valoración de riesgo y medidas de prevención en cáncer ......#seom2019 valoración de riesgo y...

#SEOM2019

Valoración de riesgo y medidas de prevención en cáncer hereditario

SESIÓN FORMATIVA DE CÁNCER HEREDITARIO IIAVANCES EN EL MANEJO CLÍNICO DEL CÁNCER HEREDITARIO

Begoña Graña Suárez MD PhDOncología Médica- Consulta Cáncer Familiar Complexo Hospitalario Universitario A Coruña [email protected]

23 de octubre de 2019

#SEOM2019

Disclosure Information

Research Funding (clinical trials): Incyte, Amgen, Celgene, Sanofi

Travel/Conference assistance Grant support: Amgen, Astra-Zeneca, Novartis , Celgene ,Servier,

NINGÚN CONFLICTO DE INTERÉS RELACIONADO CON ESTA PRESENTACIÓN

3

TYPICAL REFERRALS TO A CANCER GENETICS CLINICTYPICAL REFERRALS TO A CANCER GENETICS CLINIC

Colorectal

Rare syndromes

Breast/ovary

Other

Marc Tischkowitz MD PhD, ESMO – Hereditary Cancer Genetics 26-27 April 2019 Lugano

Relatives

Cancer patients

High riskDirect gene testing feasible

Moderate risk

4

Pepita 49 años, Cáncer de MAMA

María 25 años, sana

¿ Cuál es mi riesgo de padecer cáncer de mama ?¿Cuál es mi riesgo de ser portadora de mutación en un gen de predisposición?

5

Dr. Luis Robles Díaz #SEOM18

Historia familiar

Evaluación

Recomendaciones individualizadas

Valoración en una unidad de consejo genético especializada

Recomendaciones de lapoblación general

Intervención

Espórádico

Moderado(Familiar)

Alto(Hereditario)

Clasificacion enfunción del riesgo

¿PARA QUE SIRVE REALIZAR UNA ESTIMACIÓN DE RIESGO EN CÁNCER FAMILIAR ?

7Foulkes , NEJM 2008

Susceptibilidad hereditaria al cáncer de mama

ALTO RIESGO: RR > 4 ( RA >30%)RIESGO MODERADO: RR 2-4 (>17 % < 30%)BAJO RIESGO: RR 1.01-1.99 (<17%)RR=riesgo relativoRA=riesgo acumulado a lo largo de la vida

8

NICE 2018 https://www.nice.org.uk/guidance/cg164/

1 Lifetime risk of developing breast cancer is at least 17% but less than 30%.2 Lifetime risk of developing breast cancer is at least 30%. High risk group includes rare conditions that

carry an increased risk of breast cancer, such as Peutz-Jegher syndrome, (STK11), Cowden (PTEN), familial

diffuse gastric cancer (E-Cadherin).3 Surveillance recommendations for this group rePect the fact that women who at Orst assessment had a

30% or greater BRCA carrier probability and reach 60 years of age without developing breast or ovarian

cancer will now have a lower than 30% carrier probability and should no longer be offered MRI

surveillance.4 Surveillance recommendations for this group rePect the fact that women who at Orst assessment had a

30% or greater TP53 carrier probability and reach 50 years of age without developing breast cancer or any

other TP53-related malignancy will now have a lower than 30% carrier probability and should no longer be

offered MRI surveillance.

Terms used in this guideline

Breast cancer risk categoryBreast cancer risk category

Breast cancer risk categoryBreast cancer risk category

Near populationNear population

riskrisk

ModerModerate riskate risk High risk1High risk1

Lifetime risk from age 20Lifetime risk from age 20 Less than 17% Greater than 17% but less than

30%

30% or

greater

Risk between ages 40Risk between ages 40

and 50and 50

Less than 3% 3–8% Greater than

8%

1This group includes known BRCA1, BRCA2 and TP53 mutations and rare conditions that carry

an increased risk of breast cancer such as Peutz-Jegher syndrome (STK11), Cowden (PTEN)

and familial diffuse gastric cancer (E-Cadherin).

First-degree relativFirst-degree relativeses

Mother, father, daughter, son, sister, brother.

Second-degree relativSecond-degree relativeses

Grandparent, grandchild, aunt, uncle, niece, nephew, half-sister, half-brother.

Familial breast cancer: classi5cation, care and managing breast cancer and related risks in people with

a family history of breast cancer (CG164)

© NICE 2018. All rights reserved. Subject to Notice of rights (https://www.nice.org.uk/terms-and-

conditions#notice-of-rights).

Page 37 of

48

High risk: Annual MRI from age 25 + Mx from age 30/consideration of RRMModerate risk: Annual Mx from age 40 until age 60Population risk: Population screening

9

8

Professional barriersTeamwork by Oncologists and Clinical Geneticists

1. Interpretation of BRCA-test result (tumor vs germline & mutation vs VUS)

2. Interpretation of Family history for

• Other hereditary causes than BRCA• Familial cancer risk in non-BRCA families

3. Communication of test result and of familial cancer risk

After BRCA-germline testing face-to-face clinical geneticist in case of:

- positive BRCA-test result

- positive family history (Other hereditary or familial causes of cancer

- Complex test results (VUS, germline vs tumour, multiple genes)

Personal communication from presenterNicoline Hoogerbrugge, ESMO Preceptorship on Hereditary Cancer Genetics Lugano April 2019

10

En el 2019, al menos para BRCA1/2tenemos claros los riesgo asociados de padecer cáncer…. ¿?

WONG , Genomic Medicine 2016

11

Kuchenbaecker , JAMA 2017

Risks of Breast, Ovarian, and Contralateral Breast Cancer for BRCA1 and BRCA2 Mutation Carriers.

Cohorte prospectiva ( inclusión entre 1997-2011)6036 BRCA1 y 3820 BRCA2 5046 Mujeres sanas4810 afectas de cáncer de mama/ovario u ambos.Mediana de seguimiento 5 años

cancer at 20 years after the first breast cancer diagnosis was41% (95% CI, 32%-53%) for BRCA1 and 21% (95% CI, 15%-50%) for BRCA2 carriers.

Breast and Ovarian Cancer Risks by Family HistoryThe estimated cumulative breast and ovarian cancer risks byfamily history are shown in Table 4 and eFigure 3 in the Supple-ment. Breast cancer risk estimates for both BRCA1 and BRCA2carriers increased with the number of first- and second-degree relatives diagnosed as having breast cancer (P<.001 fortrend for BRCA1; P=.02 for BRCA2) (Table 4). For women with2 or more first- or second-degree relatives diagnosed as hav-ing breast cancer compared with those with no family historyof breast cancer, the HR for breast cancer was 1.99 (95% CI, 1.41-2.82) for BRCA1 carriers (cumulative risk estimates to age 70years: 73% [95% CI, 65%-80%] vs 53% [95% CI, 39%-69%]) andthe HR for breast cancer was 1.91 (95% CI, 1.08-3.37) for BRCA2carriers (cumulative risks to age 70 years: 65% [95% CI, 56%-74%] vs 39% [95% CI, 25%-56%]) (Table 4).

