Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out

familiality.8.24.06.ppt - Department of Genetic Epidemiology

VIEWS: 0 PAGES: 42

									 The familial/genetic component of disease
Diseases can be arranged on a continuum:

purely                                                           purely
genetic ------------------------------------------------------- environmental

(Hemophilia)                                               (poisoning)

Most lie between the two extremes and are the result of
interaction of both types of factors

How can we identify those diseases with a genetic
component?
Understanding the genetic contribution to health

   At least some genetic component can be argued for most human traits:

      – disease     (breast cancer, intracranial aneurysm)

      – health     (longevity, high bone mineral density)

      – behavior   (depression, anxiety)

      – response to medication (warfarin and CYP450)

      –…
       The familial/genetic component of disease

Evidence that inherited predisposition plays a role in disease
 risk comes from different sources:

1. A rare disease in excess in a pedigree(s)
2. Multiple pedigrees with statistical excess of common disease
3. Increased risk to relatives of affected individuals

Role of genetic predisposition is most clear in (1) and least clear
 in (3)
 The familial/genetic component of disease

Family aggregation could be attributed to:
      genes shared by family members,
      exposure to the same risk factors,
      or the interactive effects of the two

The biological interpretation of increased risks in relatives is a matter of
  speculation…

Familial versus genetic components are difficult to sort out
  The familial/genetic component of disease

Demonstration of family aggregation for a disease has a long history

No conclusions could be made regarding familial aggregation based
   on families with multiple cases of the disease of interest

For common diseases, the occurrence of multiple cases in a single
   family could easily be due to chance
 The familial/genetic component of disease
Increased risks to relatives:

Comparison of the frequency of a positive family history among
  cases to the frequency among controls


Problems:
   -information for distant relatives?

   -recall bias?
       affected individuals know more about their fam hx

   -ascertainment bias?
      source of cases/controls
 Analysis of familiality in the Utah Population Database


Goals of this course:

-Familiarity with large, genealogical databases

-Familiarity with the UPDB

-Familiarity with multiple methods of examining familiality
in the UPDB or a UPDB-like resource

-Examination of the familiality for a specific disease
  The familial/genetic component of disease

Necessary data elements for familiality studies:

Genealogy
    to provide genetic relationships

Phenotypes
    to correctly assign cases and controls
Utah Population Data Base (UPDB)

In the 1970’s Mark Skolnick proposed building a Utah
  population genealogy with linked phenotype data

He had constructed a genealogy in the Parma Valley, but
 medical records were too diffuse

The Iceland genealogy resource was already being
  developed

Skolnick, 1977 Prospects paper
                Utah Population Data Base History

Original Contributers:
Mark Skolnick, Ph.D.

Charlie Smart, M.D.            (UCR)
Lee Bean, Ph.D.                (Sociology)
Geri Mineau, Ph.D.             (Sociology graduate student)
  (now the Director of UPDB)
Roger Williams, M.D.           (Internal Medicine)

Tim Maness and                 (built the genealogy)
 Sue Dintelman
Others?
         Utah Population Data Base History
Mormons make up 75% of the state of Utah
 Individual Mormons trace their ancestries as far as possible
 Records collected in the Family History Library (FHL)

Skolnick was allowed access to 3-generation family genealogy
  sheets submitted by members of the Church of Jesus Christ of
  Latter Day Saints (in the early 1970’s)

Selected sheets from the FHL containing at least 1 life event in
  Utah or on the pioneer trail (1840-1850)

Resulting genealogy contained ~ 1.6 million individuals linked in
  genealogies 6-7 generations deep
      Funding and Administration History

UPDB was supported by NIH funding to Skolnick’s group (through LDS and Univ
  of Utah) for about 15 years then

       turned over to the State of Utah

       turned over to the University of Utah

       supported by the University School of Medicine Cancer Center
       (PI: Robert Stewart, M.D.)

