Mapping The Cancer Genome by liwenting

VIEWS: 13 PAGES: 42

									 Mapping The Cancer Genome

        Francis S. Collins and Anne D. Barker
                 Scientific American
                    Feb18, 2007
http://www.sciam.com/article.cfm?chanID=sa006&colID=1&articleID=CC1E538E
-E7F2-99DF-3F44D06D3B292CF3
“If we wish to learn more about cancer, we must now
   concentrate on the cellular genome”
        Renato Delbecco.

   This quote originates from more than 20 years ago. (pre-
    Human Genome Project)

   Recently, the NIH launched “The Cancer Genome Atlas”
    (TCGA) project to catalog the genetic changes that cause
    cancer.
                           Cancer
   Not a disease with a singular cause.
     This is not understood by the general public.


   Cancer arises from progressive accumulation
    of mutation.
       Genes that can be involved in cancer progression number
        in the hundreds
       Mutations can arise from environmental factors or simple
        errors in DNA replication
•    A cell must accomplish several things before it can become
     transformed (cancerous).
    1. Escape normal growth factor/apoptosis signaling
       –   Are normally held in non-dividing state and killed off if they appear
           to be getting out of control.
    2. Get out of normal cell cycle checkpoint control
       –   Allows replication with damaged DNA
    3. Become immortal
       –   Through activation of telomerase and other methods.


–    Each one of these steps can be accomplished in a huge
     number of ways.
–    A seemingly infinite number of combinations of mutations
     can lead to cancer.
                     Cancer
   Categorized by the location of the primary
    tumors or type of cell involved.
      Quite a naïve system.
         Tumors found in the same location in

          different people can be vastly different.
         Cells within the same tumor can be vastly

          different.
      These cryptic differences can have a huge
       impact on the outcome of the patient.
Diagnosis is Still Largely Histological
                   Easy to see that cells are
                    abnormal, and should
                    probably be removed.
                   Important characteristics
                    completely addressed
• For the patient, the important characteristics of
  a tumor are:
   1.Risk of metastasis
   2.Response to chemotherapy or other
     treatments.
                      Metastasis
   The spread of cancerous cells from an original
    tumor to other parts of the body. These cells
    then give rise to tumors in other organ systems.
       Cells detatch from the tumor, work their way
        through the extracelluar matrix
       Travels through the lymphatic or circulatory system
        to another site.
   If metastasis has occurred the prognosis is
    poor.
   Secondary tumors often occur in the brain, lung
    and liver.
         Response to Chemotherapy
   A very “Big Hammer” approach.
       Cancerous cells are out of control because they are
        dividing in an un-regulated fashion.
       Goal is to kill all dividing cells.
   However:
       Tumors in the same tissue have varying response
        to Chemo
       Very Very hard on the patient.
     Cancer in the Pre-Genomics Era

   Over the last 20 years or so, about 350 genes
    were identified that contributed to a cancerous
    state when mutated.
   Identified in a low-throughput, one at a time
    manner. Lots of money, time and PhDs.
Can we look at the genetic makeup of a
 tumor to determine its likely behavior?
     Cancer in the Post-Genomic Era

   Compiling The Cancer Genome Atlas
   Catalogue the set of genes potentially involved
    in all types of cancer through the use of high
    throughput sequencing
           Early Proof of Concept

   Wellcome Trust Sanger Institute, 2001
      Sequenced 20 genes from 378 malignant
       melanoma tumor samples
      Found B-RAF was mutated in about 70%
      A drug inhibiting the B-RAF pathway is now
       in clinical trials
   Similar investigations have been completed
    with other cancer types.
               Upping the Scale

   Johns Hopkins University
      Sequenced 13,000 genes from 22 tumor
       samples
          11 Breast Cancer, 11 Colorectal Cancer


      Discovered significant mutations in over 200
       genes.
      Around a dozen genes were thought to be
       involved in these cancers previously.
               The Scale of TCGA

   For each type of cancer:
      Sequence the genomes of hundreds of
       tumors in order to saturate the set of
       involved genes
      Sort out the important mutations from the
       background mutations accumulated by the
       genetically unstable cells.
          Factor in heterogeneity within the tumors.