There was no significant difference in ovarian cancerrisk for BRCA1 carriers with family history of ovarian cancercompared with those without (HR, 1.37; 95% CI, 0.89-2.11)(Table 4; eFigure 3 in the Supplement). A similar patternwas observed for BRCA2 carriers, but the number of eventsfor women with ovarian cancer family history was small(n = 5). Results were similar when family history of cancerwas restricted to first-degree relatives (eTable 9 in theSupplement) or when analyses were stratified by the pres-ence of family history of breast or ovarian cancer (eTables10-13 in the Supplement). For BRCA1 mutation carriers, therisk of breast cancer was lower for women with a family his-tory of ovarian cancer compared with those with no familyhistory of ovarian cancer (HR, 0.71 [95% CI, 0.51-0.99] in

women with a family history of breast cancer; HR, 0.38[95% CI, 0.21-0.70] in those without) (eTable 12).

Breast and Ovarian Cancer Risks by Mutation PositionBRCA1 mutations located outside the region bounded bypositions c.2282 to c.4071 were associated with a signifi-cantly higher breast cancer risk compared with mutationswithin the region (HR, 1.46; 95% CI, 1.11-1.93; P = .007)(Table 5; eFigure 4 in the Supplement), but there was no sig-nificant difference in ovarian cancer risk. There was nosignificant difference in the breast or ovarian cancer risks foreither the BRCA1 c.68_69delAG or c.5266dupC mutationscompared with BRCA1 mutations in the same region(Table 5). BRCA2 mutations outside the OCCR were associ-ated with a significantly higher breast cancer risk comparedwith mutations within the OCCR (based on the narrowOCCR definition: HR, 1.70 [95% CI, 1.18-2.46]; P = .005; basedon the broad OCCR definition: HR, 1.93 [95% CI, 1.36-2.74];P < .001) (Table 5), but there was no significant difference inovarian cancer risk. There was no significant differencein breast cancer risk for BRCA2 c.5946delT mutation carrierscompared with other OCCR BRCA2 mutations (HR, 0.73;95% CI, 0.35-1.54; P = .41). The associations by mutationposition remained significant after adjusting for familyhistory of breast cancer and after excluding carriers of theBRCA2 c.5946delT mutation from the OCCR (eTable 14 inthe Supplement).

DiscussionThis study estimated age-specific risks of breast, ovarian, andcontralateral breast cancer for BRCA1 and BRCA2 mutation

Figure 2. Estimated Cumulative Risks of Breast and Ovarian Cancer in Mutation Carriers

100

80

60

40

20

020

5330

80

1321

70

4135

60

138110

50

273204

40

404267

Brea

st C

ance

r Risk

, %

Age, y

No. at riskBRCA1

BRCA1 carriersBRCA1 carriers

BRCA2 carriers

BRCA2 carriers

BRCA2

30

340160

Cumulative risk of first breast cancer among BRCA1 and BRCA2mutation carriers

A

100

80

60

40

20

020

5330

80

2328

70

5459

60

131157

50

243230

40

544371

Ovar

ian

Canc

er R

isk, %

Age, y30

420190

Cumulative risk of ovarian cancer among BRCA1 and BRCA2mutation carriers

B

Kaplan-Meier estimates of cumulative risks of breast and ovarian cancers.In the breast cancer analysis, women were censored at risk-reducing bilateralmastectomy. In the ovarian cancer analysis, women were censored forrisk-reducing salpingo-oophorectomy. Number at risk indicates the number

of women who remained at risk at the end of the 10-year age category(eg, in panel A, there were 138 women with BRCA1 mutations still at risk ofbreast cancer at the end of the age 50-60 years period). The earliest follow-upstarted at age 18 years.

Research Original Investigation Risks of Breast, Ovarian, and Contralateral Breast Cancer Among BRCA Mutation Carriers

2408 JAMA June 20, 2017 Volume 317, Number 23 (Reprinted) jama.com

© 2017 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a SERGAS - Servicio Gallego de Salud User on 10/20/2019

12

LIFETIME CANCER RISKSKUCHENBAECKER, 2017

Family history & mutation position – important variables in risk assessment.

Kuchenbaecker JAMA 2017

13

BRCA1 GENOTYPE-PHENOTYPERebbeck, JAMA 2015

RelativelyHigher BC risk

RelativelyHigher OC risk

Rebbeck (CIMBA Consortium) , JAMA 2015 JAMA

AssociaWon of type and locaWon of BRCA1 and BRCA2 mutaWons with risk of breast and ovarian cancer.

Estudio observacional 19 581 BRCA1+ y 11 900 BRCA2+ 33 países 6 continentes à“With appropriate validation, thesedata may have implications for riskassessment and cancer prevention “

https://www.ncbi.nlm.nih.gov/pubmed/25849179

14

Mujeres R71G+ afectas/total portadoras

Mediana edad diagnóstico Histología Estadio TNM

Ca. MAMA 58% 42 (19-87) 75,6%¹ TripleNegativo 53% ² estadio II

2º Ca. MAMA

22,5 % de las portadoras que

padecieron un 1er

Ca. Mama

53 (32-70)81,8%¹ Triple

Negativo 55%² estadio I

Ca. OVARIO 16,3% 50 (39-82) 74%¹ Seroso 60%² estadio III/IV

MUTACIÓN FUNDADORA GALLEGA c.211A>G (BRCA1): ELEVADA FRECUENCIA Y AGRESIVIDAD Graña B SEOM 2019

-Cohorte RETROSPECTIVA -245 mujeres portadoras c.211 A>G(H.Coruña, ICO, VHV)

15

Modelos/Herramientas para la estimación de riesgo de padecer cáncer *

Mama/ovario ✓ BRCRAT (Gail)

✓ Tyrer-Cuzick ( Ibis) ✓ BOADICEA

✓ BRCAPRO

Cancer colorectal✓ MMPRO✓ PREMMOtros✓ MELPRO✓ Cleveland Score

*Diferenciar de modelos que exclusivamente estiman el riesgo de ser portador de una mutación : Manchester, Penn II...

