 A1 WO 2008125791A1
Authority
WO
WIPO (PCT)
Prior art keywords
genes
kit
expression
prognostic
binding
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/GB2007/001343
Other languages
English (en)
Inventor
Pier Paolo Di Fiore
Francesco Nicassio
Fabrizio Bianchi
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
TOLLERVEY REBECCA
IFOM Fondazione Istituto FIRC di Oncologia Molecolare
Original Assignee
TOLLERVEY REBECCA
IFOM Fondazione Istituto FIRC di Oncologia Molecolare
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by TOLLERVEY REBECCA, IFOM Fondazione Istituto FIRC di Oncologia Molecolare filed Critical TOLLERVEY REBECCA
Priority to PCT/GB2007/001343 priority Critical patent/WO2008125791A1/fr
Priority to US12/576,812 priority patent/US8747867B2/en
Publication of WO2008125791A1 publication Critical patent/WO2008125791A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
    • C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
    • C12Q1/6876—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes
    • C12Q1/6883—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes for diseases caused by alterations of genetic material
    • C12Q1/6886—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes for diseases caused by alterations of genetic material for cancer
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/106—Pharmacogenomics, i.e. genetic variability in individual responses to drugs and drug metabolism
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/118—Prognosis of disease development
    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
    • C12Q2600/00—Oligonucleotides characterized by their use
    • C12Q2600/136—Screening for pharmacological compounds