   Expand this to the 200ish known cancer types.
             Compiling the Atlas

   Pilot project by National Cancer Institute and
    National Human Genome Research Institute
      Generate the Atlas for brain, lung and
       ovarian cancer at a cost of 100 million
       dollars
   These three cancers account for 200,000
    deaths per year.
   Lots of starting material that meets the
    project’s requirements.
           Components of the Project
   Collecting and processing samples, generating and analyzing
    sequence requires infrastructure. Project will be modular,
    separated into:
    – Bio-specimen Core Research Center
          Collect specimens


    – Cancer Genome Characterization Centers
          Identify potential interesting gene targets to sequence,

           likely several thousand genes.
    – Genome Sequencing Centers
          Generate sequence data for each of the thousands of

           genes for 1500 tumor samples
    – Data Coordinating Centers
          Put it all together
             Uncharted Territory
   We’ve got the technical knowledge compile
    the Atlas (if at great cost)
   But, how do we turn this knowledge into
    strategies for treatment / prevention?
       Relate the knowledge of the genetic makeup of a
        person’s cancer to treatment options.
Ethical and Policy Concerns
          • Will need legislation to prevent
          discrimination based on genetic
          factors
          • Knowledge of whether a
          cancer could be easily treated,
          the survival rate, etc. is of great
          interest to insurance companies,
          HMOs
Genetic tests offer promise but raise
            questions too

         The New York Times Feb. 18, 2005
                 Denise Caruso




 http://www.nytimes.com/2007/02/18/business/yourmoney/18reframe.html
         DNA testing as an industry
   More than 1000 genetic tests available.
     Attempt to predict disease by looking for genetic
      markers.
   Questions of reliability
       correlation between a genetic trait and the actual
        disease?
       relative reliability if different tests for the same
        disease?
            Discovering “the gene”?
   Many discoveries linking genes to diseases
       Diabetes
       Alzheimer's
       Schizophrenia
       Depression etc…
   These are not a “one gene, one disease”
    relationships
       Often very complicated genetic and environmental
        influences.
       There is no “Gene for Alzheimer's” but there are genes
        that correlate with the disease
             Regulatory Black Hole
   No laws regulating the effectivness of these
    tests.
   Not even that a test must actually indicate the
    presense, or risk of a disease.
       The correlations are often quite thin
   FDA is working on guidelines, but has no formal
    aproval process for these tests.
       Those that have been suggested are canned by the
        testing industry as “disincentive to innovation”
    Economic benefit vs. ethical wisdom

   Cannot blame diagnostics makers wanting
    freedom to develop new and important market
   But genetic tests may lead to life-changing
    decisions
       Take drug or not
       carry child to term or terminate pregnancy
       Have surgery or not
      GENOMIC MEDICINE:
Gene expression tests foretell breast
         cancer's future
          Science, Vol 303,1754-1755; 2004



     http://www.sciencemag.org/cgi/content/full/303/5665/1754
                         Breast cancer
• National Alliance of Breast Cancer
  Organizations:
      Excluding skin cancers, breast cancer is the most
       common form of cancer in women in the US
           In US over 200,000 new cases of female invasive breast
            cancer diagnosed per year (39,000 deaths)
      Men can develop breast cancer too
           Low incidence – in US ~1,500 cases diagnosed per year
            (400 deaths)




                                http://www.nabco.org/index.php/39/index.php/137
                     Breast cancer

   In the US about 54,000 cases of                female
    pre-invasive breast cancer will be diagnosed this year
       ~ 88% will be ductal carcinoma in situ (DCIS).