Esmmación del riesgo de padecer cáncer de mama1 Breast Cancer Risk AssessmentTool(BCRAT) basado en el modelo de Gail (NCI- EEUU) ,

MUJER, 42 años, -menarquia 13, -1er hijo a los 33 - no biopsias, -1 familiar de primer gradoafecto de CM

Gail, JNCI 1989

1. Breast Cancer Risk AssessmentTool(BCRAT) basado en el modelo de Gail (NCI- EEUU) ,

MAMA, 44

MAMA, 38 MAMA, 29OVARIO, 42

El modelo de GAIL (BCRAT) infraesmma el riesgo de padecercáncer de mama al no considerar los antecedentes en linea paterna

MAMA, 60

Estimación del riesgo de padecer cáncer de mama

Modelos para la estimación del riesgo en cáncer de mama :

2. IBIS Breast Cancer Risk Evaluation Tool, basado en el modelo de Tyrer-Cuzick (UK)

Tyrer, Stat Med 2004

2. Modelos para la estimación del riesgo en cáncer de mama :

IBIS Breast Cancer Risk Evaluation Tool, basado en el modelo de Tyrer-Cuzick (UK)

Modelos para la esmmación del riesgo en cáncer de mama :

2. IBIS Breast Cancer Risk Evaluamon Tool, basado en el modelo de Tyrer-Cuzick (UK)

23

BOADICEA Pedigree Table View

Your input pedigree is listed below...

Name Tgt IndivID FathID MothID Sex MZtwin Status Age Yob 1BrCa 2BrCa OvCa ProCa PanCa Ashkn GeneticTests Pathology1 PROBAND < 1 3 2 F Alive 33 1986 2 Anon 2 5 4 F Dead 36 1955 36 3 Anon 3 9 8 M Alive 60 4 Anon 4 F Dead 34 1925 34 5 Anon 5 M Dead 6 Anon 6 5 4 F Alive 55 7 Anon 7 5 4 F Alive 50 8 Anon 8 F Dead 64 9 Anon 9 M Dead 75

Model: UK Mutation Frequencies / Default Mutation Sensitivities / UK Cancer Incidence Rates/ Percent Format

Page Up Page Down Edit Add Delete MZ Twin

Logout Reset Model Draw Switch Compute

Centre for Cancer Genetic Epidemiology

© 2016 Centre for Cancer Genetic EpidemiologyDepartment of Public Health and Primary Care, University of Cambridge


3. BOADICEA Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm

Antoniou BJC 2008

24



Mother of Consultand Mother of ConsultandEnter details of this individual...

Logout Reset Go Back Skip Continue

Centre for Cancer GeneticEpidemiology


Estrogen Receptor (ER) Unknown Positive Negative

Progestrogen Receptor(PR)

Unknown Positive Negative

Human Epidermal GrowthFactor Receptor Two (HER2)


Cytokeratin Fourteen(CK14)


Cytokeratin Five/Six(CK5/6)


Clinical history Breast cancer pathology Genetic testing

21/10/2019, 03)35BOADICEA risk estimation on the World Wide Web

Page 1 of 1https://pluto.srl.cam.ac.uk/cgi-bin/bd4/v4beta14/bd.cgi

Mother of Consultand Mother of ConsultandEnter details of this individual...

Logout Reset Go Back Skip Continue

Centre for Cancer Genetic Epidemiology


BRCA1Genetic test type Untested Mutation search Direct gene test

Genetic test result Untested Positive Negative

BRCA2Genetic test type Untested Mutation search Direct gene test


PALB2Genetic test type Untested Mutation search Direct gene test


ATM Genetic test type Untested Mutation search Direct gene test


CHEK2Genetic test type Untested Mutation search Direct gene test


Clinical history Breast cancer pathology Genetic testing

25

Modelos para la esmmación del riesgo en cáncer de mama :

3. BOADICEA Breast and Ovarian Analysis of Disease Incidence and Carrier Esmmamon AlgorithmBOADICEA Computed Results

Computed results for the Target: Anon(1)

GeneticStatus

Mutation CarrierProbabilities (Percent)

BRCA1 6.0BRCA2 4.7PALB2 1.2ATM 0.9CHEK2 1.3NoMutation 85.9

ModelParameters

Target FamilyMember Anon(1)

MutationFrequencies:UKBRCA1:6.394d-4BRCA2:0.00102PALB2:0.000575ATM:0.001921CHEK2:0.002614

Mutation SearchSensitivities: DefaultBRCA1: 0.7BRCA2: 0.8PALB2: 0.9ATM: 0.9CHEK2: 1.0

CancerIncidenceRates

UK

Age Breast CancerRisks (Percent)

Ovarian CancerRisks (Percent)

34 0.3 0.035 0.7 0.036 1.0 0.137 1.4 0.138 1.9 0.140 2.8 0.243 4.6 0.445 5.9 0.550 9.7 0.955 13.4 1.460 17.0 2.065 20.5 2.770 23.8 3.375 26.4 3.980 28.6 4.6

Logout

Reset

Go Back Graph Breast Cancer Risks Graph Ovarian Cancer Risks

Reformat Generate Report


© 2016 Centre for Cancer Genetic Epidemiology

2.0 BOADICEA risk calculation results

Index or subject of the BOADICEA calculationFirstname/identifier of index: Anon Unique identifier of index: 1

Breast Cancer Risks

Ovarian Cancer Risks



Breast Cancer Risks


26



BOADICEA Risk Calculation SummarySoftware version: v4beta14 Date: Mon Oct 21 02:46:24 BST 2019 Session number: 4900c24ca264f2 Risk calculation number: 4 Pedigree number: CORUNA Firstname/identifier of index: Anon Unique identifier of index: 1

THIS REPORT IS PROVIDED FOR RESEARCH USE ONLY This report summarises mutation carrier probabilities and breast/ovarian cancer risks computed using the BOADICEA model. The followingsections describe the input pedigree data set and computed results.

1.0 Input pedigree

1.1 Input pedigree summary tableThe input pedigree is summarised in the table(s) that follow. The index (the subject of the BOADICEA calculation) has the label ’T’ in the Tgtcolumn.

