3-Peter Devilee Session I - 12.02.2019-My genome Our...
Transcript of 3-Peter Devilee Session I - 12.02.2019-My genome Our...
Peter Devilee
Predicting breast cancer – an update
Human Genetics & Pathology
Leiden Univ Medical Centre
LEIDEN, NL
Disclosure
15-feb-192 P. Devilee (LUMC)
(potentiële) belangenverstrengeling Geen
Voor bijeenkomst mogelijk relevante
relaties met bedrijven(Bedrijfsnamen)
• Sponsoring of onderzoeksgeld
• Honorarium of andere (financiële)
vergoeding
• Aandeelhouder
• Andere relatie, namelijk …
• NWO, KWF, Horizon2020
•
••
Breast cancer in the Netherlands
1 in 8 (lifetime risk)
.
15-feb-193
van der Waal D, Verbeek AL, den Heeten GJ, Ripping TM, Tjan-Heijnen VC, Broeders MJ. Breast cancer diagnosis and death in
the Netherlands: a changing burden. Eur J Public Health 2015;25(2):320-324
P. Devilee (LUMC)
Clinical management of breast cancer risk
15-feb-194 P. Devilee (LUMC)
Risk class Low (<2) Moderate
(RR: 2-3)
High
(RR: >3)
BRCA1/2
Life time risk <20% 20-30% >30% >60%
Start prevention 50 yr 40 yr 35 yr 25 yr
Physical
examination- - + +
Mammography Population
screening
<50 yr Annually
>50 yr
population
screening
<60 yr Annually
>60 yr
population
screening
>40 yr
biannually
>60 yr
annually
MRI - - - +
Dutch guideline
Main risk factors for breast cancer
• Biology
• Age
• Hormonal factors
• Breast density on mammogram
• Lifestyle
• Reproductive factors
• Alcohol / smoking / physical exercise / obesity
• Use of HRT, radiation exposure
• Family history
• Familial relative risk (FRR)
15-feb-195 P. Devilee (LUMC)
Predicting breast cancer – current practice
15-feb-196 P. Devilee (LUMC)
B44
B56
Gene-test
Gene-specific counseling
Family history
“positive”
“negative”
Breast cancer genes: N = ?
15-feb-197
Proven:
ATM, BRCA1, BRCA2, CHEK2, PALB2
Syndromic:
PTEN, STK11/LKB1, TP53, NF1
Possible:
BRIP1, NBN, FANCM, MRE11A, RECQL,
MUTYH, PPM1D, RAD51C, RAD51D,
XRCC2, RINT1, MMR genes, . . .
Challenges:
• Which are truly associated?
• Risk estimates uncertain
• Subtype-specific risks
• Allelic diversity (missense changes)
• Risks to other cancers?
P. Devilee (LUMC)
Gene-specific risks
15-feb-198 P. Devilee (LUMC)
High risk
Moderate risk
Near-PopulationRisk
Lee et al. (2016); Lee et al. (2019)
0
10
20
30
40
50
60
70
80
20 30 40 50 60 70 80
Bre
ast
Ca
nce
r R
isk
(%
)
Age (years)
BRCA1 BRCA2 PALB2 CHEK2 ATM Untested
Confidence intervals . . .
15-feb-19P. Devilee (LUMC)9
Easton et al. 2015
Low
Moderate
High
Assumptions:
• Applying estimates of relative risk to
population incidence (England 2003-2007)
• Ignoring competing mortality
• Ignoring other risk factors as modifiers
• “Typical” mutation carrier, i.e., ignoring
allelic diversity
Assumptions:
• Applying estimates of relative risk to
population incidence (England 2003-2007)
• Ignoring competing mortality
• Ignoring other risk factors as modifiers
• “Typical” mutation carrier, i.e., ignoring
allelic diversity
2x lifetime risk
4x lifetime risk
PALB2
CHEK2
DNA Sequencing, big data
15-feb-19P. Devilee (LUMC)10
Study Target Cases Controls
BRIDGES Gene Panel N=35 60,000 60,000
CARRIERS Gene Panel N=30 30,000 30,000
Li et al. 2018 Gene Panel N=625 11,400 4,000
Couch et al. 2017 Gene Panel N=21 30,000 GnomAD
Large-scale case-control studies
BRIDGES – preliminary results major genes
15-feb-1911 P. Devilee (LUMC)
BRIDGES
(Pop-based)
Meta-analysis[1] Ambry[2]
Gene OR (95%CI) OR (90%CI) OR (95% CI)
BRCA1 11.60 (8.54-15.78) 11.4 -
BRCA2 5.63 (4.58-6.91) 11.7 -
PALB2 5.32 (3.79-7.47) 5.3 (3.0–9.4) 7.46 (5.12-11.19)
ATM 1.90 (1.52-2.39) 2.8 (2.2–3.7) 2.78 (2.22-3.62)