       now supported by Huntsman Cancer Institute


Other funding support from LL Bean, K Smith, L Cannon Albright,… (U of U)
                   UPDB Progress
UPDB has become a dynamic database
     now receives annual updates for:
           Utah births and deaths
           Utah Cancer Registry
           Utah driver’s licenses
           Idaho Cancer Registry

UPDB is now person-oriented
      includes over 9 million records
      names and other info for over 12 million individuals
      multiple records represent over 9 million individuals

No new genealogy data, but triplets added from vital records
                          /www.hci.utah.edu/groups/ppr/webdata.html
           Records available in UPDB
Records Type                              Records
Genealogy                                 1.6 million

Cancer Records (Utah and Idaho)           300,000

Birth Certificates & Fetal Deaths        1.9 million
Death Certificates                       700,000
Fetal Deaths                               6,000
Marriage Certificates                    500,000
Divorce Records                          220,000
Driver License                           2.6 million
1880 Census                              143,000
HCFA follow-up                           139,000

Total Records                             8.7 million

Current population of Utah is ~ 2.2 million
                          New Data in UPDB
GenMed File

New in 2003
Link between medical information in the UUHSC Enterprise Data Resource Center and UPDB

UUHSC Data Resource Center contains over 1.4 million patient demographic records
Over 1 million (70%) have been matched to a “person record” in UPDB with about 730,000 (50%)
   matched to a person with two or more generations
For Utah residents the rates increase to 79% matched and 57% with two or more generations
Neither the patient demographic records nor the medical information from the UUHSC are incorporated
   into UPDB; each resource maintains its own records. However, the medical data and UPDB records are
   accessed at the time that a research specific project has been approved.

                                     http://www.hci.utah.edu/groups/ppr/webdata.html
 Utah Population Data Base (UPDB)
Mormons make up 75% of the state of Utah
    Individual Mormons trace their ancestries as far as possible
    Records collected in the Family History Library

UPDB used 3-generation family genealogy sheets submitted by
  members of the Church of Jesus Christ of Latter Day Saints

UPDB selected sheets containing > 1 life event in Utah or on the
  pioneer trail (1840-1850)

Original UPDB included 1.6 million individuals linked in genealogies
  6 or 7 generations deep
                         UPDB History
UPDB was supported by Skolnick’s group for about 15 years then

       given to the State of Utah

       given to the University of Utah

       supported by the Cancer Center

       absorbed by HCI

       GenData/Lineagen


Access governed by Resource for Genetic Epidemiology (RGE)
                   UPDB Permissions

Administered by RGE

Each research project must have an RGE project and permissions

Each research project must have IRB approval

**Each investigator must sign confidentiality statement
  Gen Epi Version of UPDB
No-name version of the UPDB at Gen Epi

Individuals who are part of at least 4 generations of a family are flagged

6497093 individuals, 2595249 are “G flagged”
         Utah “G” flagged Phenotype Data

Death Certificates   369,000+ death certificates
                     dating back to 1904 , up to 2004
                     coded with ICD revisions 6 - 10


Cancer Records       93,000+ NCI SEER registry records
                     coded with ICD-0
                     age, stage, grade, survival,
                     100% Utah ascertainment,
                     > 95% follow-up
                     Utah “G” Flag Resource
                          birth year      Males      Females
Death Certificates    }     <1856         39,924      38,159
     250,000          }
                      }   1856 - 1880     66,111       62,411
Cancer Data }         }
    80,000    }       }   1881 - 1905    118,479     111,519
              }       }
              }       }   1906 - 1930    160,228     148,735
              }       }
              }       }   1931 - 1955    193,159     182,124
              }       }
              }       }   1956 - 1980    304,396     308,987
              }       }
              }       }   1981 - 2003    258,011     244,811
                                        1,140,308   1,096,746
Sample Utah Prostate Pedigree (60 cases)
           Utah “G” flagged Phenotype Data
Death Certificates    360,000+ death certificates
                      dating back to 1904
                      coded with ICD revisions 6 - 10

Cancer Records        93,000+ NCI SEER registry records
                      coded with ICD-0
                      age, stage, grade, survival,
                      100% Utah ascertainment,
                      > 95% follow-up