NameTgtIndivIDFathIDMothIDSexMZtwinStatusAgeYob 1BrCa2BrCaOvCaProCaPanCaAshknGeneticTests Pathology 1Anon < 1 3 2 F Alive 33 1986 BRCA1–[ms]&nbspBRCA2–[ms] 2Anon 2 5 4 F Dead 36 195536 3Anon 3 9 8 M Dead 65 4Anon 4 F Dead 34 192534 5Anon 5 M Dead 65 6Anon 6 5 4 F Alive 55 1955 7Anon 7 5 4 F Alive 50 1955 8Anon 8 F Dead 64 9Anon 9 M Dead 65



The BOADICEA model predicts the following mutation carrier probabilities and breast/ovarian cancerrisks for this individual:

Genetic status Mutation carrier probabilities (Percent) BRCA1 1.9 BRCA2 1.0 PALB2 1.3 ATM 1.0 CHEK2 1.4 No mutation 93.3

Model parameters Family member Anon (1) Mutation frequencies UK

BRCA1: 6.394d-4BRCA2: 0.00102PALB2: 0.000575ATM: 0.001921CHEK2: 0.002614

Mutation search sensitivities DefaultBRCA1: 0.7BRCA2: 0.8PALB2: 0.9ATM: 0.9CHEK2: 1.0

Cancer incidence rates UK

AgeBreast cancer risks (Percent)

Ovarian cancerrisks (Percent)

34 0.2 0.0 35 0.4 0.0 36 0.6 0.0 37 0.8 0.0 38 1.1 0.1 40 1.7 0.1 43 2.9 0.2 45 3.8 0.2 50 6.7 0.4 55 9.8 0.7 60 12.9 1.0 65 16.0 1.3 70 19.0 1.7 75 21.5 2.1 80 23.6 2.5



Breast Cancer Risks




Breast Cancer Risks












34 0.2 0.0 35 0.4 0.0 36 0.6 0.0 37 0.8 0.0 38 1.1 0.1 40 1.7 0.1 43 2.9 0.2 45 3.8 0.2 50 6.7 0.4 55 9.8 0.7 60 12.9 1.0 65 16.0 1.3 70 19.0 1.7 75 21.5 2.1 80 23.6 2.5

29

Modelos para la es-mación del riesgo en cáncer de mama :

3. BOADICEA Breast and Ovarian Analysis of Disease Incidence and Carrier Es-ma-on Algorithm

BOADICEA Risk Calculation SummarySoftware version: v4beta14 Date: Mon Oct 21 02:46:24 BST 2019 Session number: 4900c24ca264f2 Risk calculation number: 4 Pedigree number: CORUNA Firstname/identifier of index: Anon Unique identifier of index: 1

THIS REPORT IS PROVIDED FOR RESEARCH USE ONLY This report summarises mutation carrier probabilities and breast/ovarian cancer risks computed using the BOADICEA model. The followingsections describe the input pedigree data set and computed results.

1.0 Input pedigree

1.1 Input pedigree summary tableThe input pedigree is summarised in the table(s) that follow. The index (the subject of the BOADICEA calculation) has the label ’T’ in the Tgtcolumn.

NameTgtIndivIDFathIDMothIDSexMZtwinStatusAgeYob 1BrCa2BrCaOvCaProCaPanCaAshknGeneticTests Pathology 1Anon < 1 3 2 F Alive 33 1986 BRCA1–[ms]&nbspBRCA2–[ms] 2Anon 2 5 4 F Dead 36 195536 3Anon 3 9 8 M Dead 65 4Anon 4 F Dead 34 192534 5Anon 5 M Dead 65 6Anon 6 5 4 F Alive 55 1955 7Anon 7 5 4 F Alive 50 1955 8Anon 8 F Dead 64 9Anon 9 M Dead 65











34 0.2 0.0 35 0.4 0.0 36 0.6 0.0 37 0.8 0.0 38 1.1 0.1 40 1.7 0.1 43 2.9 0.2 45 3.8 0.2 50 6.7 0.4 55 9.8 0.7 60 12.9 1.0 65 16.0 1.3 70 19.0 1.7 75 21.5 2.1 80 23.6 2.5



Breast Cancer Risks




Breast Cancer Risks












34 0.2 0.0 35 0.4 0.0 36 0.6 0.0 37 0.8 0.0 38 1.1 0.1 40 1.7 0.1 43 2.9 0.2 45 3.8 0.2 50 6.7 0.4 55 9.8 0.7 60 12.9 1.0 65 16.0 1.3 70 19.0 1.7 75 21.5 2.1 80 23.6 2.5

27



BOADICEA Model ParametersUse the menus below to change BOADICEA model parameters...

Warning: computed risks are critically dependent on these settings

Mutationfrequencies: UK BRCA1: 6.394d-4 PALB2: 0.000575

BRCA2: 0.00102 ATM: 0.001921

CHEK2: 0.002614

Mutationsearchsensitivities:

Default BRCA1: 0.7 PALB2: 0.9

BRCA2: 0.8 ATM: 0.9

CHEK2: 1.0

Cancerincidence rates: UK

Output datadisplay format: Percent

Update Model




4. CANCERGENE ,incluye modelo BRCAPRO, (Dallas, EEUU)

29

MODELO IBIS

RA à39%Riesgo próximos 10 añosà 5.4%

Si BRCA negativo (no informativo)RA à26% [20% si “eslovena”]

Riesgo próximos 10 añosà 1.9 %

MODELO BOADICEA

RA à28%(mama) 4.6% (ovario)Riesgo próximos 10 añosà 4%(mama)

Si BRCA negamvo (no informamvo)RA à23% Riesgo próximos 10 añosà 2.7 %

30

Comment

www.thelancet.com/oncology Vol 20 April 2019 463

Risk-assessment tools are used in routine clinical practice to identify women at increased risk of breast cancer and to inform counselling about lifestyle changes, genetic testing, screening timing or modality, and eligibility for risk-reducing drugs or surgery. In The Lancet Oncology, Mary Beth Terry and colleagues1 report a comparative validation of four models—the Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm model (BOADICEA), BRCAPRO, the Breast Cancer Risk Assessment Tool (BCRAT), and the International Breast Cancer Intervention Study model (IBIS)—used in clinical practice to provide absolute risk estimates for breast cancer on the basis of different sets of factors. Their work is an important contribution to the field in view of the need for robust, comparative assessments of risk models. Terry and colleagues assessed model calibration with data from a combination of family-based cohorts in Australia, Canada, and the USA—the Breast Cancer Prospective Family Study Cohort. In this study population, BOADICEA and IBIS were the best-performing models in terms of calibration and risk discrimination. Although the study population was large (15 732 women without breast cancer at baseline, 619 of whom developed invasive breast cancer within 10 years of follow-up), important subgroup analyses by country or by age and mutation status were limited by size. 5-year and 10-year risk estimates were well calibrated overall, but both models overpredicted risk in women in the highest risk quantile for breast cancer. This overprediction was small (eg, the predicted vs observed 10-year risk of breast cancer in BRCA-negative women was 7·1% vs 6·1% for BOADICEA and 7·5% vs 6·5% for IBIS). However, the highest risk quantile included both moderate-risk and high-risk women (ie, women with a 10-year risk ≥5% or a 5-year risk ≥2·5%), and thus might not fully reflect prediction accuracy in women at high risk of breast cancer. It is also important to assess model performance in several independent study

populations because both model calibration and risk discrimination are population dependent. Consistent with findings in the Breast Cancer Prospective Family Study Cohort,1 two other reports2,3 based on large prospective cohorts in the USA and the UK have shown overprediction of breast cancer risk in women at high risk, showing that further model improvements are needed. These studies mostly included non-Hispanic white women, similar to the make-up of the Breast Cancer Prospective Family Study Cohort. Other racial and ethnic groups have been traditionally understudied, but efforts by different groups are underway to address this important research gap.