Definitions

  • the present invention relates to novel markers for cancer, and the use of these markers in assessment of disease conditions and particularly in prognosis and in therapy.
  • cancer signatures which can be used in patient management or which can identify the targets subverted in neoplasia.
  • These efforts are mainly concentrated on unbiased screening of cancer transcriptomes .
  • one approach is to identify genes whose expression is significantly modified in tumours as compared to normal cells, or in tumours of different grades (e.g., Beer, D. G., et al . 2002, Nature Medicine Vol. 8, No. 8, 816-824) and to select from these a subset which are associated with survival.
  • a difficulty of this approach is that the resultant signatures often represent the end point of complex upstream interactions, and cannot readily be allocated to particular molecular pathways .
  • Signatures produced in the prior art are often not highly robust, and often fail to provide good results from datasets that have been obtained in different clinical environments and from different patients. Additionally, prior art signatures often include a large number of genes, which increases the cost and difficulty of clinical screening in patients. In a clinical setting the use of small signatures is desirable as analysis is rendered much easier. For instance, smaller numbers of genes are amenable to analysis with readily available technology such as Real Time PCR.
  • the present inventors have developed an approach which uses well defined molecular tools capable of forcing terminally differentiated cells in culture to re-enter the cell cycle.
  • the present invention is therefore based in part on a biased method of identifying cancer signatures, in which the examination of the cancer transcriptome is biased towards (or indeed, focused primarily or entirely on) genes which have been shown to be modulated in response to agents which force re-entry of terminally differentiated cells into the cell cycle.
  • This is based on the hypothesis that the molecular tools mimic pathways subverted in naturally occurring tumours and that a limited .number of altered signalling pathways lead to the malignant state.
  • the inventors have found that the signatures obtained by the method, and in particular by focusing on genes modulated in response to ElA, can provide good indicators for the assessment of cancer and cancer progression.
  • the genes identified in such a screen can be more readily reverse engineered into signalling pathways, and thus, pathways of particular interest in human cancers can be identified.
  • one aspect of the invention concerns methods of assessment of cancer which comprise assessing the status of such genes.
  • ElA-modulated genes of interest are genes whose expression is not strongly modulated, e.g., induced, by inactivation of a pocket protein, preferably Rb, and which are significantly modulated (e.g., induced) by an ElA pocket binding mutant, particularly the ElA pocket binding mutant YH47.
  • the mutant is described in Wang HG et al, "Identification of specific adenovirus ElA N-terminal residues critical to the binding of cellular proteins and to the control of cell growth", J Virol. 1993 Jan,- 67(1): 476-88.
  • genes in this class e.g., DDX21, SF3B1, ch-TOG, SKIN, TRPC4AP and SMU-I
  • genes in this class are upregulated in a significant proportion of human cancers (relative to normal tissue) , and can also be used as predictors of cancer progression.
  • this ElA induced pathway appears to represent a useful therapeutic target.
  • the inhibition of expression of an example of this class of genes, SKIN is able to dramatically reduce proliferation in cancer cell lines overexpressing SKIN while having no effect on normal cells.
  • the inventors have also identified other classes of genes which are induced by ElA and which can be used as predictors of cancer progression.
  • genes are genes whose expression is (a) strongly modulated, e.g., induced, by inactivation of a pocket protein, preferably Rb, not modulated (e.g., induced) by an ElA pocket binding mutant, particularly the ElA pocket binding mutant YH47 and strongly induced by E2F1 overexpression
  • the present invention provides a method for the assessment of cancer in a subject, the method comprising: providing an assay sample obtained from said subject; determining the expression level in the sample of one or more genes from figure 2.
  • the method involves determining the expression level of a plurality of genes from figure 2 (i.e., two or more genes), most preferably 3, 4, 5 or more genes.
  • the method may comprise determining the expression level of 10 or more or 15 or more genes.
  • the cancer is breast cancer.
  • the method may comprise determining the expression level in the sample of at least one (preferably at least two or three) of the ch-TOG, SKIN, and TRPC4AP genes.
  • the invention may also relate to method for the assessment of cancer in a subject, the method comprising: providing an assay sample obtained from said subject; determining the expression level in the sample of one or more genes (preferably 2, 3, 4, 5, 8, 10, 11, 12 or all genes) from table 1.
  • the cancer is colon cancer.
  • the method may comprise determining the expression level in the sample at least one (preferably at least two or three) of SKIN, SMU-I and ch-TOG.
  • the cancer is non-small cell lung carcinoma (NSCLC) .
  • NSCLC non-small cell lung carcinoma
  • the invention provides a method for assessment of NSCLC (non small cell lung cancer) in a subject, and preferably a method for assigning a prognosis to the subject, comprising: providing an assay sample obtained from said subject; determining the expression level in said sample of at least one gene from table 2 and/or table 3.
  • the above methods comprise determining the expression level of at least 2, 3, 4, 5, 8, 10, 11, 12 or more, or all, of the genes in table 2 and/or 3.
  • the method may comprise determining the expression level of at least 15 or 20 of the genes in that table.
  • the present inventors have also used an integrated strategy for analysing the cancer transciptome, which integrates meta- analysis of microarray expression data and expression profiling of ElA modulated genes. In doing so, the present inventors have identified further signatures as set out in tables 8, 8a, 8b and 9, associated with NSCLC.
  • the present invention also relates to a method for assessment of NSCLC in a subject, and preferably a method for assigning a prognosis to the subject, comprising: providing an assay sample obtained from said subject; determining the expression level in said sample of at least one gene from table 8, preferably a plurality of genes.
  • the method comprises determining the expression level of a plurality of genes from table 8, i.e., at least 2, 3, 4, 5, 6, 7, 8, 9 or 10 of said genes.
  • the method may comprise determining the expression level of at least 15, 20, 25 or 30 of said genes.
  • the method comprises determining the expression level of at least 2, 3, 4, 5, 6, 7, 8, 9 or 10 genes from table 8a.
  • Preferred genes of table 8a may be or include E2F1, MCM6, SF3B1, RRM2 and/or NUDCDl .
  • the method comprises determining the expression level of at least 2, 3, 4, 5, 6, 7, 8, 9 or 10 genes from table 8b.
  • Preferred genes of table 8b may be or include H0XB7, SerpinB5, E2F4, and/or HSPG2.
  • the method comprises determining the expression level of at least 2, 3, 4, 5, 6, 7, 8, 9 or 10 genes from table 2 , 3 or 9.
  • Other genes which may be used in some embodiments are those marked with the symbol t in table 8.
  • the genes for use in the method may include any one of these - e.g., the methods may comprise measuring the expression level of any 5, 8, 10 or more of the genes of table 8 including any one (or 2, 3, 4, 5 etc) of the genes marked with the symbol t.
  • the present invention relates to a kit/apparatus suitable for carrying out any of the above methods .
  • the present invention provides a kit for use in assessing cancer in a subject, and preferably a method for assigning a prognosis to the subject, based on a sample obtained therefrom, the kit comprising specific binding partners capable of binding to an expression product of: one or more genes (preferably each of a plurality of genes, more preferably 2, 3, 4, 5, 10, 15 or more genes) from figure 2 ; one or more genes (preferably each of a plurality of genes, more preferably 2, 3, 4, 5, 8, 10, 11, 12 or more genes) from table 2 ; one or more genes (preferably each of a plurality of genes, more preferably 2, 3, 4, 5, 8, 10, 11, 12 or more genes) from table 3; one or more genes (preferably each of a plurality of genes, more preferably 2, 3, 4, 5, 8, 10, 15, 20 or more genes) from table 8; and/or one or more genes (preferably each of a plurality of genes, more preferably at least 2, 3, 4, 5, 7, 8 or 9, or all
  • the binding partners may be nucleic acids which bind the gene itself, and which are detectably labelled.
  • the present invention also provides an apparatus comprising a solid support bearing binding partners capable of binding to an expression product of: one or more genes (preferably each of a plurality of genes, more preferably 2, 3, 4, 5, 10, 15 or more genes) from figure 2,- one or more genes (preferably each of a plurality of genes, more preferably 2, 3, 4, 5, 8, 10, 11, 12 or more genes) from table 2; one or more genes (preferably each of a plurality of genes, more preferably 2, 3, 4, 5, 8, 10, 11, 12 or more genes) from table 3; one or more genes (preferably each of a plurality of genes, more preferably 2, 3, 4, 5, 8, 10, 15, 20 or more genes) from table 8; and/or one or more genes (preferably each of a plurality of genes, more preferably at least 2, 3, 4, 5, 7, 8 or 9, or all 10 genes) from table 9.
  • one or more genes preferably each of a plurality of genes, more preferably 2, 3, 4, 5, 10, 15 or more genes
  • the present inventors have identified genes which provide a high prognostic power in NSCLC, and in particular which provide a remarkably high prognostic power with a relatively small number of genes .
  • the inventors have compared the newly-identified signature to the 21 gene signature of table 3 (also described in co-pending application PCT/EP2005/ 010153) .
  • the reduced gene set actually showed an increased prognostic power (sensitivity, accuracy and/or specificity) as compared to the 21-gene set.
  • the present invention relates to a method for the assessment of NSCLC in a subject, most preferably for assigning a prognosis to a subject having NSCLC, comprising: providing an assay sample obtained from said subject; and determining the expression level in said sample of a plurality of genes from table 9 (preferably 5 or more,- in some embodiments 6, 7, 8 or 9 or more genes, or all ten genes) .
  • the genes measured in the method consist of: no more than 14 (or 13, 12, 11 or 10) genes from table 3, including at least 6 (or 7, 8, 9 or 10) genes from table 9,- optionally other genes (e.g., prognostic genes) not listed in either table 3 or table 9 (in some embodiments, not listed in table 8); and
  • control genes are those not associated with assessment of cancer in the method.
  • the present invention also relates to a method for assigning a prognosis to a subject having NSCLC, the method comprising measuring in a sample taken from said subject the expression level of prognostic genes, wherein said prognostic genes include no more than 14 genes from table 3 and wherein said prognostic genes include at least 6 genes from table 9.
  • the prognostic genes include at least 7, 8 or 9, and more preferably, all ten of the genes from table 9.
  • the prognostic genes include no more than 13, 12, 11 or no more than 10 genes from table 3. Where the prognostic genes include no more than 13, 12, 11 or genes from table 3 , it may be preferred that at least 8 are from table 9; more preferably this includes 9 or all ten from table 9. Where said prognostic genes comprises no more than 10 genes from table 3 , these ten genes may be all ten genes from table 9.
  • the method may also comprise measuring genes as controls, wherein in the course of said method no association is made between the expression level of the controls and prognosis .
  • the assignment of a prognosis may take account of other factors and clinical parameters such as tumour size, tumour grading, tumour infiltration, and degree of metastasis of the tumour.
  • the prognostic genes may optionally include, in addition, other genes which are not listed in either table 3 or 9 , in order to further increase prognostic power (e.g., accuracy, sensitivity and/or specificity) .
  • the additional genes may be genes other than those listed in table 8.
  • the prognostic genes may also include p53, KRAS or other markers: e.g., selected from RB, EGFR, MYC, APC, CDH13, RARB, DAPKl, DAPK2 , FHIT, RASSFlA, BCL2 , ERBB2 (Her2/Neu) , GRP, KIT, p21, p27, pl6, FAS, CASP3 (Caspase 3), BIRC5 (Survivin) , VEGF, PDGF, FGF2 , COL18A1 (Collagen XVIII) , CCNBl, CCNDl, TERT, SEMA3B, PTEN, hOGGl , BAPl, TIMP3 , MGMT, FUSl, ROBOl, TSLCl, NPRL2, CYB561D2, GSTPl and/or MGMT.
  • markers e.g., selected from RB, EGFR, MYC,
  • the' prognostic genes measured may consist of: no more than 14 (or 13, 12, 11 or 10) genes from table 3 including at least 6 (or 7, 8, 9 or 10) genes from table 9; plus optionally other genes not listed in either table 3 or table 9 (in some embodiments, not listed in table 8) .
  • the genes measured in the method may consist of: no more than 14 (or 13, 12, 11 or 10) genes from table 3, including at least 6 (or 7, 8, 9 or 10) genes from table 9; plus optionally other genes not listed in either table 3 or table 9 (in some embodiments, not listed in table 8); plus optionally control genes (control genes being those not associated with assessment of cancer in the method) .