CHEK2 2.45 (2.10-2.86) 3.0 (2.6–3.5) 2.26 (1.89-2.72)
[1] Easton et al. (2015) N. Engl. J. Med. 372, 2243; [2] Couch et al. (2017) JAMA Oncol. 3, 1190
Polygenic risk scores
15-feb-1912 P. Devilee (LUMC)
77 SNPs – max 154 risk alleles
Mavaddat et al, J Natl Cancer Inst. 2015 Apr 8;107(5)
(N = 33,381)
(N = 33,673)
Relative risk
RR = 1
Common SNPs: united we stand, divided we fall
Risk stratification: 313-SNP PRS
15-feb-1913 P. Devilee (LUMC)
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
25 30 35 40 45 50 55 60 65 70 75
Life
time
Abs
olut
e R
isk
Age (years)
>99%
95-99%
90-95%
80-90%
60-80%
40-60%
20-40%
10-20%
5-10%
1-5%
<1%
Overall breast cancer – Test set
Mavaddat et al, Am. J. Hum. Genet., 2019
Middle quintile
Interaction of BRCA1/2 and PRS
15-feb-1914 P. Devilee (LUMC)
BRCA1(N=15252)
BRCA2(N=8211)
Breast cancer Ovarian cancer
CIMBA Data – Kuchenbaecker et al. (2017) J Natl Cancer Inst. 109(7): djw302
N=7797 events N=2462 events
N=4330 events N=631 events
Impact on risk management?
15-feb-1915 P. Devilee (LUMC)
• BRCA1• Breast cancer• 88 SNPs
12 jr
Risk-prediction tools: BOADICEA
• Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation
Algorithm
• Online tool
• Computes chance of carrying BRCA1 and BRCA2 mutation
• Computes age-specific lifetime risks to breast and ovarian cancer
• Uses family history and BRCA1, BRCA2, ATM, CHEK2, PALB2 prevalences and
risks
• Includes polygenic risk score
• Includes non-genetic risk factors in simple additive risk model
• CE-mark procedure ongoing
15-feb-1916 P. Devilee (LUMC)
Lee, Antoniou, et al. Genet Med, 2019
BOADICEA – Family history only
15-feb-1917 P. Devilee (LUMC)
2 2 2
B32
d34
B42 l
B45 r
d51
Age: 35
Pr64
25% 25% 19% 24% 19%
Family history only
BOADICEA – Family history + PRS
15-feb-1918 P. Devilee (LUMC)
2 2 2
B32
d34
B42 l
B45 r
d51
Age: 35
Pr64
31% 19% 12% 28% 22%
Family history + PRS
~20% may shift risk class
15-feb-1919 P. Devilee (LUMC)
No change
~100 BRCAx multiple case families
N=340
N=272
Lakeman et al., submitted
Moderate-risk gene carrier (CHEK2*1100delC)
15-feb-1920 P. Devilee (LUMC)
35.3% 44.5% 20.2%
Lee et al., Genet. Med., 2019
Stop doing just panel testing!
15-feb-1921 P. Devilee (LUMC)
0.1 1 10
CHEK2ATM
PALB2BRCA2
BRCA1
Approx. Centile 10% <0.1%1%
Relative Risk
Practice now:
• Aimed at detecting and excluding
high risk
• Testing involves mostly affected
women
• In positive families:
• Carriers receive gene-specific risk
estimates
• Non-carriers receive population
risks
• In negative families: family history
determines risk
Future:
• Aimed at assessing individual risk
• Anyone can be tested
• No “positive” or “negative” results
• Risks can be lower than average
• Non-genetic risk factors
incorporated
• Some risk factors are modifiable
15-feb-1922
Future of genetic counseling
P. Devilee (LUMC)
Risk-based population screening?
15-feb-1923 P. Devilee (LUMC)
0.00
0.02
0.04
0.06
0.08
0.10
0.12
20 25 30 35 40 45 50 55 60 65
10 y
ear
risk
Age (years)
All Breast Cancers
>99%
95-99%
90-95%
80-90%
60-80%
40-60%
20-40%
10-20%
5-10%
1-5%
<1%
20% of women reach threshold before age 40
20-40% of women never reach threshold
Screening threshold
10-year risk all breast cancers, by SNP profile
Mavaddat et al (2015), JNCI 107, djv036
Acknowledgements
15-feb-1924 P. Devilee (LUMC)
• Breast Cancer Association Consortium (BCAC)
• Consortium of Investigators of Modifiers of BRCA1/2 (CIMBA)
• Breast Cancer Risk After Diagnostic Gene Sequencing (BRIDGES)
• PERSPECTIVE, CARRIERS, kConFAB, B-CAST, ENIGMA
• Hebon NL