Simpler models, such as BCRAT, are sometimes preferred over complex models because they are easier and faster to use. However, the consequence of this simplicity is lower risk discrimination at the population level and less accurate risk scores for individual women. Although differences in measures of risk discrimination, such as the concordance statistic used by Terry and colleagues,1 might be small, more comprehensive information about risk factors could substantially improve the ability to identify women at high or low extremes of risk. For instance, a personal history of atypical hyperplasia, lobular carcinoma in situ, or high mammographic breast density can place women in the high-risk category, but these risk factors were not assessed by Terry and colleagues because of data limitations. Polygenic risk scores, which are derived from genetic testing of many common genetic variants, are a new, important risk factor for breast cancer. Although the variants are associated with small risks individually, when aggregated as a polygenic risk score, they can identify women with or without a family history of breast cancer at substantially different levels of risk.2,4–6 Both IBIS and BOACIDEA have been extended to include information on polygenic risk (although this information was not included in the versions analysed by Terry and colleagues),5,7 and clinical tests to measure

7 Giuliani M, Mathew AS, Bahig H, et al. SUNSET: stereotactic radiation for ultracentral non-small-cell lung cancer-a safety and efficacy trial. Clin Lung Cancer 2018; 19: e529–32.

8 Popp I, Grosu AL, Niedermann G, Duda DG. Immune modulation by hypofractionated stereotactic radiation therapy: therapeutic implications. Radiother Oncol 2016; 120: 185–94.

9 Nyman J, Hallqvist A, Lund JA, et al. OC-0565: SPACE—a randomized study of SBRT vs conventional fractionated radiotherapy in medically inoperable stage I NSCLC. Radiother Oncol 2016; 121: 1–8.

10 Chang JY, Senan S, Paul MA, et al. Stereotactic ablative radiotherapy versus lobectomy for operable stage I non-small-cell lung cancer: a pooled analysis of two randomised trials. Lancet Oncol 2015; 16: 630–37.

Assessment of breast cancer risk: which tools to use?

Published Online February 21, 2019 http://dx.doi.org/10.1016/ S1470-2045(19)30071-3

See Articles page 504

Stua

rt Ki

nlou

gh/Ik

on Im

ages

/ Scie

nce P

hoto

Libr

ary

Comment












Stua

rt Ki

nlou

gh/Ik

on Im

ages

/ Scie

nce P

hoto

Libr

ary

Comment












Stua

rt Ki

nlou

gh/Ik

on Im

ages

/ Scie

nce P

hoto

Libr

ary

Comment

464 www.thelancet.com/oncology Vol 20 April 2019

polygenic risk scores are already available. However, there is an urgent need for both rigorous assessment of the calibration of risk scores based on integration of polygenic risk scores and other known risk factors (such as family history, lifestyle, or benign breast disease), and training of health-care providers on how to interpret results.

Ideally, risk assessment tools should be flexible and easy to update as and when new data about risk emerge, and should be able to accommodate for missing data to predict risk on the basis of a subset of risk factors.6,8,9 Furthermore, the complexity and heterogeneity of disease might necessitate risk stratification by subtype.10 User-friendly interfaces for risk assessment and communication tailored to specific clinical scenarios could facilitate the use of complex underlying risk models and provide guidance to users. Such an approach would also address the challenge of choosing from the available models without a clear understanding of the pros and cons of each one.

In summary, application of clinical guidelines and progress towards new precision prevention strategies (eg, risk-stratified screening strategies that are being assessed in trials) requires the development of flexible, comprehensive models with robust validation in diverse populations to provide accurate personalised risk estimates, particularly in women at high risk of cancer, for whom clinical decisions have the greatest potential impact. Furthermore, when possible, models should be validated in the populations in which they

are intended to be used—eg, in health-care delivery settings.

*Montserrat Garcia-Closas, Nilanjan ChatterjeeDivision of Cancer Epidemiology and Genetics, National Cancer Institute, Shady Grove Campus, Rockville, MD 20850 (MG-C); and Bloomberg School of Public Health and Department of Oncology, School of Medicine, Johns Hopkins University, Baltimore, MD, USA (NC) [email protected] declare no competing interests.

1 Terry MB, Liao Y, Whittemore AS. 10-year performance of four models of breast cancer risk: a validation study. Lancet Oncology 2019; published online Feb 21. http://dx.doi.org/10.1016/S1470-2045(18)30902-1.

2 Choudhury PP, Wilcox AN, Brook MN, et al. Comparative validation of breast cancer risk prediction models and projections for future risk stratification. bioRxiv 2018; published online Oct 19. DOI:10.1101/440347.

3 Brentnall AR, Cuzick J, Buist DSM, Bowles EJA. Long-term accuracy of breast cancer risk assessment combining classic risk factors and breast density. JAMA Oncol 2018; 4: e180174.

4 Mavaddat N, Michailidou K, Dennis J, et al. Polygenic risk scores for prediction of breast cancer and breast cancer subtypes. Am J Hum Genet 2019; 104: 21–34.

5 Cuzick J, Brentnall AR, Segal C, et al. Impact of a panel of 88 single nucleotide polymorphisms on the risk of breast cancer in high-risk women: results from two randomized tamoxifen prevention trials. J Clin Oncol 2017; 35: 743–50.

6 Maas P, Barrdahl M, Joshi AD, et al. Breast cancer risk from modifiable and nonmodifiable risk factors among white women in the United States. JAMA Oncol 2016; 2: 1295.

7 Lee A, Mavaddat N, Wilcox AN, et al. BOADICEA: a comprehensive breast cancer risk prediction model incorporating genetic and nongenetic risk factors. Genet Med 2019; 31: 33.

8 Choudhury PP, Wilcox AN, Wheeler B, et al. iCARE: an R package to build and apply absolute risk models. bioRxiv 2016; published online Oct 12. DOI:10.1101/079954.

9 Grill S, Ankerst DP, Gail MH, Chatterjee N, Pfeiffer RM. Comparison of approaches for incorporating new information into existing risk prediction models. Stat Med 2017; 36: 1134–56.