  • a useful feature of the present invention is that it provides a small, and hence more clinically useful, signature. Accordingly, it may be preferred that the total number of prognostic genes whose expression levels are measured is 50 or less, more preferably 30 or less, 25 or less, 20 or less, or 19, 18, 17, 16, 15 or less. In some embodiments, the total number of prognostic genes measured may be only 14, 13, 12, 11 or 10 (or less) . Of course, in any one of these embodiments, the prognostic genes may include 7, 8, 9 or all ten of the genes from table 9.
  • the genes from table 9 may in some embodiments include one or more, e.g., 2, 3 or all of HOXB7, SerpinB ⁇ , E2F4, and/or HSPG2. Additionally or alternatively, the genes from table 9 may in some embodiments include two or more, e.g., 3, 4 or all of E2F1, MCM6, SF3B1, RRM2 and/or NUDCDl. In some embodiments, it may be preferred that the genes from table 9 include HOXB7 , SerpinB5, E2F4, HSPG2, E2F1, MCM6 and RRM2 , and optionally also SF3B1.
  • the present invention provides a kit/apparatus suitable for use in any one of the above methods .
  • the invention provides a kit for use in assigning a prognosis to a subject based on a sample obtained therefrom, the kit comprising specific binding partners capable of binding to an expression product of each of a set of prognostic genes, wherein said prognostic genes include no more than 14 (optionally no more than 13, 12, 11 or 10) genes from table 3 and wherein said prognostic genes include at least 6 genes (optionally at least 7, 8, 9 or all ten genes) from table 9.
  • the binding partners may be nucleic acids which bind the gene itself, and which are detectably labelled.
  • the set of prognostic genes may comprise no more than 50, 30, 25, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11 or 10 genes.
  • the kit could also comprise binding members for controls (genes not associated with prognosis) .
  • the kit may consist of specific binding partners for each of: no more than 14 (or 13, 12, 11 or 10) genes from table 3, including at least 6 (or 7, 8, 9 or 10) genes from table 9; optionally other genes not listed in either table 3 or table 9 (in some embodiments, not listed in table 8) ; and optionally control genes (controls genes being those not associated with assessment of cancer in the method)
  • the total number of genes for which binding partners are provided in the kit may be fewer than 50, 30, 25, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11 or 10 genes (including 6, 7, 8, 9, or 10 genes from table 9, optionally 1, 2, 3 or 4 other genes from table 3, plus optionally other prognostic markers and/or controls) .
  • the invention also provides an apparatus comprising a solid support bearing binding partners capable of binding to an expression product of each of a set of prognostic genes, wherein said prognostic genes include no more than 14 genes from table 3 (optionally no more than 13, 12, 11 or 10 genes) and wherein said prognostic genes include at least 6 genes
  • the apparatus may consist of said solid support bearing binding members for 6, 7, 8, 9 or 10 genes from table 9, plus optionally 0, 1, 2, 3, or 4 other genes from table 3, plus optionally up to 30, 20, 10, 5, 4, 3, 2 or 1 other genes
  • kits/apparatus suitable for use in such a method relate equally to the kits/apparatus suitable for use in such a method.
  • RRM2 is associated with resistance of a variety of chemotherapeutic agents, including gemcitabine (Zhou, B. S et al. 1995, Cancer Res 55:1328-1333, Goan, Y. G., et al . 1999, Cancer Res 59:4204-4207).
  • gemcitabine In pancreatic adenocarcinomas, direct targeting of RRM2 by siRNA enhanced chemosensitivity to gemcitabine both in vitro and in vivo (Duxbury, M. S et al 2004, Oncogene 23:1539-1548).
  • gemcitabine is one of the most active chemotherapeutic agents (Natale, R. 2005, Lung Cancer 50 Suppl l:S2-4).
  • the present invention also provides a method of determining whether or not to treat a subject identified as having NSCLC with gemcitabine or with an agent which inhibits RRM2 protein or which downregulates RRM2 expression (e.g., an antibody, or a nucleic acid inhibitor of expression such an antisense, ribozyme or siRNA molecule) , the method comprising determining the status/expression level of RRM2 in a sample obtained from said subject.
  • the subject has a stage I tumour, e.g., a stage IA or stage IB tumour.
  • RRM2 Overexpression of RRM2 may suggest that the subject should not be treated with gemcitabine, or should be treated with an agent which inhibits RRM2 protein or which downregulates RRM2 expression (e.g., in conjunction with treatment with gemcitabine) .
  • the method may be used in conjunction with any of the methods for assessment/prognosis of NSCLC described herein (in which case, RRM2 should obviously be one of the prognostic genes whose expression level is examined) .
  • the invention provides a method of determining whether a patient would benefit from adjuvant therapy (by assigning a prognosis to the patient) , and at the same time determining whether the adjuvant therapy should comprise the use of gemcitabine or an agent which inhibits RRM2 protein/downregulates RRM2 expression.
  • Figure 1 shows genes induced by ElA, and the results of RTQ- PCR in ElA (dl520) infected TD C2C12 myotubes, proliferating (MYB) C2C12 myoblasts, ElA (dl520) infected TD MSC (mouse satellite cells) and proliferating (MYB) MSC myoblasts.
  • the first column gives the mouse accession number.
  • the second column gives the name and description in mouse.
  • the fourth column gives the accession number of the human sequence.
  • Figure 2 shows ElA induced genes allocated to classes A, B, C or D according to their mechanism of regulation.
  • the columns show the ratio of induction under the named conditions with induction with wild type ElA.
  • the column headed "Ratio 24h/36h” shows the ratio of induction at 24h and 36h.
  • Figure 3. 3a shows the percentage of positive tumour samples for the named genes in different tissues.
  • Figure 3b shows bright field and dark field microscope analysis showing the specific signal from the cancer cells of tumour samples (T) compared to a matched normal counterpart (N) .
  • Figure 3c shows a cell cycle plot of relative mKNA levels of 4 ElA induced genes in GO synchronized serum starved NIH 3T3 cells stimulated by serum addition and HeLa cells released after nocodazole induced G2/M arrest. Almost all the class D genes are not cell cycle regulated in both serum response dependent and independent manner, while all the class A and B genes are cell cycle regulated and the class C genes marginally cell cycle regulated.
  • ClassA-XTPl filled squares
  • classB-MGC22679 empty squares
  • classD-TRPC4AP empty circles
  • classD-SKIN filled circles
  • TMA colon specific tissue microarrays
  • Figures 5 A-C show that selected class D genes predict disease outcome in breast cancer.
  • Figure 5D shows the probability of remaining free of distant metastasis for a patient having a good or bad prognosis based on the Class A, B and C genes predictor.
  • Class A, B and C were used together as a predictor of prognostic outcome on a subgroup of breast tumours with no lymph nodes involvement at surgery, which either developed metastatic disease (NO+ patients) or stayed disease-free (NO- T/GB2007/001343
  • tumour cell lines Six different tumour cell lines (as indicated) were treated with SKlN-specific siRNA (empty circles in A; RNAi in B and C), or a control scrambled oligo (filled triangles in A; scr. in B and C) or mock-treated (filled squares in A; mock in B and C) . Twenty-four hours after treatment, cells were re- plated to measure cell growth (A) , or analyzed for SKIN transcript levels by Q-RT-PCR (B) . A. Cells, re-plated in standard growth medium, were counted at the indicated time points. Data are expressed relative to the number of cells present in the plate 24 h after re-plating (assumed as 1) . B.
  • Q-RT-PCR data are expressed relative to those detected in growing MCFlOA cells, to allow for comparison among cell lines.
  • C In the case of DLDl and HT-29 cells, levels of SKIN were also measured by Western Blot with an anti-SKIN antibody.
  • Figure 7 shows the results of an ONCOMINE analysis of Class-D genes.
  • the genes which pass the statistical filter p-value ⁇ 0.05 with Bonferroni correction
  • Iog2 median value in every class considered where: "N” stands for normal samples; "T” for primary tumours and "M” for distant metastasis.
  • Figure 8 shows that SKIN is amplified in colon cancers.
  • Figure 9 shows the probability of remaining metastasis-free of patients with a good (dashed line) or poor (solid line) expression signature based on the inventor's NSCLC predictor
  • Figure 9A shows the results for the 12 gene predictor for the dataset (Michigan cohort) of Beer et al . (Beer, D. G., et al, 2002. Gene-expression profiles predict survival of patients with lung adenocarcinoma. Nat Med, 8: 816-824), and Figure 9B shows the results for the 12 gene predictor for the dataset (Harvard cohort) of Bhattacharjee, et al . (Bhattacharjee, A., et al, 2001. Classification of human lung carcinomas by mRNA expression profiling reveals distinct adenocarcinoma subclasses. Proc Natl Acad Sci U SA, 98:
  • Figure 9D shows the results of Q-RT-PCR analysis of the 21 genes predictor performed on an independent set of 30 patients, all with stage I NSCLC adenocarcinomas .
  • the 49-gene model predicts overall survival.
  • the 49-gene model of table 8b was used to predict overall survival in the Stage I subset of lung adenocarcinomas from the Michigan (67 of the 86 patients of the original datasets) , Harvard (62 of the 84 patients of the original datasets) and Duke (34 patients) cohorts (BiId, A. H. , et al, 2006. Oncogenic pathway signatures in human cancers as a guide to targeted therapies. Nature 439:353-357). Data are shown as the probability of survival, in a Kaplan-Meier plot, as a function of a "favorable" (upper line) , or "unfavorable” (lower line) signature .
  • the 28-gene biased signature and the 80-gene model predict overall survival.
  • the 28-gene biased signature (A) and the 80-gene model (B) were used to predict overall survival in Stage I lung adenocarcinomas of the Duke cohort (34 patients) .
  • Data are shown as the probability of survival, in a Kaplan-
  • the 10-gene model predicts overall survival.
  • the 10-gene model was tested to predict overall survival in the indicated cohorts of Stage I (A) and Stage IA (B) lung adenocarcinomas .
  • Data are shown as the probability of survival, in a Kaplan-Meier plot, as a function of a "favorable" (upper line) , or "unfavorable” (lower line) signature.
  • FIG. 13 Performance of the 49-gene model.
  • the 49-gene model was used to predict overall survival in lung adenocarcinoma patients of the Harvard and Michigan cohorts, as shown. Patients were divided according to tumor stage, and 7 001343
  • Kaplan-Meyer analyses are shown for Stage I patients, Stage II-III patients, and for the entire cohort (Stage I-II-III) .
  • the 49-gene model had no predictive power in Stage II-III adenocarcinomas.
  • the performance of the 49-gene model is compared to those of the 50- and 100-gene signature of Beer et al. (1) in the prediction of overall patient survival, in the Duke cohort, by Kaplan-Meyer analysis.
  • Upper line- favorable signature; lower line- unfavorable signature In respect of the Harvard Stage II-III group here and in figure 14, the favourable signature is black and the unfavourable signature is grey.
  • FIG 14. Performance of the 10-gene model on the Michigan and Harvard cohorts.
  • the 10-gene signature was used to predict overall patient survival within the Michigan and Harvard cohorts .
  • 7 genes could be used (SF3B1, NUDCDl and SCGB3A1 were not present on HU6800 microarray, used in that study)
  • 8 genes could be used (NUDCDl and SCGB3A1 were not present on HU95av2 microarray used in that study) .
  • Patients from the two cohorts were grouped according to the tumor stage, as described in figure 12. Remarkably, despite the reduction in the number of genes utilizable, the signature could still predict overall survival in patients with Stage I disease.
  • the prediction in the Duke cohort (all 10 genes could be used in this case) is the same as in Fig. 12A and is reported for comparison.
  • DDX21 Deadbox polypeptide 21.
  • SF3B1 is splicing factor 3b, subunit 1.
  • Ch-TOG is colonic and hepatic tumour overexpressed protein and is also known as KIAA097.
  • SKIN similar to KIAA0493 induced in tumour
  • TRPC4AP is the transient receptor potential cation channel, subfamily C, member 4 associated protein, and is also known as RRIP.
  • SMU-I is the Suppressor of MEC-8 and UNC-52 homolog.
  • Figure 1 provides the accession number for the human and mouse sequences, but reference to the gene or protein may include other mammalian sequences . The short names used herein are, for convenience, the names of the human homolog, but this is not intended to exclude other mammalian homologs.
  • Figure 2 provides the accession numbers for the four classes of ElA-induced genes .
  • ElA is intended to be reference to any adenoviral ElA expression product capable of inducing re-entry of a terminally differentiated cell into the cell cycle .
  • it refers to the ElA 12S mRNA product (which is the short splicing variant) , or to a fragment or variant thereof which retains the biological activity.
  • Tables 1, 2, 3, 7, 8, 8a, 8b, and 9 provide accession numbers for the genes therein.
  • reference to the gene or protein may include other mammalian sequences, i.e., mammalian homologs .
  • a patient or subject as referred to herein is preferably a mammalian patient or subject and most preferably human.
  • a sample or assay sample from said patient or subject is preferably a sample of tumour tissue.
  • providing a sample obtained from a patient and determining expression levels in the sample is reference to an in vitro method practiced on a sample after removal of the sample from the body, e.g. , by biopsy or during the course of surgery.
  • An assessment of cancer as referred to herein may be diagnosis or prognosis of the cancer.
  • the methods are methods of assigning a prognosis to a subject.
  • the assigned prognosis may be "good” or “poor” prognosis.