Driver’s License      BMI
Birth Certificates    birth weight, APGAR,
Other                 longevity, fertility, offspring sex ratio

UUHSC Hospital data   dx, procedure, medications, images, labs?
   The familial/genetic component of disease

We will concentrate on 3 different methods:

- average relatedness measures

- estimation of risk of disease in relatives of affecteds

- identification of pedigrees with a statistical excess of
      affecteds
   The familial/genetic component of disease

We will concentrate on 3 different methods:

- average relatedness measures

- estimation of risk of disease in relatives of affecteds

- identification of pedigrees with a statistical excess of
      affecteds
     Utah Analysis of Familiality
– Genealogical Index of Familiality (GIF) developed
  for the UPDB in the 1970s

– Estimates average relatedness of all pairs in a
  group, using Malécot coefficient of kinship

– Significance measured by comparison to the mean
  GIF of 1000 sets of matched controls

– Similar to the familiality measure Iceland is using
Utah Analysis of Cancer Familiality
                                               mean
                                        case   control
Site                      n             GIF    GIF          p value
lip               1243           4.65   2.67    < 1.0 x e -16
small intestine    256           4.31   2.68     0.010
prostate      13969              3.48   2.56    < 1.0 x e -16
thyroid           1180           3.36   2.04    < 1.0 x e -16
melanoma                 3798           3.22   2.27       < 1.0 x e -16
testis                    531           3.21   2.06        2.5 x e -5
colorectal               8048           3.20   2.56       < 1.0 x e -16
stomach           1496           3.18   2.60     9.3 x e -5
lung                     4124           3.03   2.57        3.8 x e -8
lymphoma                 3451           2.94   2.37        8.8 x e -12
breast                   10924          2.92   2.42       < 1.0 x e -16
ovarian           1643           2.89   2.42     0.0006
Familiality of Prostate Cancer by Age at Diagnosis


Group                cases     GIF significance
Prostate Cancer      13969     3.48 <.000001
 early (<61)          1070     4.36 <.000001
 late (>70)           8492     3.59 <.000001
 later (>80)          2692     3.85 <.000001
 latest (>90)          229     3.62 0.2122
Familiality of Prostate Cancer by Clinical Features

                                 number of
Group                            cases       GIF    significance
By stage:
Local                             8247       3.47   <.00001
Regional                          2459       3.61   <.00001
Distant                           1300       3.39   0.00070

By grade:
Well-moderately differentiated   11514       3.49   <.00001
Poorly differentiated             2263       3.71   <.00001
Undifferentiated                   192       8.29   <.00001

By survival:
Survivors (> 10 yrs)              2113       4.11   <.00001
Non-survivors                      874       2.51   0.740
Utah Analysis of Familiality from Death Certificate Data
                                                     mean
                                            case     control
Site                                  n     GIF      GIF         p value
Diverticulosis               238    16.90    3.07   <.00001
Multiple Sclerosis           242    15.58    2.72   <.00001
Influenza                            325    13.83    2.92 <.00001
Asthma                               400    9.34     2.89 <.00001
Congenital anomalies         304    7.27     2.29   <.00001
Emphysema                           1096    6.86     2.91 <.00001
Cardiomyopathy                       647    6.45     2.83 <.00001
Parkinson’s                  756    6.18     2.94   <.00001
Alcohol related                      459    4.81     2.89 <.00001
Prostate Cancer                     3323    4.75     2.89 <.00001
Chronic Airway Obstruction   1782   4.26     2.88    <.00001
   The familial/genetic component of disease

We will concentrate on 3 different methods:

- average relatedness measures

- estimation of risk of disease in relatives of affecteds

- identification of pedigrees with a statistical excess of
      affecteds
                     Relative Risks


•Diseases with a heritable component are expected to occur at higher
frequency among relatives of the cases than among random controls.

