Familial associations and disease risk steinar tretli

Familial associations and diseaserisk.

Linkage possibilitiesBy

Steinar Tretli, PhD, Professor

The Cancer Registry of Norway

Cancer are closely connected to both genetic conditions and environmental exposures.

• We observed for several types of cancer a familial association /clustering.

• This may be caused by genetic, epigenetic or environmental (incl lifestyle) effects.

• Relatives to a cancer case are usually concerned about their own/family cancer risk.

Familial associations?

• For several familial associations, we know the genetic condition behind.This is the case for breast cancer among women with a AT-mutation («Ataxia telengiasia») or

• elevated risk of breast cancer risk in carriers of BRCA1 mutation (indirectly detected by breast cancer in early age)

• We usually need large studies to locate and characterize the familiar component

• Familial associations in disease incidence may give important indications of the cause of the disease.

• Familial associations are relevant in the discussion of genetic effects and may be a basis for genetic councelling.

• Familial associations may be relevant in the discussion of environmental effects and a basis for envirionmental intervention.

Environmental and heritable causes of cancer among 9.6 million individuals in the Swedish family‐cancer database.

International Journal of CancerVolume 99, Issue 2, pages 260-266, 8 MAR 2002 DOI: 10.1002/ijc.10332http://onlinelibrary.wiley.com/doi/10.1002/ijc.10332/full#fig1

Familiality = genetic factors + shared environmental factors + childhood shared environmental factors

Environmental and heritable causes of cancer among 9.6 million individuals in the Swedish family‐cancer database.

International Journal of CancerVolume 99, Issue 2, pages 260-266, 8 MAR 2002 DOI: 10.1002/ijc.10332http://onlinelibrary.wiley.com/doi/10.1002/ijc.10332/full#fig1

Familiality = genetic factors + shared environmental factors + childhood environmental factors

Environmental and heritable causes of cancer among 9.6 million individuals in the Swedish family‐cancer database.

International Journal of CancerVolume 99, Issue 2, pages 260-266, 8 MAR 2002 DOI: 10.1002/ijc.10332http://onlinelibrary.wiley.com/doi/10.1002/ijc.10332/full#fig1

Familiality = genetic factors + shared environmental factors + childhood environmental factors

Int J Cancer 99,260-266 (2002)

Relation n, both Correlation (95% CI) n, both Correlation (95% CI) n, both Correlation (95% CI)affected affected affected

Stomach Colon Rectum

Mother-father 203 0.15 (0.13–0.17) 422 0.12 (0.11–0.14) 122 0.09 (0.07–0.11)Daughter-daughter 1 0.29 (0.21–0.45) 4 0.22 (0.14–0.31) 0 Son-son 4 0.41 (0.37–0.52) 5 0.28 (0.16–0.39) 2 0.30 (0.09–0.43)Daughter-son 0 9 0.25 (0.18–0.33) 1 0.17 (0.07–0.29) Mother-daughter 5 0.12 (0.04–0.15) 73 0.20 (0.19–0.23) 12 0.15 (0.10–0.21) Mother-son 7 0.14 (0.04–0.21) 60 0.20 (0.17–0.24) 15 0.17 (0.11–0.22) Father-daughter 13 0.16 (0.09–0.21) 58 0.17 (0.13–0.20) 20 0.17 (0.12–0.22) Father-son 16 0.16 (0.10–0.22) 50 0.17 (0.13–0.21) 14 0.12 (0.06–0.18) Half sibs 0 2 0.18 (0.11–0.31) 0Adoptives 0 3 0.09 (-0.05–0.21) 0

Kamila CZENE, Paul LICHTENSTEIN and Kari HEMMINKI:

ENVIRONMENTAL AND HERITABLE CAUSES OF CANCER AMONG 9.6 MILLION INDIVIDUALS IN THE SWEDISH FAMILY-CANCER DATABASE

This means that we need large populations to locate and caracterize the effect.

Example

Nordic study of nervous system tumours (CNS).Among 63307 CNS cases 32 347 belonged to the offspring generation

Brain (Concordant family history):

SIR= 1.7 (1.5,1.8) in offspring of affected patients (n=353)SIR= 2.0 (1.7,2.3) in siblings (n=201)SIR= 9.4 (5.5,15.1) parent and sibling affected (n= 17) Spine (Concordant family history):

SIR= 4.4 (1.1,11.3) in offspring of affected patients (n=4)SIR= 7.1 (1.8,18.4) in siblings (n=4)Peripheral nerves (Concordant family history):

SIR= 16.3 (9.3,26.6) in offspring of affected patients (n=16)SIR= 27.7 (16.1,44.4) in siblings (n= 17)SIR= 943.9 ( 245.5, 2440.6) parent and sibling affected (n=4)

Challenges in cancer studies:

Old enough registries: Parents- children (most types of cancer occur in old ages.)

Few offspring per family: Difficult to study genetic factors and shared environmental factors

The personal identification nummer (PIN) given to all Norwegian citizen ( and the same in the other nordic countries), has made it possible to build up a generation register. The PIN for every child is linked to the mother’s PIN and also father’s PIN (Father defind by information from the mother). Maternity and patternity histories are created by Statistics Norway (SSB) from 1935 and onwards and the registry is own by Skattedirektoratet. From 1967 the same information can be find in the Norwegian birth registry.

Several outcome registries in Norway may be linked to the Norwegian population registry(own by Skattedirektoratet):• The Cancer Registry (1953-)• The Medical Birth Registry (1967-)• The Patient Registry (2008-)• The Description Registry ( 2004-)• The Cause of Death Registry (1951-)

Affected family member(s)

Number of families

• No affected family members

• Father only First son only

• Second son only Third son only

• Fourt son only• Fifth son only• Father & first son• Father & sec. sons• Father & First &

second sons• First & second sons• First & third sons• First & fourth sons• First & fifth sons• Second & third sons• Second & fourth sons• Second & fifth sons• Affected

family members

1,127,796

2,2683,487 1,272

32865181910

233

711431

7,524

Familial testicular germ-cell tumors in Norway

The Relative risk given:• affected father 4.03 (95% CI: 3.12, 5.19)• One affected brother 5.88 (95% CI: 4.70,7,36)• Two affected brothers 21.71 (95% CI: 8.93, 52.76)

Example 2. (Valberg et al, Am J Epidemiol, 2014)

Example 3

Association study on birth defects and cancer incidence.Result: Birth defects are associated to increased cancer risk.The cancer risk is not elevated in sisters, brothers or parents of children with birth defects. (Show table)

Thank you for your attention!

[email protected]