  • the assigned prognosis may be a probability of survival after a given period of time, e.g. , at five years (categorical or numerical probability) .
  • Cancers may be clinically assigned to various stages.
  • the most commonly used staging system for both breast cancer and NSCLC is the TNM system, which is widely described in the art. It may be preferred that the methods are methods of assigning a prognosis to a patient having an early stage cancer.
  • an early stage tumour is defined as a tumour that has not spread beyond the breast or the axillary lymph nodes (Tis, Tl, T2 , NO, MO) .
  • an early stage tumour is defined as a tumour that has not lymph node metastases or spread is confined to hilar lymph nodes. This includes stage I and stage II lung cancer .
  • Cancer may be e.g., melanoma, or cancer of the breast, colon, kidney, larynx, lung, prostate, stomach, uterus or brain.
  • the cancer is NSCLC.
  • the cancer is an adenocarcinoma, e.g. , an NSCLC adenocarcinoma .
  • NSCLC non-small cell lung carcinoma
  • stage I tumours are defined as TlNOMO (stage IA) or T2N0M0 (stage IB) .
  • T refers to the primary tumour as below:
  • Tl - Tumour with diameter of 3 cm or smaller and surrounded by lung or visceral pleura, without bronchoscopic evidence of invasion more proximal than lobar bronchus (i.e., not in the main bronchus)
  • T2 -Tumour with any of the following features of size or extent :
  • NO refers to no lymph node involvement and MO refers to no distant metastasis.
  • Methods of the present invention are preferably used to assign a prognosis to a patient having a stage I tumour, e.g., a stage IA or stage IB tumour.
  • a stage I tumour e.g., a stage IA or stage IB tumour.
  • NSCLC is an adenocarcinoma, e.g., a stage I (A or B) NSCLC adenocarcinoma.
  • "good" prognosis patients may be advised not to have adjuvant therapy, whereas "poor" prognosis patients may be advised to have adjuvant therapy.
  • the method may comprise assigning the patient to a treatment group, which may be a "with" or "without” adjuvant therapy group; in other words, the method may comprise determining whether a patient should have adjuvant therapy (by assigning a prognosis to the patient) .
  • Adjuvant therapy is therapy applied in addition to a primary therapy (e.g. , surgery) and is intended to reduce the risk of cancer recurrence.
  • adjuvant therapies include chemotherapy, radiotherapy, immunotherapy, targeted therapy or hormone therapy.
  • the methods of the invention may also comprise comparing the protein status or expression level to that of a control sample, as explained in more detail below.
  • a control sample is a sample of normal cells or a sample of tumour cells having good prognosis
  • a poor prognosis may be suggested by a gene status or by a level of a gene expression product which is divergent from the level in the control.
  • poor prognosis may be suggested by gene status or a level of gene expression products which is in line with or similar to the control sample.
  • more than one control may be used.
  • Poor prognosis may be associated with downregulation or upregulation of expression of a given gene, e.g., with downregulation or upregulation of a mKNA or protein expression product, relative to a reference expression level, such as a control or a good prognosis group.
  • patients may be assigned to the poor/good prognosis group depending on the number of genes they have following the poor/good prognosis pattern.
  • they may be put in the poor/good prognosis class if they have more than 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65% or 70% of genes following a poor/good prognosis pattern.
  • the chosen level will depend on the individual circumstances .
  • the method may also comprise measuring genes as controls (e.g., housekeeping or universally expressed genes) , wherein in the course of said method no association is made between the expression level of the controls and prognosis.
  • controls e.g., housekeeping or universally expressed genes
  • prognostic genes herein is intended to exclude such controls .
  • the assignment of a prognosis may take account of other factors and clinical parameters such as the tumour size, tumour grading, tumour infiltration, and degree of metastasis of the tumour.
  • the methods may include measuring other prognostic genes in order to further increase prognostic power (e.g., accuracy, sensitivity and/or specificity) .
  • the prognostic genes may also include p53, KRAS or other markers: e.g., selected from RB, EGFR, MYC, APC, CDHl3 , RARB, DAPKl, DAPK2 , FHIT, RASSFlA, BCL2, ERBB2 (Her2/Neu), GRP, KIT, p21, p27, pl6, FAS, CASP3 (Caspase 3), BIRC5 (Survivin) , VEGF, PDGF, FGF2 , COL18A1 (Collagen XVIII) , CCNBl, CCNDl, TERT, SEMA3B, PTEN, hOGGl, BAPl, TIMP3, MGMT, FUSl, ROBOl, TSLC
  • the total number of prognostic genes i.e., genes other than controls whose expression levels are measured is 100, 75, 50 or less, more preferably 30 or less, and in some embodiments 25 or less, 20 or less, or 19, 18, 17, 16, 15 or less.
  • the total number of prognostic genes measured maybe only 14, 13, 12, 11 or 10 (or less).
  • the total number genes (e.g., including both prognostic and control genes) measured is 100, 75, 50 or less, more preferably 30 or less, and in some embodiments 25 or less, 20 or less, or 19, 18, 17, 16, 15 or less .
  • Determination of protein, gene or transcript level may be made by any of the methods known in the art .
  • the method may comprise measuring the level of a protein or a nucleic acid, which may be DNA or mRNA.
  • a gene expression product as referred to herein may be a protein or a transcript (i.e., an RNA molecule expressed by the gene) .
  • suitable methods for assessing protein levels include immunohistochemistry (e.g., immunofluorescence), western blotting, and solid phase methods such as ELISA (enzyme-linked immunoabsorbant assay) .
  • an assessment of protein level can be made by determining the proportion of cells showing labelling (e.g., staining or fluorescence).
  • Transcript level may be determined by in situ hybridisation, e.g., accompanied by assessment of the proportion of cells showing hybridisation.
  • quantitative PCR methods may be used, e.g. based upon the ABI TaqManTM technology, which is widely used in the art. It is described in a number of prior art publications, for example reference may be made to WO00/05409. PCR methods require a primer pair which target opposite strands of the target gene at a suitable distance apart (typically 50 to 300 bases) . Suitable target sequences for the primers may be determined by reference to Genbank sequences .
  • a convenient method is by hybridisation of the sample (either directly or after generation of cDNA or cRNA) to a gene chip array and/or micro fluidic card (Low density array) .
  • Quantitative PCR methods may use microfluidic techniques, e.g., a microfluidic card.
  • the method may comprise contacting the sample with a substrate bearing a plurality of nucleic acids .
  • the substrate bears nucleic acids hybridising specifically to transcripts of each of the genes whose expression level is to be detected.
  • genes may be present in commercially available chips from Affymetrix, and these chips may be used in accordance with protocols from the manufacturer.
  • examples of methods for the provision of microarrays and their use may also be found in, for example, WO84/01031, WO88/1058, WO89/01157, WO93/8472, WO95/18376/ WO95/18377, WO95/24649 and EP-A-0373203 and reference may also be made to this and other literature in the art .
  • microfluidic card technology the genes may for instance be present in commercially available microfluidic cards from Applied Biosystem, also known as Low Density Arrays . These cards may be used in accordance with protocol from the manufacturer.
  • TaqMan® Low Density Arrays are customizable, easy-to-use, 384- well micro fluidic cards for real-time PCR-based quantitative gene expression applications (ABI TaqManTM technology) . Over than 40,000 inventoried TaqMan® assays covering human, mouse, and rat genes, are commercially available.
  • the micro fluidic technology uses 8 sample-loading ports, each connected to 48 reaction wells.
  • Gene copy number may be determined using techniques known in the art, including in situ hybridisation (ISH) with nucleic acid probes which may be labelled with e.g. a fluorescent label (FISH) , or PCR of genomic DNA.
  • ISH in situ hybridisation
  • FISH fluorescent label
  • a method of the invention comprises determining the gene status of an assay sample obtained from a patient and/or determining the expression level
  • the method may also comprise comparing the determination made on that sample with a determination made on a reference or control sample.
  • Reference or control samples for the above methods may be a sample of normal (unaffected) cells, preferably cells of the same type as the assay sample.
  • the sample may be a sample of cells affected by cancer, preferably a cancer of the same type as is in the patient or is suspected to be in the patient .
  • the control sample may preferably be taken from a tumour cell having one of the states of interest.
  • the control sample may be a sample taken from a tumour from a metastatic tumour, or may be a sample from a non-metastatic tumour.
  • control sample may be taken from one or more of hyperplastic polyps, adenomas and carcinomas.
  • control sample may be a sample of cells from a tissue type associated with the presence or absence of cancer, and/or from a tumour with good or with poor prognosis .
  • the control sample may be obtained from the patient, from another subject or from a population of subjects. Where a population of subjects is used, the comparison may be made with the average (e.g., mean or median) in samples of cells from said population.
  • One advantage of using a control of normal tissue from the same patient is that it accounts for any individual variation. Where the control is from another patient (either of normal or affected tissue) , this may also be a reason why results based on a population of patients may be preferred.
  • the method may comprise the use of more than one control; for example the sample to be tested may be compared to a normal sample from the same patient and the transcript level of an affected sample from another patient or patients. In another example, the sample to be tested may be compared to one or more sample from a metastatic tumour and one or more samples from a non-metastatic tumour.
  • the assay sample is a sample of affected tissue obtained from the patient
  • the control sample is obtained from the patient at an earlier time point, so as to provide a historical record. In one embodiment, this allows for monitoring of the progression of the condition over time.
  • this allows for assessment of the effectiveness of a particular treatment.
  • comparing the severity of the condition in a patient at two time points it is possible to determine whether a particular treatment regime is having a positive effect or not.
  • the effectiveness of any one regime may differ from patient to patient, or during the course of the disease .
  • Comparison to the gene status or to the level of a gene expression product in a control sample may of course be comparison to previously determined data, and need not comprise the step of analysing the control sample.
  • the specific binding partner for a protein may be an antibody, as defined below, and is preferably a monoclonal antibody.
  • the antibody may be detectably labelled.
  • the specific binding partner may be a nucleic acid sequence capable of specifically hybridising to said transcript.
  • the nucleic acid sequence may be detectably labelled. It may be a primer or primer pair, e.g., for quantitative PCR.
  • binding partner which is suitable for detection of the transcript or protein in a complex mixture.
  • the binding partner may bind to the gene expression product preferentially over other transcripts/proteins in the same species and may have no or substantially no binding affinity for other proteins or transcripts.
  • the transcript is preferably capable of distinguishing the target transcript from other transcripts in the mixture at least under stringent hybridisation conditions .
  • the invention relates to a kit or an apparatus which comprise a specific binding partner for a gene expression product.
  • the specific binding partner may be immobilised on a solid support.
  • the kit for assessment of a sample may comprise an apparatus comprising or consisting of a solid support bearing binding partners for each of the genes of interest .
  • the kit may further comprise a detectably labelled moiety capable of binding to a complex between the protein and its specific binding partner. Additionally or alternatively, the kit may include one or more of the following reagents :
  • a detection system to reveal the enzymatic activity coupled to the primary antibody or the secondary moiety (e.g., secondary antibody), where the label is an enzyme, such as peroxidase.
  • the kit may be for immunohistochemical techniques, and may comprise a first antibody capable of binding the protein to be detected, and a second, labelled antibody capable of binding said first antibody.
  • the kit may comprise a first, immobilised antibody capable of binding the protein to be detected and a second, labelled antibody capable of binding the protein when bound to the first antibody.
  • a label may be a radioactive, fluorescent chemiluminescent or enzyme label.
  • Radioactive labels can be detected using a scintillation counter or other radiation counting device, fluorescent labels using a laser and confocal microscope, and enzyme labels by the action of an enzyme label on a substrate, typically to produce a colour change.
  • the result of the assay is obtained by contacting the enzyme with a substrate on which it can act to produce an observable result such as a colour change, the extent of which depends on the amount of analyte originally in the sample .