10 Li K, Anderson G, Viallon V, et al. Risk prediction for estrogen receptor-specific breast cancers in two large prospective cohorts. Breast Cancer Res 2018; 20: 147.

Precision oncology giveth and precision oncology taketh away

Published Online March 8, 2019

http://dx.doi.org/10.1016/ S1470-2045(19)30095-6

Cent

re O

scar

Lam

bert/

Phan

ie/ S

cienc

e Pho

to Li

brar

y

Efforts to improve outcomes for cancer patients often centre on molecular profiling of individual tumours to identify actionable mutations and alterations that allow doctors to tailor treatment to the individual patient. Although it is preferable to find an inclusionary marker—a particular molecular vulnerability that sup-ports the use of a treatment that would otherwise not have been considered—exclusionary markers, mutations, or amplifications that preclude activity of an otherwise standard and seemingly reasonable treatment strategy, are also encountered. RAS mutations in colorectal cancer are a good example of an exclusionary marker, since the presence of a RAS

mutation excludes the realistic possibility of benefit from an anti-EGFR monoclonal antibody and therefore precludes the use of anti-EGFR therapies in practice, thus sparing the patient the toxicity, expense, and false hope of a treatment that doctors know in advance will not work. Although inclusionary markers are preferred, both types of markers can improve the quality of patient care.

In this issue of The Lancet Oncology, Funda Meric-Bernstam and colleagues,1 investigators of the MyPathway study, present updated data regarding the targeting of HER2-amplified colorectal cancer with a combination of two anti-HER2 drugs, pertuzumab and trastuzumab. The finding of anti-tumour activity

31


Articles

Lancet Oncol 2019; 20: 504–17

Published Online February 21, 2019

http://dx.doi.org/10.1016/ S1470-2045(18)30902-1

See Comment page 463

Department of Epidemiology (Prof M B Terry PhD, Y Liao MS,

N Leoce DrPH, N Zeinomar PhD) and Department of

Biostatistics (R Buchsbaum), Mailman School of Public

Health, Department of Pediatrics (Prof W K Chung MD),

and Department of Medicine (Prof W K Chung), Columbia

University, New York, NY, USA; Herbert Irving Comprehensive

Cancer Center, Columbia University Medical Center,

New York, NY, USA (Prof M B Terry, Prof W K Chung); Department of Health Research

and Policy (Prof A S Whittemore PhD), Department of Biomedical

Data Science (Prof A S Whittemore),

Department of Medicine (Prof E M John PhD),

and Stanford Cancer Institute (Prof E M John), Stanford

University School of Medicine, Stanford, CA, USA; Centre for

Epidemiology and Biostatistics (G S Dite PhD, R L Milne PhD,

Prof G G Giles PhD, P C Weideman GDip, Prof K-A Phillips MD, Prof J L Hopper PhD,

R J MacInnis PhD), Genetic Epidemiology Laboratory, Department of Pathology

(Prof M C Southey PhD), Department of Medicine

(S-A McLachlan MBBS), and Sir Peter MacCallum

Department of Oncology (Prof K-A Phillips), University of

Melbourne, Parkville, VIC, Australia; Lunenfeld-

Tanenbaum Research Institute, Sinai Health System, Toronto,

ON, Canada (Prof J A Knight PhD, G Glendon MSc,

Prof I L Andrulis PhD); Dalla Lana School of Public Health

(Prof J A Knight), Department of Molecular Genetics

10-year performance of four models of breast cancer risk: a validation studyMary Beth Terry, Yuyan Liao, Alice S Whittemore, Nicole Leoce, Richard Buchsbaum, Nur Zeinomar, Gillian S Dite, Wendy K Chung, Julia A Knight, Melissa C Southey, Roger L Milne, David Goldgar, Graham G Giles, Sue-Anne McLachlan, Michael L Friedlander, Prue C Weideman, Gord Glendon, Stephanie Nesci, Irene L Andrulis, Esther M John, Kelly-Anne Phillips, Mary B Daly, Saundra S Buys, John L Hopper, Robert J MacInnis

SummaryBackground Independent validation is essential to justify use of models of breast cancer risk prediction and inform decisions about prevention options and screening. Few independent validations had been done using cohorts for common breast cancer risk prediction models, and those that have been done had small sample sizes and short follow-up periods, and used earlier versions of the prediction tools. We aimed to validate the relative performance of four commonly used models of breast cancer risk and assess the effect of limited data input on each one’s performance.

Methods In this validation study, we used the Breast Cancer Prospective Family Study Cohort (ProF-SC), which includes 18 856 women from Australia, Canada, and the USA who did not have breast cancer at recruitment, between March 17, 1992, and June 29, 2011. We selected women from the cohort who were 20–70 years old and had no previous history of bilateral prophylactic mastectomy or ovarian cancer, at least 2 months of follow-up data, and information available about family history of breast cancer. We used this selected cohort to calculate 10-year risk scores and compare four models of breast cancer risk prediction: the Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm model (BOADICEA), BRCAPRO, the Breast Cancer Risk Assessment Tool (BCRAT), and the International Breast Cancer Intervention Study model (IBIS). We compared model calibration based on the ratio of the expected number of breast cancer cases to the observed number of breast cancer cases in the cohort, and on the basis of their discriminatory ability to separate those who will and will not have breast cancer diagnosed within 10 years as measured with the concordance statistic (C-statistic). We did subgroup analyses to compare the performance of the models at 10 years in BRCA1 or BRCA2 mutation carriers (ie, BRCA-positive women), tested non-carriers and untested participants (ie, BRCA-negative women), and participants younger than 50 years at recruitment. We also assessed the effect that limited data input (eg, restriction of the amount of family history and non-genetic information included) had on the models’ performance.

Findings After median follow-up of 11·1 years (IQR 6·0–14·4), 619 (4%) of 15 732 women selected from the ProF-SC cohort study were prospectively diagnosed with breast cancer after recruitment, of whom 519 (84%) had histologically confirmed disease. BOADICEA and IBIS were well calibrated in the overall validation cohort, whereas BRCAPRO and BCRAT underpredicted risk (ratio of expected cases to observed cases 1·05 [95% CI 0·97–1·14] for BOADICEA, 1·03 [0·96–1·12] for IBIS, 0·59 [0·55–0·64] for BRCAPRO, and 0·79 [0·73–0·85] for BRCAT). The estimated C-statistics for the complete validation cohort were 0·70 (95% CI 0·68–0·72) for BOADICEA, 0·71 (0·69–0·73) for IBIS, 0·68 (0·65–0·70) for BRCAPRO, and 0·60 (0·58–0·62) for BCRAT. In subgroup analyses by BRCA mutation status, the ratio of expected to observed cases for BRCA-negative women was 1·02 (95% CI 0·93–1·12) for BOADICEA, 1·00 (0·92–1·10) for IBIS, 0·53 (0·49–0·58) for BRCAPRO, and 0·97 (0·89–1·06) for BCRAT. For BRCA-positive participants, BOADICEA and IBIS were well calibrated, but BRCAPRO underpredicted risk (ratio of expected to observed cases 1·17 [95% CI 0·99–1·38] for BOADICEA, 1·14 [0·96–1·35] for IBIS, and 0·80 [0·68–0·95] for BRCAPRO). We noted similar patterns of calibration for women younger than 50 years at recruitment. Finally, BOADICEA and IBIS predictive scores were not appreciably affected by limiting input data to family history for first-degree and second-degree relatives.