   After lung cancer, breast cancer is second leading
    cause of cancer death for all women
   It is the leading overall cause of cancer death in
    women between the ages of 40 and 59



                              http://www.nabco.org/index.php/39/index.php/137
                         Therapy options
   If cancer is not in lymph nodes, most patients are
    cured by surgery and tamoxifen (estrogen-inhibiting
    drug)
   Development of metastases usually leads to death
       Reduction of this risk if you take chemotherapy
       But chemotherapy can lead to
            heart failure
            leukemia
            life threatening infections
       Have to decide if you want to take this route
         How to decide what therapy?

   Current pathology examinations and molecular
    markers not conclusive

       May broadly hint if tumour will metastasize

       Not individualized approach
                          Oncotype DX

   Individualized testing service for
    determining a breast cancer
    patient’s chance of developing
    metastases

   Test based on genetic activity in
    tumor

   Test may be relevant to up to 1/3 of
    breast cancer patients
                    Oncotype DX
   Tumour biopsy sample is encased in paraffin wax
    and mailed to company

   Company does RNA extraction and gene
    expression analysis

   Physician receives report
    – http://www.genomichealth.com/oncotype/about/how.aspx
              How test was developed
   250 genes selected that were
    believed to have influence over
    cancer progression

   Measured genes’ expression levels
    in tumor samples
        Found 16 genes that predicted
         metastases and picked five control
         genes

   Tested 668 tumours from another
    trial to validate results
                  Validation study
   Gene expression patterns were analyzed and
    classified
    - Low, medium and high risk groups

   Good results: After ten years only 6.8% in low risk
    group developed metastases, compared to 30.4% in
    high risk group

   More accurate than current methods to determine
    metastasis risk
                   Some limitations…
   Cannot predict which patients are good candidates
    for chemotherapy
       Useful to predict outcome, but not benefits from
        chemotherapy
   Based on tumours that express estrogen receptor
       Based on patients who took tamoxifen
   Based on patients whose tumours did not spread to
    lymph nodes
    = fewer than half of patients
   Assumed that those with high risk of metastases will
    benefit most from chemotherapy
   Based on tumours from study 20 years ago
       Tumours may have changed over time
   Sample may have been selective
       only from tumours large enough to have tissue left over
        after pathology tests
 Other genes may be better predictors for benefit
  from chemotherapy
• Clinical trial is planned – track metastases,
  outcome and patient response to chemotherapy
   Oncotype DX falls into “homebrew” category of
    diagnostic test.
       i.e. it is a “clinical laboratory reference service”
            Thus exempt from standard FDA review
       Approved by California State Licensing Agency
            Regulates labs in state
               Agendia Mammaprint

   Breast cancer profiling test
   Generated gene expression data for 25,000
    different genes from group of 117 women
       Breast cancer that had not spread to lymph nodes
       Clinical course was followed for more than 5 years
        after surgery
             Agendia Mammaprint

   Identified 70-gene expression signature to
    indicate good or bad prognosis
   Showed that very early, small tumours can be
    predicted to become bad players later
    – First to show that gene expression patterns can
      identify these
                 Validation study
   Studied 295 tumours from Netherlands
    Cancer Institute

   Only tumours of untreated patients tested

   Almost 56% of lymph-node negative patients with
    poor prognosis developed metastases after 10 years

   Only 13% of patients with good prognosis developed
    metastases after 10 years
                   Flaws in study

   Also used stored samples

   Only women under age of 53 were tested

   Only samples from single institution

   Test group (where 70 indicator genes came from)
    was included in results
                 PRESS RELEASE
   Agendia's MammaPrint breast cancer prognosis test
    cleared by U.S. Food and Drug Administration
    (FDA).
   MammaPrint® is the first multi-gene expression test
    to receive market clearance by the FDA




    http://www.agendia.com/index.php?option=com_content&task=view&id
    =162&Itemid=239
              Future of this approach

   Tests may allow women with low risk signature to
    forego chemotherapy
       Expect 25-35% reduction in chemotherapy in Europe
       US doctors more likely to keep chemotherapy for marginal
        cases – everything will stay same for now


   Test may show which small tumours may turn out
    bad and help point out where chemotherapy may be
    helpful

								
To top