  • Suitable enzyme labels may give rise to detectable changes such as colorimetric, fluorometric , chemiluminescent or electrochemical changes, and include horseradish peroxidase and alkaline phosphatase, as well as lysozyme (detectable for example by lysis of organisms such as microccocus lysodeikticus) , chymotrypsin, and E. coli DNA polymerase.
  • Other possible labels include macromolecular colloidal particles or particulate material such as latex beads that are coloured, magnetic or paramagnetic, and biologically or chemically active agents that can directly or indirectly cause detectable signals to be visually observed, electronically detected or otherwise recorded.
  • These molecules may be enzymes which catalyse reactions that develop or change colours or cause changes in electrical properties, for example. They may be molecularIy excitable, such that electronic transitions between energy states result in characteristic spectral absorptions or emissions. They may include chemical entities used in conjunction with biosensors. Other methods may also be used to detect interaction between the protein and the antibody, including physical methods such as surface plasmon resonance, agglutination, light scattering or other means .
  • the kit may comprise primers for PCR analysis of KNA samples or genomic DNA from patients, i.e., primers which are capable of hybridising to an RNA expression product of the gene in question, or to the gene itself, and of serving as extension primers.
  • the PCR may be quantitative PCR.
  • the kit may be a gene chip array, in 1 which case it preferably comprises a control specific for said at least one transcript; and optionally at least one control for the gene chip .
  • the kit may comprise probes for FISH analysis of gene copy number or other genetic alterations .
  • the kit may comprise nucleic acids capable of binding to each of a set of prognostic genes , wherein said nucleic acids are detectably labelled (preferably fluorescently labelled) .
  • the number of sequences in the array will be such that where the number of nucleic acids suitable for detection of the marker transcript is n, the number of control nucleic acids specific for individual transcripts is n' , where n' is from 0 to 2n, and the number of control nucleic acids (e.g. for detection of "housekeeping" transcripts, transcripts having normally high levels in the cell type being assessed, or the like) on said gene chip is m where m is from 0 to 100, preferably from 1 to 30, then n + n' + m represent at least 50%, preferably 75% and more preferably at least 90% of the nucleic acids on said chip.
  • the array may comprise binding partners for a total of e.g. , less than 1000, 500, 400, 300, 200, 100, 50, 40, 30, 25, 20 or 15 prognostic genes .
  • kit or apparatus comprises binding partners for no more than 500, 400, 300, 200, 100, 50, 30, 25, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11 or 10 prognostic genes.
  • kit or apparatus could also comprise binding members for controls (genes not associated with prognosis) .
  • the total number of genes for which binding partners are provided in the kit or apparatus, including any controls may be fewer than 500, 400, 300, 200, 100, 50, 30, 25, 20, 19, 18, 17, 16 or 15.
  • Antibodies Methods of producing antibodies are known in the art.
  • Preferred antibodies are isolated, in the sense of being free from contaminants such as antibodies able to bind other polypeptides and/or free of serum components. Monoclonal antibodies are preferred for some purposes, though polyclonal antibodies are within the scope of the present invention.
  • kits comprise more than one antibody
  • these are preferably mixtures of isolated antibodies as described above.
  • Antibodies may be obtained using techniques which are standard in the art. Methods of producing antibodies include immunising a mammal (e.g. mouse, rat, rabbit) with a polypeptide of the invention. Antibodies may be -obtained from immunised animals using any of a variety of techniques known in the art, and screened, preferably using binding of antibody to antigen of interest. For instance, Western blotting techniques or immunoprecipitation may be used (Armitage et al, Nature, 357:80-82, 1992).
  • an antibody specific for a protein may be obtained from a recombinantly produced library of expressed immunoglobulin variable domains, e.g. using lambda bacteriophage or filamentous bacteriophage which display functional immunoglobulin binding domains on their surfaces; for instance see WO92/01047.
  • a suitable library is a Llama library.
  • Antibodies according to the present invention may be modified in a number of ways. Indeed the term “antibody” should be construed as covering any binding substance having a binding domain with the required specificity, e.g., antibody fragments, derivatives, functional equivalents and homologues of antibodies, including synthetic molecules and molecules whose shape mimics that of an antibody enabling it to bind an antigen or epitope .
  • antibody should be construed as covering any binding substance having a binding domain with the required specificity, e.g., antibody fragments, derivatives, functional equivalents and homologues of antibodies, including synthetic molecules and molecules whose shape mimics that of an antibody enabling it to bind an antigen or epitope .
  • Example antibody fragments capable of binding an antigen or other binding partner, are the Fab fragment consisting of the
  • VL, VH, Cl and CHl domains the Fd fragment consisting of the VH and CHl domains; the Fv fragment consisting of the VL and VH domains of a single arm of an antibody; the dAb fragment ' which consists of a VH domain,- isolated CDR regions and F(ab')2 fragments, a bivalent fragment including two Fab fragments linked by a disulphide bridge at the hinge region.
  • Single chain Fv fragments are also included.
  • Humanized antibodies in which CDRs from a non-human source are grafted onto human framework regions , typically with the alteration of some of the framework amino acid residues, to provide antibodies which are less immunogenic than the parent non-human antibodies, are also included within the present invention.
  • the present invention also provides for the use of a protein selected from DDX21, SF3B1, ch-TOG, SKIN, TRPC4AP and SMU-I or other gene listed in Fig 2, a protein of table 2, a protein of table 3 or a protein of table 8 or 9 for screening for a candidate agent for the treatment of cancer in a patient.
  • the invention provides a method of screening for a candidate agent for the treatment of cancer in a patient, comprising: a) providing a protein selected from DDX21, SF3B1, ch- TOG, SKIN, TRPC4AP and SMU-I or other gene listed in Fig 2, or a protein of table 2 , table 3 or table 8 or 9 ; b) bringing the protein into contact with a test agent; c) determining whether said test agent is capable of binding and/or modulating the activity of the protein.
  • the invention provides a method of screening for a candidate agent for the treatment of cancer in a patient, comprising: identifying a gene whose expression is modulated in a terminally differentiated mammalian cell in culture by contacting the cell with ElA so as to cause its re-entry into the cell cycle,- providing a protein expressed by the gene,- bringing the protein into contact with a test agent; and determining whether said test agent is capable of binding and/or modulating the activity of the protein.
  • the invention also provides a method of screening for a candidate agent for the treatment of cancer in a patient, wherein said method comprises a) providing a transformed cell in culture; b) bringing said cell into contact with a test agent; and c) determining whether said test agent is capable of modulating the level of a transcript selected from DDX21, SF3B1, ch-TOG, SKIN, TRPC4AP and SMU-I or other transcript listed in Fig 2, or a transcript of table 2, table 3 or table 8 or 9.
  • the invention provides the use of an agent obtainable in one of the above screening methods for the manufacture of a medicament for the treatment of cancer.
  • the cancer may be e.g., melanoma, or cancer of the breast, colon, kidney, larynx, lung, prostate, stomach, uterus or brain.
  • the cancer is NSCLC.
  • the cancer is an adenocarcinoma, e.g., an NSCLC adenocarcinoma.
  • the protein used in the above assays may be a mammalian protein, preferably a human protein. It may also be a fragment or variant of the full length mammalian protein. Preferred fragments and variants are those which retain the activity of the mammalian protein. Fragments may comprise at least 10, more preferably at least 20, 30, 40 or 50 consecutive amino acids of the mammalian protein sequence. A variant may have at least 70%, 80%, 90%, 95% or 99% identity to a full length mammalian sequence, preferably to the human sequence, assessed over the full length of the mammalian sequence .
  • the percentage identity of amino acid sequences can be calculated using commercially available algorithms.
  • the following programs (provided by the National Center for Biotechnology Information) may be used to determine homologies: BLAST, gapped BLAST, BLASTN and PSI-BLAST, which may be used with default parameters .
  • the protein for use in the assay may be fused to a heterologous sequence, e.g. , a sequence allowing the protein to be isolated and/or immobilised.
  • Binding assays may be competitive or non-competitive.
  • the assay method may comprise determining whether the test agent is capable of inhibiting the protein, or determining whether the test agent is capable of activating the protein.
  • the assay is preferably for an activator of the protein, and the assay preferably involves determining whether the test agent is capable of increasing the activity of the protein.
  • the assay may be carried out under conditions where the protein normally shows low or no activity.
  • the assay is preferably for an inhibitor of the activity of the protein, and the assay preferably involves determining whether the test agent is capable of reducing the activity of the protein.
  • the assay may be carried out under conditions in which the protein is normally active.
  • proteins with enzymatic function may be assayed in the presence of a substrate for the enzyme, such that the presence of a test agent capable of modulating the activity results in a faster or slower turnover of substrate.
  • the substrate may be the natural substrate for the enzyme or a synthetic analogue. In either case, the substrate may be labelled with a detectable label to monitor its conversion into a final product .
  • test agent may be examined for ligand binding function in a manner that leads to antagonism or agonism of the ligand binding property.
  • DNA binding or transcriptional activating activity may be determined, wherein a modulator is able to either enhance or reduce such activity.
  • DNA binding may be determined in a mobility shift assay.
  • the DNA region to which the protein bind may be operably linked to a reporter gene (and additionally, if needed, a promoter region and/or transcription initiation region between said DNA region and reporter gene) , such that transcription of the gene is determined and the modulation of this transcription, when it occurs, can be seen.
  • Suitable reporter genes include, for example, chloramphenicol acetyl transferase or more preferably, fluorescent reporter genes such as green fluorescent protein.
  • Test agents may be natural or synthetic chemical compounds used in drug screening programmes .
  • Extracts of plants , microbes or other organisms, which contain several characterised or uncharacterised components may also be used.
  • Combinatorial library technology (including solid phase synthesis and parallel synthesis methodologies) provides an efficient way of testing a potentially vast number of different substances for ability to modulate an interaction.
  • Such libraries and their use are known in the art, for all manner of natural products, small molecules and peptides, among others . Many such libraries are commercially available and sold for drug screening programmes of the type now envisaged by the present invention.
  • test agents or candidate modulators are antibodies or binding fragment thereof which bind a protein target, as described above.
  • test agents are peptides based upon a fragment of the protein sequence to be modulated.
  • fragments of the protein corresponding to portions of the protein which interact with other proteins or with DNA may be a target for small peptides which act as competitive inhibitors of protein function.
  • Such peptides may be for example from 5 to 20 amino acids in length.
  • the peptides may also provide the basis for design of mimetics, as explained in more detail below.
  • the invention provides methods comprising the step of providing a transformed cell in culture, and determining whether a test agent is capable of modulating (inhibiting or activating) the levels of a gene transcript.
  • the transformed cell may be a tumour cell, e.g., isolated from a human subject, or may be a cell which has been contacting with a transforming agent or an agent which causes re-entry of a terminally differentiated cell into the cell cycle.
  • the cell may be a cell which has been contacted with an ElA protein as described above, e.g., by infecting the cell with an adenovirus.
  • the cell may be a terminally differentiated cell.
  • Cell based assay methods can be configured to determine expression of the gene either at the level of transcription or at the level of translation. Where transcripts are to be measured, then this may be determined using the methods described above, e.g. on gene chips, by multiplex PCR, or the like.
  • the assay is preferably for agents which increase the expression of the gene (e.g., by increasing the quantity of the transcript) .
  • agents which increase the expression of the gene e.g., by increasing the quantity of the transcript
  • Such an agent may comprise the coding sequence of the gene itself (i.e., it may be a gene therapy vector) .
  • the assay is preferably for agents which decrease the expression of the gene.