Interpretation Our results suggest that models that include multigenerational family history, such as BOADICEA and IBIS, have better ability to predict breast cancer risk, even for women at average or below-average risk of breast cancer. Although BOADICEA and IBIS performed similarly, further improvements in the accuracy of predictions could be possible with hybrid models that incorporate the polygenic risk component of BOADICEA and the non-family-history risk factors included in IBIS.

Funding US National Institutes of Health, National Cancer Institute, Breast Cancer Research Foundation, Australian National Health and Medical Research Council, Victorian Health Promotion Foundation, Victorian Breast Cancer Research Consortium, Cancer Australia, National Breast Cancer Foundation, Queensland Cancer Fund, Cancer Councils of New South Wales, Victoria, Tasmania, and South Australia, and Cancer Foundation of Western Australia.

Copyright © 2019 Elsevier Ltd. All rights reserved.

Terry MB, Lancet Oncology 2019

ProF-SC http://www.bcfamilyregistry.org+18.856 mujeres (Australia, Canada, EEUU)Prospectivo, inclusión 1992-2011

Terry MB, Int J Epidemiology 2016

VALIDACIÓN de los 4 modelos BCRAT,IBIS, BOADICEA,BRCAPRO

+15732 mujeres [20-70 años, No ca de ovario ni cirugía mamaria reductora de riesgo]

*Mediana de seguimiento 11 años*519 casos de cáncer de mama

RIESGO DE CÁNCER DE MAMA A 10 AÑOS

*CALIBRACIÓN: casos esperados/observados

*DISCRIMINACIÓN:capacidad para separar afectos de sanos

http://www.bcfamilyregistry.org/

32


Articles


breast cancer in the complete validation cohort. The ratio of expected to observed cases was 1·05 (95% CI 0·97–1·14) for BOADICEA, 0·59 (0·55–0·64) for BRCAPRO, 0·79 (0·73–0·85) for BRCAT, and 1·03 (0·96–1·12) for IBIS (table 3). Figure 1 shows discriminatory power of the models in terms of relative sensitivity and specificity in the complete validation cohort, and subgroups by BRCA mutation status and age. The estimated sensitivity and specificity in the complete validation cohort were 81·9% (95% CI 78·6–84·9) and 42·8% (42·1–43·6) for BOADICEA, 45·9% (41·9–49·9) and 76·1% (75·4–76·8) for BRCAPRO, 57·8% (53·8–61·8) and 55·0% (54·2–55·8) for BCRAT, and 79·5% (76·1–82·6) and 46·7% (45·9–47·5) for IBIS.

The overall results did not change substantially when we did sensitivity analyses in which women diagnosed with in situ disease were not censored and were followed up until invasive diagnosis or last follow-up (appendix p 2), nor did results for IBIS when we considered different assumptions regarding competing mortality events (appendix p 3).

In BRCA-negative participants, the models’ discrimi-natory ability based on C-statistic estimates ranged from

0·62 (95% CI 0·59–0·64) to 0·66 (0·64–0·68; table 3). The assigned risks from the different models were moderately correlated with each other (appendix p 5), particularly for BRCA-negative participants, and model discrimination based on the C-statistic did not differ significantly different between BOADICEA, IBIS, and BCRAT for BRCA-negative women (appendix p 5). BOADICEA, BCRAT, and IBIS were well calibrated for BRCA-negative participants, but BRCAPRO substantially underpredicted breast cancer risk (table 3). Although BOADICEA and IBIS were well calibrated in both BRCA-positive and BRCA-negative participants, they overpredicted risk for women in the highest risk quantile (figures 2, 3). BCRAT underpredicted risk for women in the two lowest risk quantiles, whereas BRCAPRO underpredicted risk for all except for those in the highest risk quantile (all p<0·0001 for difference between predicted risk and observed risk; figures 2, 3). The country-specific calibration analysis showed that BOADICEA, but not the other three models, was well calibrated for participants from all three countries (appendix pp 6–7).

Discrimination was lower for BRCA-positive than for BRCA-negative participants (table 3). Model discrimination

0p=0·0004 p<0·0001

10–y

ear b

reas

t can

cer r

isk (%

)

5

10

15

20A BOADICEA B BRCAPRO

Observed riskAssigned risk

0<1·7% ≥1·7 to <3·4%

Risk quantile Risk quantile≥3·4 to <5% ≥5% <1·7% ≥1·7 to <3·4% ≥3·4 to <5% ≥5%

10–y

ear b

reas

t can

cer r

isk (%

)

5

10

15

20C BCRAT D IBIS

1·0%

3·1%4·7%

8·2%

0·8%2·6%

4·2%

9·7%

2·4%

4·6%6·1%

16·7%

0·8%2·5%

3·8%

13·6%

p<0·0001 p=0·00012·8%

4·5% 5·0%7·0%

0·8%2·5%

4·1%

7·8%

1·4%2·9%

4·8%

8·5%

0·8%2·5%

4·2%

10·0%

Figure 2: Calibration of 10-year breast cancer risk prediction scores for BOADICEA (A), BRCAPRO (B), BCRAT (C), and IBIS (D) in overall cohort, by risk quantileTriangles represent the mean risk predicted by the models, whereas squares represent the mean observed risk for each quantile. Error bars represent 95% CIs. p values represent the test of goodness of fit across all four risk quantiles. All eligible participants included in the analysis (n=15 732). BOADICEA=Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm model. BCRAT=Breast Cancer Risk Assessment Tool. IBIS=International Breast Cancer Intervention Study model.

See Online for appendix

33

VALIDACIÓN de los 4 modelos BCRAT,IBIS, BOADICEA,BRCAPROTerry MB, Lancet Oncology 2019

¿FUTURO MODELO HIBRIDO?