•Must compare the rate of a particular disease among the relatives of
cases with that disease to the rate in the population

•e.g. first degree relatives of prostate cancer cases are thought to be at
approximately 3 times the population risk of prostate cancer


                                                                   2/98
First Degree Relative Risks for Aneurysms


581 Intracranial Aneurysm (IA) cases in the UPDB

5280 of their first degree relatives also have a death
  certificate

                              Obs Exp RR      p
Intracranial Aneurysm         19   4.7 4.05 <0.0001
   First Degree Relative Risks for All Aneurysm Types
                               Obs   Exp    RR      p
581 IA cases
5280 first degree relatives
Intracranial Aneurysm           19    4.7   4.05 <0.00001

1362 AA cases
14322 first degree relatives
Aortic Aneurysm                 69   33.3    2.07 <0.00001

82 OA cases
950 first degree relatives
Other Aneurysm                   2    0.1 14.13   0.0091
 Relative risks to study disease associations


•Relative risks for associated diseases

   •Diseases with a heritable component are expected to occur at higher
   frequency among relatives of the cases than among random controls.

   •Compare the rate of a particular disease among the relatives of cases
   with another disease to the rate in the population




                                                                  2/98
First Degree Relative Risks for Different Aneurysm Types

Aneurysm type         Obs   Exp      RR          p

IA with AA            12    12.4      0.97    0.59

IA with OA            0      0.7       --        --

AA with OA            7      1.9      3.64    0.0037
     Disease Associations (using Relative Risks)
2901 melanoma cases                23,419 first degree relatives

Cancer site           obs    exp            RR (95% CI)
Melanoma               190    56            3.39 (2.79, 3.55)
Eye                     14     5            2.61 (1.16, 5.00)
Multiple myeloma        43    23            1.85 (1.20, 2.71)
Pancreas                57    34            1.67 (1.16, 2.33)
Kidney                  42    26            1.60 (1.03, 2.35)
Stomach                 47    31            1.54 (1.02, 2.21)
Colon                  172   124            1.39 (1.13, 1.69)
Prostate               413   311            1.33 (1.17, 1.51)
Breast                 307   237            1.29 (1.11, 1.50)
   The familial/genetic component of disease

We will concentrate on 3 different methods:

- average relatedness measures

- estimation of risk of disease in relatives of affecteds

- identification of pedigrees with a statistical excess of
      affecteds
     Utah High Risk Pedigree Studies

Disease                 Time period      Pedigrees   DNAs
Breast Cancer           1972-1992          490       6215
Colon Cancer            1981-1991          272       3993
Melanoma                1986-1994          179       2361
Prostate Cancer         1990 - present     362       7868
Asthma                  1995-2000          198       2494
Osteoporosis            1995-2000          323       1918
Depression              1996-2000          438       2939
Intracranial Aneurysm   1998 - present     151        625
                                           Total     28,413
 Genes Localized and Isolated using
     extended Utah pedigrees
1987          Neurofibromatosis

1988          Alport Syndrome

1992-1994     Melanoma locus, p16 melanoma gene

1994          BRCA1 gene, BRCA2 locus
1996          BRCA2 gene

1997 - 2001   prostate locus, HPC2/ELAC2 gene

2003          Depressive Disorder locus
                       Semester Schedule 2006
Aug 24             Introduction                           Project 1
Aug 31             Utah tools and phenotypes              Project 2
Sep 7              Building a genealogy database          Sue and Tim
Sep 14             Genealogy Databases: UPDB, UINFO       Steve
Sep 21              Utah GIF method                       Project 3
Sep 28
Oct 5              NO CLASS - Fall Break
Oct 12             Utah RR method
Oct 19                                                    Project 4
Oct 26             Disease Association and Syndromes
Nov 2
Nov 9              Utah High Risk Pedigree Methods        Project 5
Nov 16             NO CLASS AMIA/IGES
Nov 23             NO CLASS - Thanksgiving
Nov 30
Dec 7              Presentations

Hand in projects 1 and 2 by Sep 21
Hand in projects 3 - 5 the week after they are assigned
   Project Presentations

disease phenotype, familial/genetic evidence
GIF’s
RR’s
disease associations
high risk pedigrees found
problems
lessons

								
To top