  • Cell based assay methods may be used to test agents of the sorts described above. They may also be used to screen further " classes of test agents/candidate modulators, including antisense oligonucleotides .
  • antisense oligonucleotides are typically from 12 to 25, e.g. about 15 to 20 nucleotides in length, and may include or consist of modified backbone structures, e.g. methylphosphonate and phosphorothioate backbones, to help stabilise the oligonucleotide.
  • the antisense oligonucleotides may be derived from the coding region of a target gene or be from the 5 ' or 3 ' untranslated region.
  • Test agents may further include RNAi, i.e.
  • RNA molecules which are sequence specific for a gene transcript. They may also include ribozymes which specifically target the transcript mRNA, i.e., a catalytic RNA molecule which cleaves other RNA molecules of a particular nucleic acid sequence. General methods for the construction of ribozymes are known in the art .
  • Agents obtained, in accordance with the present invention may be used in methods of treating cancer in a patient.
  • the modulator will be formulated with one or more pharmaceutically acceptable carriers suitable for a chosen route of administration to a subject.
  • pharmaceutically acceptable carriers include, for example, pharmaceutical grades of mannitol, lactose, cellulose, cellulose derivatives, starch, magnesium stearate, sodium saccharin, talcum, glucose, sucrose, magnesium carbonate, and the like may be used.
  • Liquid pharmaceutically administrable compositions can, for example, be prepared by dissolving, dispersing, etc, a modulator and optional pharmaceutical adjuvants in a carrier, such as, for example, water, saline aqueous dextrose, glycerol, ethanol, and the like, to thereby form a solution or suspension.
  • a carrier such as, for example, water, saline aqueous dextrose, glycerol, ethanol, and the like
  • the pharmaceutical composition to be administered may also contain minor amounts of non-toxic auxiliary substances such as wetting or emulsifying agents, pH buffering agents and the like, for example, sodium acetate, sorbitan monolaurate, triethanolamine sodium acetate, sorbitan monolaurate, triethanolamine oleate, etc.
  • composition or formulation to be administered will, in any event, contain a quantity of the active compound(s) in an amount effective to alleviate the symptoms of the subject being treated. Routes of administration may depend upon the precise condition being treated.
  • Mimetics Once candidate substance have been found in the assays and screens according to the present invention, they may be used to design mimetic compounds for development as drugs.
  • the designing of mimetics to a known pharmaceutically active compound is a known approach to the development of pharmaceuticals based on a "lead" compound. This might be desirable where the active compound is difficult or expensive to synthesise or where it is unsuitable for a particular method of administration, e.g. peptides are unsuitable active agents for oral compositions as they tend to be quickly degraded by proteases in the alimentary canal.
  • Mimetic design, synthesis and testing is generally used to avoid randomly screening large number of molecules for a target property.
  • the pharmacophore Once the pharmacophore has been found, its structure is modelled to according its physical properties, e.g. stereochemistry, bonding, size and/or charge, using data from a range of sources, e.g. spectroscopic techniques, X-ray diffraction data and NMR. Computational analysis, similarity mapping (which models the charge and/or volume of a pharmacophore, rather than the bonding between atoms) and other techniques can be used in this modelling process.
  • a range of sources e.g. spectroscopic techniques, X-ray diffraction data and NMR.
  • Computational analysis, similarity mapping which models the charge and/or volume of a pharmacophore, rather than the bonding between atoms
  • other techniques can be used in this modelling process.
  • the three-dimensional structure of the ligand and its binding partner are modelled. This can be especially useful where the ligand and/or binding partner change conformation on binding, allowing the model to take account of this in the design of the mimetic.
  • a template molecule is then selected onto which chemical groups which mimic the pharmacophore can be grafted.
  • the template molecule and the chemical groups grafted on to it can conveniently be selected so that the mimetic is easy to synthesise, is likely to be pharmacologically acceptable, and does not degrade in vivo, while retaining the biological activity of the lead compound.
  • the mimetic or mimetics found by this approach can then be screened to see whether they have the target property, or to what extent they exhibit it . Further optimisation or modification can then be carried out to arrive at one or more final mimetics for in vivo or clinical testing.
  • Example 1 The following examples are provided by way of illustration. Example 1
  • TD C2C12 myotubes were infected with either the adenovirus dl520 (expressing only the 12S ⁇ iRNA of ElA) or the control adenovirus dl312 (expressing no ElA mRNA) . Only the dl520 infected myotubes displayed 48h p.i. S-phase re-entry phenotype (about 70%) . 2 ⁇ g of time course pooled polyA + RNA from dl520 and dl312 infected myotubes was used as starting material for cDNA retro-transcription (Invitrogen) and subtraction procedures (Clonetech) to obtain a library of about 800 clones.
  • ElA induced library was screened by Reverse Northern. 14 filters (7 plates, 2 filter per plate) contained all the cloned sequences as single purified PCR bands and some controls (DNA ladder IX as negative control, adenoviral cDNA and NP95 sequence as positive control, GAPDH as internal standard) . Each plate (2 filters) was hybridized in duplicate with two different labelled cDNA pools (dl520 and dl312 infected myotubes cDNA) to fish out by comparing the radioactive signals only the ElA (dl520) induced clones. The single positives clones were picked, then grown and sequenced to retrieve by blast analysis the corresponding gene.
  • the Reverse Northern positive ElA induced genes are validated by SYBR GREEN based quantitative RT PCR on RNA from ElA (dl520) infected TD C2C12 myotubes, proliferating (MYB) C2C12 myoblasts, ElA (dl520) infected TD MSC (mouse satellite cells) and proliferating (MYB) MSC myoblasts.
  • Figure 1 shows the results of the validation. It shows 55 non- redundant clones of which 29 (henceforth referred to as ElA- induced genes) showed reproducible, and greater than 2-fold, induction, upon ElA expression in both TD C2C12 mouse myotubes and primary TD muscle satellite cells (MSC) .
  • ElA- induced genes 29 (henceforth referred to as ElA- induced genes) showed reproducible, and greater than 2-fold, induction, upon ElA expression in both TD C2C12 mouse myotubes and primary TD muscle satellite cells (MSC) .
  • a mathematical ratio calculated between 24h/EARLY and 36h/LATE ElA induction defined the timing of activation.
  • EARLY >0,4 ;
  • LATE ⁇ 0,4.
  • Example 2 ElA exerts pleiotropic effects on TD myotubes. It suppresses tissue-specific genes, through its binding to the transcriptional co-activators p300/CBP and MyoD, and reactivates the cell cycle, through a mechanism in which binding to pocket proteins (mainly pRb and pl30) and restoration of E2F activity is pivotal.
  • pocket proteins mainly pRb and pl30
  • the ectopic expression of E2Fs in TD myotubes does not induce S phase, indicating that other ElA-activated pathways are concomitantly needed.
  • ElA-regulated pocket/Rb-independent mechanisms are known, which involve CycE/CDK2-, CtBP-, TRAPP- or p400- regulated pathways, and other chromatin remodelling activities .
  • Real-time PCR was carried out on the ABI/Prism 7700 Sequence Detector System (Perkin-Elmer/Applied Biosystems), using a pre-PCR step of 10 min at 95 0 C, followed by 40 cycles of 15 s at 95°C and 60 s at 60 0 C. Specificity of the amplified products was confirmed by melting curve analysis (DISSOCIATION CURVETM Perkin-Elmer/Applied Biosysterns) and by 6% PAGE. Preparations with KNA template without reverse transcriptase were used as negative controls.
  • Class B Pocket-dependent, E2Fl-independent (or scarcely-dependent) genes (8 genes) .
  • Class D Pocket-independent (or substantially-independent) genes. This group of 6 genes is well activated by YH47 and scarcely by Rb removal. In addition almost all of them are
  • a first genetic cluster comprising class A and B genes, is constituted by "typical" ElA-responsive genes, whose induction is stringently pocket protein-dependent (regardless of the E2Fl-dependen.ce) . All the early-induced genes belong to this group. Of interest, a subset of genes in this genetic cluster (MCM7, MCM4 and MCM6) , which is widely known to be under the transcriptional control of E2F1 in non-post-mitotic cells, does not seem to be responsive to the overexpression of this protein in a TD environment, despite retaining pocket protein- dependence.
  • a second genetic cluster (class C and D) is made up of pocket- indifferent or pocket-independent genes. It is not clear why all these genes are "late” genes, albeit the correlation is too strong to be due to chance. More importantly, within this cluster, class D genes constitute a transcriptional signature, induced by a well-defined genetic alteration, through a yet unknown mechanism.
  • TMA tissue microarrays
  • tumour proliferative index assessed by immunostaining with anti-Ki-67
  • levels of four of six class D genes SKIN, TRPC4AP, SMU-I and ch-TOG/KIAA0097
  • SF3B1 was also overexpressed in adenomas (Fig. 4), albeit with overall less intense staining than in adenocarcinomas (not shown) , consistent with the possibility that its overexpression represents an early event in tumour progression.
  • DDX21 is also overexpressed in adenomas.
  • the Class D genes were able to predict the risk of disease relapse with a p-value ⁇ 0.05 on the data set generated in- house and a p-value ⁇ 0.04 on the data set from van't Veer.
  • the predictive strength of the 3-genes model was further confirmed by Q-RT-PCR (p-value 0.003) on 15 randomly selected breast tumor patients (all lymph node negative at diagnosis) , which were all homogeneous for estrogen receptor status (Er pos) (Fig 5 C) .
  • Patients having enhanced expression of ch-TOG and SKIN and a reduced expression of TRPC4AP were designated as having a poor prognosis, whereas patients with a reduced expression of ch- TOG and SKIN and an enhanced expression of TRPC4AP were designated as having a "good" prognosis.
  • class A/B/C prognosis predictors
  • their predictive ability was tested on a subgroup of tumours with no lymph nodes involvement at surgery, which either developed metastatic disease (NO+ patients) or stayed disease-free (NO- patients) over a 5 year follow up period analysed by Affymetrix.
  • NO+ patients metastatic disease
  • NO- patients stayed disease-free
  • Affymetrix Affymetrix
  • Class C genes were able to predict the risk of disease relapse (p-value ⁇ 0.004, Figure 5D).
  • tumour cell lines Three other tumour cell lines (DLDl, SKMEL28, and MDA- MD415) showed normal levels of SKIN expression (Figure 6 B-C).
  • tumour cell lines were selected to represent matched samples (overexpressing/not overexpressing) from the same type of tumour: colon carcinoma (HT29 and DLDl), melanoma (SKMEL28 and SKMEL5) and breast carcinoma (SKBR3, and MDA-MB-415) .
  • the KD of SKIN expression by siRNA dramatically reduced proliferation of all the overexpressing cell lines, whilst a control scrambled oligo had little, if any, effect.
  • SKIN KD did not inhibit proliferation of tumour lines displaying no overexpression of SKIN ( Figure 6A) .
  • KNAs from control/ElA expressing cells were prepared, and profiled by Affymetrix Genechip technology using standard techniques .
  • HG-U133 chip A+B Affymetrix GeneChip technology
  • HG-U133 Human Genome U133 (HG-U133) Set, consisting of two GeneChip® arrays, contains almost 45,000 probe sets representing more than 39,000 transcripts derived from approximately 33,000 well-substantiated human genes. This set design uses sequences selected from GenBank®, dbEST, and RefSeq.
  • sequence clusters were created from the UniGene database (Build 133, April 20, 2001). They were then refined by analysis and comparison with a number of other publicly available databases including the Washington University EST trace repository and the University of California, Santa Cruz Golden Path human genome database (April 2001 release) .
  • the HG-U133A Array includes representation of the RefSeq database sequences and probe sets related to sequences previously represented on the Human Genome U95Av2 Array.
  • the HG-U133B Array contains primarily probe sets representing EST clusters.
  • Affymetrix® Microarray Suite version 5.0 was used to normalised and pre-filter the data, with the following procedure : -The detection algorithm of the software was used to calculate a Detection p-value (see Manual for further details) and assign a Present, Marginal, Absent call of the signal for each spot on the array. Features (gene) always called Absent in every arrays were excluded. - The intensity signal of each transcript probed on the array, should be more than 200 (the range of signal is normally between 10 and 20.000) after MAS5 computing and normalisation.
  • Chip normalisation The median intensity of the signals of all the transcripts probed (probe pairs) on the array was computed (global median) and this value is used to divide again the signal of each probe pair. This procedure is called Chip normalisation.
  • the class prediction isolates a gene.
  • step 3 Selects the smallest p-value calculated in step 2 and converts it into prediction strength by taking negative natural log of the p-value.
  • Genespring 6.2 ® (www.silicongenetics.com) was used to perform the analyses .
  • the predictor is able to determine the risk to develop metastasis within 5 years .