34

Articles


Registry and the Kathleen Cunningham Foundation Consortium for Research into Familial Breast Cancer [kConFab]) recruited between March 17, 1992, and June 29, 2011.17–19 Full details of ProF-SC,19 and baseline data collection and follow-up methods for the Breast Cancer Family Registry17 and kConFab,18 have been published previously.19,20 At baseline, ProF-SC included 18 856 women who did not have breast cancer. We restricted analyses to women who were aged 20–70 years at baseline and had no previous history of bilateral prophylactic mastectomy or ovarian cancer, at least 2 months of follow-up data since recruitment, and information available about family history from family pedigrees (n=15 732). We limited eligibility to women aged between 20 and 70 years at baseline so that we could validate a 10-year predicted risk score during our average follow-up time of 10 years (some risk models do not predict risk beyond age 80 years). All participants in the Breast Cancer Family Registry and kConFab provided written informed consent before enrolment, and the

study protocols at all sites were approved by institutional review boards at each participating institution.

ProceduresParticipants completed the same baseline risk factor questionnaire, which included questions about participants’ demographic characteristics, height, weight, history of benign and malignant breast disease, history of breast and ovarian surgeries, reproductive history, and lifestyle factors. The questionnaire also included questions about family history regarding breast and other cancers in the participants and their first-degree and second-degree relatives, including ages at cancer diagnoses and age last known to be alive. Family history and epidemiological information for the families participating in the ProF-SC cohort were collected at baseline and at each systematic follow-up interview; additional information was collected annually when families were contacted.

Screening for germline BRCA1 and BRCA2 mutations in the Breast Cancer Family Registry and kConFab17,21 typically involved screening the youngest affected family member at baseline. If that person carried a mutation, other family members were also tested. BRCA-negative women were defined as women who were not known to be BRCA1 or BRCA2 carriers, which included both tested and untested women.

Outcome information about breast cancer diagnosis, including pathological report, biopsies, and treatment questionnaires, were collected at diagnosis, or after the diagnosis was identified by personal or family reports or by cancer registry confirmation. Because the cohorts are based on family information, cancer information for each participant was typically self-reported or reported by a first-degree relative and could come from more than one source.

In the validation cohort, data were generally available for all items required by these models, except for mammographic density and polygenic risk score used in IBIS. However, data were scarce for pathologically confirmed lobular carcinoma in situ (used in IBIS) and atypical hyperplasia (used in BCRAT and IBIS).

We used the latest version (at the time of analysis) of software packages for each risk prediction model, IBIS (version 8b), BOADICEA (version 3), BCRAT (version 2.0), and BRCAPRO (version 2.1-3) to calculate 10-year risk scores. The data inputs required for each model to calculate breast cancer risk are summarised in table 1. BOADICEA, BRCAPRO, and IBIS all incorporate familial risk via detailed family history information across multiple generations. The breast cancer hazard ratio calculated by the IBIS model depends on several lifestyle and personal risk factors in addition to familial risk.5 BOADICEA6 and BRCAPRO7 do not include lifestyle and personal risk factors. BCRAT accounts for familial risk on the basis of the number of affected first-degree relatives diagnosed with breast cancer, and also includes other risk factors, such as age at menarche,

BOADICEA BRCAPRO BCRAT IBIS

Personal information

Current age Yes Yes Yes Yes

Year of birth Yes NA NA NA

Race or ethnicity NA Yes Yes NA

Age at menarche NA NA Yes Yes

Parity NA NA NA Yes

Age at first birth NA NA Yes Yes

Menopausal status NA NA NA Yes

Menopausal hormone therapy use NA NA NA Yes

Body-mass index NA NA NA Yes

History of atypical hyperplasia NA NA Yes Yes

History of lobular carcinoma in situ NA NA NA Yes

Previous breast biopsy NA NA Yes Yes

Mammographic density NA NA NA Yes

Polygenic risk score NA NA NA Yes

Information about the individual and their family members

First-degree relatives with breast cancer Yes Yes Yes Yes

Second-degree and third-degree relatives with breast cancer

Yes Yes NA Yes

Identical twin with breast cancer Yes Yes NA NA

Age at cancer diagnosis Yes Yes NA Yes

Bilateral breast cancer* Yes Yes NA Yes

Ovarian cancer* Yes Yes NA Yes

Pancreatic cancer* Yes NA NA NA

Prostate cancer* Yes NA NA NA

Molecular subtype of breast tumour Yes Yes NA NA

Vital status of family members† Yes Yes NA Yes

BRCA1 and BRCA2 mutation status Yes Yes NA Yes

Ashkenazi Jewish heritage Yes Yes NA Yes

These models do not necessarily require all data inputs for calculation of a risk estimate. BOADICEA=Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm model. BCRAT=Breast Cancer Risk Assessment Tool. IBIS=International Breast Cancer Intervention Study model. NA=not applicable. *In any family member. †Current age or age at death if deceased.

Table 1: Summary of inputs needed for the four assessed models of breast cancer risk

For more on IBIS see http://www.ems-trials.org/

riskevaluator/

For more on BOADICEA see https://pluto.srl.cam.ac.uk/cgi-

bin/bd3/v3/bd.cgi

For more on BCRAT see https://dceg.cancer.gov/tools/risk-assessment/bcrasasmacro

For more on BRCAPRO see https://projects.iq.harvard.edu/

bayesmendel/brcapro


#SEOM2019

Valoración de riesgo en cáncer hereditario Mensaje para casa (I):

• Para avanzar en las estrategias de cribado y prevención de cáncer de mama debemos disponer de modelos que estimen de manera adecuada el riesgo de padecer la enfermedad.

• BCRAT,IBIS, BOADICEA y BRCAPRO son los 4 modelos más usados para estimación de riesgo en cáncer de mama.

• Un estudio prospectivo de validación Terry MB, Lancet Oncology 2019 de una cohorte de más de 15.000 mujeres, concluye que IBIS y BOADICEA son los modelos que mejor discriminan el riesgo de padecer cáncer de mama a 10 años (aunque sobreestiman el riesgo en las mujeres de mayor riesgo y lo infraestiman en las de menor riesgo)

• Es necesario que los modelos estén validados en las poblaciones en las que se van a usar

• En el futuro sería deseable disponer de modelos híbridos que integren factores de riesgo poligénico (BOADICEA) con factores de riesgo ”individual “ como la densidad mamaria (IBIS)

#SEOM2019

Valoración de riesgo y medidas de prevención en cáncer hereditario Mensaje para casa:

URGE DISPONER DE DATOS PROSPECTIVOS DE NUESTRO PAIS à ESTUDIO DE LA SECCION SEOM de CANCER HEREDITARIO

valoración de riesgo y medidas de prevención en cáncer ......#seom2019 valoración de riesgo y...

Documents