  • Upregulation of one or more genes (e.g., of mRNA levels) in table 1 may be associated with worsening of the prognosis.
  • mRNA upregulation is associated with a poor prognosis. Patients were considered as having a "good” signature if they had lower values of at least 7 of the 13 genes, compared to other individuals in the sample (the "poor” group) .
  • the 13 gene predictor of the present invention is able to identify four more patients which went on to develop metastasis, as compared to the Van't Veer predictor.
  • the 46 ER positive patients it is able to correctly identify 6 more patients as compared to the Van't Veer dataset.
  • the Van't Veer predictor comprises 70 genes, whereas the present predictor makes use of only 13.
  • the ability to use a smaller set of genes without comprising accuracy is important in the clinical application of the predictor, diminishing costs and allowing a larger range of techniques to be used. Alternatively, more genes could be added to the set to provide a further improvement in accuracy.
  • the expression profile of the 13 breast gene predictor on 36 NO breast cancer patients analysed by Affymetrix was further confirmed by Q-RT-PCR. The classifier performance was also confirmed by Q-RT-PCR. Q-RT-PCR reactions were performed using default settings suggested by Applied Biosystem.
  • RNA expression values of patients with lung adenocarcinomas from two independent cohorts, and more precisely: the Beer dataset (Affymetrix GeneChip HU6800) is composed by 23 patients with disease-free-survival (DFS) more than 52 months and 18 patients with relapse time (Dead-of- disease) less than 29 months,- the Bhattacharjee dataset (Affymetrix GeneChip HG-U95Av2.1) is composed by 33 patients with DFS more than 30 months and 27 with relapse time (Dead- of-disease) less than 25 months.
  • DFS disease-free-survival
  • Bhattacharjee dataset Affymetrix GeneChip HG-U95Av2.1
  • Figure 9 shows the survival probability with a good or poor expression signature based on the NSCLC predictor, using the dataset of Beer et al ( Figure 9A) or the dataset of Bhattacharjee et al ( Figure 9B) .
  • downregulation of HLA-DQBl, LU, GNS, POLR2C, PBXIPl and RAFTLIN and upregulation of E2F4 , PAICS, PFN2, SERPINB5 , HSPDl , and ARL4A may be associated with a worsening of the prognosis.
  • a good signature was considered to be one which has at least 7 out of the 12 genes (i.e., the majority of genes) which are:
  • the expression profile of the 12 lung gene predictor on an independent set of patients composed of 30 tissue specimens (all stage I NSCLC adenocarcinomas) was also evaluated by Q- RT-PCR.
  • the "test" set of patients was composed of 15 patients without evidence of disease (the good outcome group) and 15 patients died of disease (the poor outcome group) .
  • the results of the "test” screening confirmed the good performance of our 12 genes classifier (see figure 9C) .
  • downregulation of HLA-DQBl, SCGB3A1 and RAFTLIN and upregulation of PFN2, SERPINB5, E2F4, E2F1, MCM7, RRM2, MCM4, MCM6 , CML66, SF3B1, ATP13A3, CXCL6, GABPB2 , GAPDH, GARS, HOXB7, HSPG2 and KIAA0186 may be associated with a worsening of the prognosis .
  • a good signature was considered to be one which has at least 11 genes out of the 21 genes (i.e., the majority of genes) which are:
  • SERPINB5 E2F4, E2F1, MCM7, RRM2 , MCM4, MCM6, CML66, SF3B1, ATP13A3, CXCL6 , GABPB2 , GAPDH, GARS, HOXB7 , HSPG2, KIAA0186 mRNA, downregulated compared to other individuals in the analysis (the poor prognosis group) .
  • Class D genes encode for rather heterogeneous proteins, including proteins involved in RNA splicing (SAPl and Sit ⁇ u-1) , a nucleolar RNA helicase (DDX21) , a microtubule-associated protein (Ch-TOG) , a component of the TNF-Rl pathway leading to activation of NF-KB (TRPC4AP) , and a previously unknown protein displaying no distinguishing dominial feature (SKIN) .
  • SAPl and Sit ⁇ u-1 proteins involved in RNA splicing
  • DDX21 nucleolar RNA helicase
  • Ch-TOG microtubule-associated protein
  • TRPC4AP a component of the TNF-Rl pathway leading to activation of NF-KB
  • SKIN previously unknown protein displaying no distinguishing dominial feature
  • Microarray expression datasets of the Michigan obtained on the HU6800 Affymetrix chip
  • of the Harvard cohorts obtained on the HU95Av2 Affymetrix chip
  • details of patient selection criteria and methods for data normalization can be downloaded from http : / /dot .ped.med.umich. edu : 2000/ourimage/pub/Lung/index . html.
  • Affymetrix CEL format files were processed using Affymetrix Microarray Suite v.5 software (MAS 5) .
  • Amy Peng Lam Microarray spot intensities below the minimum value of 10 (the BRB software default for Affymetrix array analysis) were excluded and arrays were then normalized (centered) using the median value of the signal over the entire array.
  • genes were excluded if less than 20% of their expression data across the patients had at least a 1.5-fold change in either direction from the gene's median value. Genes were also excluded if the percentage of data missing or filtered out exceeded 75%. All data were log- transformed (base 2) . The two-sample parametric t-test was used to select significant genes.
  • TaqMan ® Low Density arrays were purchased from Applied Biosystems. Total RNA (0.5 ⁇ g) was reverse transcribed with 200 units of Superscript II RT (Invitrogen) and 250 ng random examers, using manufacturer's instructions. A reaction mix containing 75 ng of cDNA and 50 ⁇ l of 2X PCR Master Mix (Euregentec) in a final volume of 100 ⁇ l was then prepared and loaded in the array. PCR conditions were as follows: 2 min at 50 0 C, 10 min at 94.5°C, followed by 45 cycles at 97°C for 30 s and 59.7°C for 1 min, on an Applied Biosystems 7900HT PCR System.
  • the expression level of each gene was measured in triplicate, and a panel of 8 reference genes ⁇ RPL14, RPL18, AGPATl, ACTB, TBP, GUSB, PPIA, 18S) was used.
  • GeNorm software (Vandesompele, J., et al, 2002. Genome Biol 3 : RESEARCHO034) was used to evaluate the expression stability of the reference genes .
  • the average Ct value of each target gene was normalized against the geometric mean of the Ct values of the 8 reference genes. Universal Reference RNA (Stratagene) was used as calibrator for all the samples analyzed.
  • TaqMan ® assay IDs were: NUDCDl/CML66-Hs00292614_ml, CXCL6- HsO0237017_ml, E2Fl-Hs00153451_ml, E2F4-Hs00608098_ml, GABPB2- Hs00242573_ml, HLA-DQBl-HsOO409790_ml , HOXB7-Hs00270131_ml , HSPG2-Hs00194179_ml, MCM4-Hs00381533_ml, MCM6-Hs00195504_ml, MCM7-Hs00428518_ml, RAFTLIN-Hs00412084_ml , RRM2-Hs00357247_gl , SCGB3Al-Hs00369360_gl, SERPINB5-Hs00184728_ml , SF3B1- H
  • Table 5 The 49-gene model was tested for prognostic predictive accuracy by leave-one-out cross-validation. Two other models of 50 and 100 genes, respectively, from Beer et al . (as above) were tested, as a comparison. Models were tested on the reduced datasets (top) and on the original datasets (bottom) from the Michigan, Harvard, and Duke cohorts (N, number of patients in the dataset) .
  • the performance of the 71-gene model in the leave-one-out cross-validation is shown here, in comparison to the 49-gene model and to the 50- and 100-gene models from Beer et al. (as above), for the modified reduced datasets, the reduced datasets, and the original datasets. As shown, the 71-gene model did not perform better than the 49-gene model, despite containing 22 additional genes .
  • the performance of the 49-gene model was tested by Kaplan-Meier analysis on stage I adenocarcinomas (Fig. 10) .
  • the 49-gene model was very effective in predicting overall survival in the stage I patients both from the Michigan and Harvard cohorts (Fig. 10 and Fig. 13) .
  • the 49-gene model proved remarkably effective in predicting prognosis (Table 5, and table S), and overall survival (Fig. 10, and fig 13) .
  • the 49-gene model performed better that the 50- and 100-gene models (Beer et al) both in the prediction of prognosis (Table 5) and of overall survival (Fig. 13B), when tested on the Duke cohort, which might be considered as a validation cohort for the three models .
  • the biased signature could effectively predict overall survival, further confirming that a biased approach can lead to the discovery of cancer-relevant signatures (see also table 7, below) .
  • Table 7 These data are supplemental to those shown in Fig. HA. Since datasets from the Michigan, Harvard and Duke studies were obtained on different generations of chips, not all of the 28 genes of the biased signature were present on the chips. In particular, only 5 and 11 genes of 28 were present on the chips used in the Michigan and Harvard studies, respectively. This prevented Kaplan-Meyer analysis on the Michigan and Harvard cohorts, which, on the other hand, could be meaningfully performed on the Duke cohorts, since all the 28 genes were present in the datasets (see Fig. HA) .
  • Table 8 Analysis of the 80-gene model in the, Michigan, Harvard and Duke cohort and in the IFOM training cohort .
  • Table 8 The genes of the 80-gene model are shown with their human gene symbol, name, human accession number and source (Liter., from literature; ElA, from the 28-gene biased signature; Meta. , from meta-analysis of the reduced Michigan and Harvard datasets) .
  • Table 8a simply separates out the genes having an ElA source;
  • Table 8b simply separates out the genes having a meta-analysis source.
  • the column 'TREND' indicates the assigned regulation trend of each gene transcript in the poor prognosis group compared to the good prognosis group, based on the dominant trend, and is intended to provide an indication as to how these genes may be associated with worsening of prognosis. All the 28 ElA genes are considered to be up regulated in the poor prognosis class, based on the results of our previous ElA screening. B2007/001343
  • the final prognostic model was obtained by the leave-one-out cross-validation procedure, with independent gene selection (p ⁇ 0.05 as cutoff; parametric t-test) .
  • a 10-gene model displayed a predictive accuracy of 84% (sensitivity, 90%; specificity, 80%) and a p-value of 0.004, after 2000 random permutations of class labels (Table 9) .
  • Table 9 Final prognostic model obtained by the leave-one-out cross-validation procedure of the IFOM training cohort.
  • the genes of the 10-gene model are shown with their human gene symbol, name and accession number and sorted by the t-values derived from the parametric t-test. Fold ratio indicates the average mRNA fold increase (>1.0) or decrease ( ⁇ 1.0) in the bad prognosis group, compared to the good prognosis group (followed by its p-value) .
  • Upregulation of all genes other than SCGB3A1, and downregulation of SCGB3A1 may be associated with a worsening of the prognosis .
  • we used it on an additional, independent cohort of 45 stage-I lung adenocarcinomas (henceforth "the IFOM validation cohort", see Methods) .
  • Univariate and multivariate analysis showed that the 10-gene model predicted survival of patients more accurately than tumor size, grading, age, sex or presence of mutated KRAS (Table 10) .
  • Table 10 Univariate and multivariate analysis of various prognostic markers.
  • the 10-gene model was tested for prediction of survival in the indicated cohorts of patients, in comparison to other biological or biochemical parameters, in univariate and multivariate analysis. Data are expressed as odds ratio (OD) at 95% confidence interval (95% CI) . Asterisks indicate statistically significant values.
  • a poor prognosis patient was considered as one having 3 or more genes following the poor prognosis pattern, namely: upregulated in the poor prognosis compared to the good prognosis group for all genes except for SCGB3A1; downregulated in the poor prognosis group compared to the good prognosis group in the case of SCGB3A1.
  • patients could be considered “poor prognosis” having a different number of genes following the poor prognosis pattern.
  • patients having 2 or more, or 4 or 5 or more genes following the "poor prognosis” pattern could be placed in the poor prognosis class .
  • stage IA lung adenocarcinoma The 5-year survival rate of clinical stage IA lung cancer patients ranges from 67% to 77% (28-30) after surgery alone. Thus, in this group, molecular tools for patient stratification are greatly needed, to select high-risk patients eligible for adjuvant chemotherapy. As shown in Fig. 12B, our 10-gene model displayed very good predictive power both in the IFOM and in the Duke (stage IA) cohorts .
  • the inventors measured performance of the 10-genes model compared to the 21-genes set of table 3.
  • the prognostic predictive accuracies of the 10-genes model, the 21-genes model and the 11-genes model obtained after subtracting the 10 genes of the 10-genes model from the 21- genes set) on the training cohort (25 patients) , testing cohort (45 patients) and the entire data set (70 patients, training + testing cohorts) were measured.
  • the definition of the classifier was performed by applying diagonal linear discriminant analysis (DLDA) and leave-one-out cross-validation.
  • DLDA diagonal linear discriminant analysis
  • the 'Test Cohort' prediction of prognosis was performed by applying the diagonal linear discriminant analysis (DLDA) classification method using the gene expression values of the training cohort to train the classifier. The results are shown below. Remarkably, it can be seen that the reduced, ten gene dataset performs better than the 21 gene dataset despite containing fewer genes .
  • Accuracy (Ace), percentage of correctly predicted patients; sensitivity (Sen.), probability for a patient with poor prognosis to be predicted as with poor prognosis; specificity (Spec), probability for a patient with good prognosis to be predicted as with good prognosis .
  • composition of the final 10- gene prognostic model confirmed the hypothesized efficacy of our integrated approach, as it consisted of 5 genes derived from the ElA signature (NUDCDl, E2F1, MCM6, RRM2 and SF3B1) , 4 genes from the metanalysis of microarray data (H0XB7, SERPINB5, E2F4 and HSPG2) , and 1 gene (SCGB3A1) from the literature .

Landscapes

  • Chemical & Material Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Organic Chemistry (AREA)
  • Proteomics, Peptides & Aminoacids (AREA)
  • Engineering & Computer Science (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Analytical Chemistry (AREA)
  • Zoology (AREA)
  • Genetics & Genomics (AREA)
  • Wood Science & Technology (AREA)
  • Physics & Mathematics (AREA)
  • Biotechnology (AREA)
  • Microbiology (AREA)
  • Molecular Biology (AREA)
  • Hospice & Palliative Care (AREA)
  • Biophysics (AREA)
  • Oncology (AREA)
  • Biochemistry (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)

Abstract

La présente invention concerne des procédés de diagnostic et de pronostic du cancer, et en particulier le NSCLC, les procédés comprenant la détermination du niveau d'expression d'un ou de plusieurs gènes. Selon certains modes de réalisation, l'invention concerne le pronostic du NSCLC à un stade précoce.
PCT/GB2007/001343 2004-09-30 2007-04-12 Marqueurs de cancer Ceased WO2008125791A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/GB2007/001343 WO2008125791A1 (fr) 2007-04-12 2007-04-12 Marqueurs de cancer
US12/576,812 US8747867B2 (en) 2004-09-30 2007-04-12 Cancer markers

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/GB2007/001343 WO2008125791A1 (fr) 2007-04-12 2007-04-12 Marqueurs de cancer

Related Parent Applications (3)

Application Number Title Priority Date Filing Date
US11/664,137 Continuation-In-Part US7901876B2 (en) 2004-09-30 2005-09-20 Cancer markers
PCT/EP2005/010153 Continuation-In-Part WO2006037462A2 (fr) 2004-09-30 2005-09-20 Marqueurs du cancer
US66413707A Continuation-In-Part 2004-09-30 2007-03-29

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US11/576,812 Continuation US7944306B2 (en) 2004-10-08 2005-10-05 Dual bias control circuit

Publications (1)

Publication Number Publication Date
WO2008125791A1 true WO2008125791A1 (fr) 2008-10-23

Family

ID=38123853

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/GB2007/001343 Ceased WO2008125791A1 (fr) 2004-09-30 2007-04-12 Marqueurs de cancer

Country Status (1)

Country Link
WO (1) WO2008125791A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103205433A (zh) * 2013-04-03 2013-07-17 复旦大学附属肿瘤医院 用于肺癌预后的基因及其应用

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050272061A1 (en) * 2004-02-19 2005-12-08 Seattle Genetics, Inc. Expression profiling in non-small cell lung cancer
WO2006138275A2 (fr) * 2005-06-13 2006-12-28 The Regents Of The University Of Michigan Compositions et procedes de traitement et de diagnostic de cancer

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050272061A1 (en) * 2004-02-19 2005-12-08 Seattle Genetics, Inc. Expression profiling in non-small cell lung cancer
WO2006138275A2 (fr) * 2005-06-13 2006-12-28 The Regents Of The University Of Michigan Compositions et procedes de traitement et de diagnostic de cancer

Non-Patent Citations (7)

* Cited by examiner, † Cited by third party
Title
CHEN HSUAN-YU ET AL: "A five-gene signature and clinical outcome in non-small-cell lung cancer.", THE NEW ENGLAND JOURNAL OF MEDICINE 4 JAN 2007, vol. 356, no. 1, 4 January 2007 (2007-01-04), pages 11 - 20, XP009086044, ISSN: 1533-4406 *
DAVIDSON JENNIFER D ET AL: "An increase in the expression of ribonucleotide reductase large subunit 1 is associated with gemcitabine resistance in non-small cell lung cancer cell lines.", CANCER RESEARCH 1 JUN 2004, vol. 64, no. 11, 1 June 2004 (2004-06-01), pages 3761 - 3766, XP002440531, ISSN: 0008-5472 *
ENDOH HIDEKI ET AL: "Prognostic model of pulmonary adenocarcinoma by expression profiling of eight genes as determined by quantitative real-time reverse transcriptase polymerase chain reaction.", JOURNAL OF CLINICAL ONCOLOGY : OFFICIAL JOURNAL OF THE AMERICAN SOCIETY OF CLINICAL ONCOLOGY 1 MAR 2004, vol. 22, no. 5, 1 March 2004 (2004-03-01), pages 811 - 819, XP002440527, ISSN: 0732-183X *
KAWAI HIDEKI ET AL: "Estrogen receptor alpha and beta are prognostic factors in non-small cell lung cancer.", CLINICAL CANCER RESEARCH : AN OFFICIAL JOURNAL OF THE AMERICAN ASSOCIATION FOR CANCER RESEARCH 15 JUL 2005, vol. 11, no. 14, 15 July 2005 (2005-07-15), pages 5084 - 5089, XP002440528, ISSN: 1078-0432 *
NICASSIO FRANCESCO ET AL: "A cancer-specific transcriptional signature in human neoplasia.", THE JOURNAL OF CLINICAL INVESTIGATION NOV 2005, vol. 115, no. 11, November 2005 (2005-11-01), pages 3015 - 3025, XP002440529, ISSN: 0021-9738 *
ROSELL RAFAEL ET AL: "Gene expression as a predictive marker of outcome in stage IIB-IIIA-IIIB non-small cell lung cancer after induction gemcitabine-based chemotherapy followed by resectional surgery.", CLINICAL CANCER RESEARCH : AN OFFICIAL JOURNAL OF THE AMERICAN ASSOCIATION FOR CANCER RESEARCH 15 JUN 2004, vol. 10, no. 12 Pt 2, 15 June 2004 (2004-06-15), pages 4215s - 4219s, XP002440530, ISSN: 1078-0432 *
VALLADARES ET AL: "Genetic expression profiles and chromosomal alterations in sporadic breast cancer in Mexican women", CANCER GENETICS AND CYTOGENETICS, ELSEVIER SCIENCE PUBLISHING, NEW YORK, NY, US, vol. 170, no. 2, 15 October 2006 (2006-10-15), pages 147 - 151, XP005671350, ISSN: 0165-4608 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103205433A (zh) * 2013-04-03 2013-07-17 复旦大学附属肿瘤医院 用于肺癌预后的基因及其应用
CN103205433B (zh) * 2013-04-03 2014-06-04 复旦大学附属肿瘤医院 用于肺癌预后的基因及其应用

Similar Documents

Publication Publication Date Title
CA2688477C (fr) Test pronostique multigenique pour le cancer du poumon
US20090280493A1 (en) Methods and Compositions for the Prediction of Response to Trastuzumab Containing Chemotherapy Regimen in Malignant Neoplasia
US10407736B2 (en) Expression of ETS related gene (ERG) and phosphatase and tensin homolog (PTEN) correlates with prostate cancer capsular penetration
US20160168649A1 (en) Gene expression predictors of cancer prognosis
Chang et al. Comparison of genomic signatures of non-small cell lung cancer recurrence between two microarray platforms
US20100178651A1 (en) Bifunctional Predictors of Cancer Treatment Sensitivity and Resistance
JP2011500071A (ja) 遺伝子に基づくアルゴリズム的ガン予後及び患者の臨床結果
CA2699385A1 (fr) Biomarqueurs du cancer de la prostate
Ho et al. Promoter methylation status of HIN-1 associated with outcomes of ovarian clear cell adenocarcinoma
Adam et al. Epidermal growth factor receptor in adrenocortical tumors: analysis of gene sequence, protein expression and correlation with clinical outcome
EP2061885A1 (fr) Indicateur du cancer du sein dérivé du stroma
US20150344962A1 (en) Methods for evaluating breast cancer prognosis
US8747867B2 (en) Cancer markers
US20110159509A1 (en) Use of runx3 and mir-532-5p as cancer markers and therapeutic targets
JP2008514209A (ja) 癌マーカー
WO2012112645A1 (fr) Marqueurs pour l'identification des modalités de traitement du cancer du sein
EP2278026A1 (fr) Procédé de prédiction du résultat clinique de patients atteints d'un carcinome du sein
US9606122B2 (en) Prognosis of oesophageal and gastro-oesophageal junctional cancer
WO2007132156A2 (fr) Matières et procédés associés au diagnostic, au pronostic et au traitement de cancers reposant sur la détermination de nouveaux marqueurs moléculaires dans des tumeurs
WO2009068409A1 (fr) Procédé d'évaluation d'un pronostic et de prédiction d'une réponse thérapeutique à un traitement endocrinien
WO2001004343A2 (fr) Procede permettant d'etablir un pronostic pour des patients atteints d'un cancer, par mesure des niveaux d'expression de la proteine bag
KR102384992B1 (ko) 대장암 환자의 연령 특이적 바이오마커 및 이의 용도
WO2008125791A1 (fr) Marqueurs de cancer
EP2065473A1 (fr) Procédé d'évaluation de pronostic et de prédiction de succès thérapeutique dans un cancer gynécologique
KR20230086458A (ko) 전이성 유방암의 항암제 치료 반응성 및 예후 예측용 신규한 바이오마커 및 이의 용도

Legal Events

Date Code Title Description
DPE2 Request for preliminary examination filed before expiration of 19th month from priority date (pct application filed from 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 07732386

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 07732386

Country of ref document: EP

Kind code of ref document: A1