Genetic testing for breast cancer Susan M. Domchek, MD Basser Professor of Oncology

1. Genetic testing for breast cancer Susan M. Domchek, MD Basser Professor of Oncology University of Pennsylvania

2. Sex Age Family history Depends on specific of family history Depends on whether there is a known genetic susceptibility Reproductive history Early first period Late last period Postmenopausal estrogen use Late first child No breast feeding ETOH Obesity Lack of exercise Risk Factors for Breast Cancer

3. Germline – the genes you are born with Can be passed on to relatives Does not mean that disease will happen Increased risk of disease There is no one “breast cancer gene” Somatic – changes in tumors that are acquired over time Can not pass on to relatives Can be tested as part of decision making for therapy for cancer Germline vs Somatic Genetics

4. Genetics :Cancer Risk Variants Common Variants Single nucleotide polymorphisms Allele Frequency CHEK2, ATM, NBN BRCA1, BRCA2, TP53 Rare variants (moderate) Rare variants (high) 1 2 5 ≥10 Relative Risk

5. Hereditary breast cancer Adapted from Couch, Nathanson, & Offit, Science 2014

6. Risk Assessment Disease Prevention Therapeutics Germline genetic testing as a paradigm for individualized care BRCA1/2 as the prototype

7. Breast cancer: 50%-70% Second primary breast cancer: 40%-50% Ovarian cancer: 15-55% BRCA1>BRCA2 Increased risk of other cancers: Male breast cancer BRCA2>BRCA1 Pancreatic cancer BRCA2 Prostate cancer BRCA2 Melanoma BRCA2 BRCA1/2-associated cancers: lifetime risk

8. Who should be considered for testing? • Breast cancer <45 • Ovarian cancer cases (particularly high grade serous) • Male Breast Cancer • Breast and ovarian cancer in a single lineage • 2 or more women with breast cancer <50 • Ashkenazi Jewish with breast or ovarian cancer • Breast cancer < 60 and triple negative • Bilateral breast cancer <60 • Pay attention to pancreatic cancer and high grade prostate cancer • Ashkenazi Jewish individuals? • All women at age 30? Many issues related to population screening

10. Estimates of breast cancer risk in BRCA1 carriers: Significant variability in penetrance

11. Genetic modifiers: CIMBA Polygenic risk scores (PRS) using BC susceptibility SNPs identified through population-based GWAS 15,252 BRCA1 8,211 BRCA2 BRCA1 carriers Kuchenbaecker et al in press 2016

12. Risk Reducing Salpingo-Oophorectomy and the risk of breast cancer RRSO and the risk of ovarian cancer Domchek et al, JAMA 2010 PROSE Consortium

13. RRSO and all-cause mortality Domchek et al, JAMA 2010 Domchek et al, JAMA 2010

14. Treatment of BRCA1/2-associated cancers: Platinum and PARP inhibitors • Olaparib • Veliparib • Rucaparib • Niraparib • BMN-673 Poly ADP ribose polymerase (PARP) plays a role in the repair of single strand breaks through base excision repair Significant responses observed in patients with germline BRCA1/2-associated breast and ovarian cancer Tutt et al, Lancet 2010 Audeh et al, Lancet 2010 Gelmon et al, Lancet Oncology 2011

16. Tumor shrinkage Tutt et al, Lancet 2010

17. Approval is for germline BRCA1 and BRCA2 associated ovarian cancer after treatment with >3 lines of therapy The FDA did not approve maintenance therapy EMA did approve maintenance

18. Multiple tumor types Cisplatin-resistant ovarian cancer Breast cancer with >3 lines of therapy in metastatic setting Pancreatic and prostate cancer Kaufman et al, JCO 2015 Domchek et al, GynOnc2016

19. Kaufman et al JCO, 2015

20. Platinum resistance and PARP treatment • Data from the gynonc paper Domchek et al, GynOnc 2016

21. Genetic testing has become complicated…. Single nucleotide polymorphism panels

22. Key Points • Not comprehensive sequencing of genes – such as BRCA1/2 • Not a stand alone for those with a strong family history • Some change in reclassification (change in how you consider someone from a risk perspective) • Calibration: How closely the predicted probabilities agree with the actual outcome • Clinical utility (or actionability?) • Will more women take tamoxifen? • How should this impact screening in the era of changing screening recommendations? • Ongoing studies

23. In the US - thishas become very complicated….

24. *Rest of genes on Fulgent: BLM, BUB1B, CTNNB1, CYLD, DDB2, DICER1, EGFR, EGLN1, ERCC2, ERCC3, ERCC4, ERCC5, EXO1, EXT1, EXT2, FANCA, FANCB, FANCD2, FANCE, FANCF, FANCG, FANCI, FANCL, FANCM, GPC3, HRAS, KIF1B, KIT, MC1R, MPL, MSH3, NF2, PDGFRA, PICALM, PMS1, PRKAR1A, PRKDC, PTPN11, RB1, RBBP8, RBM15, RECQL4, ROBO2, SBDS, SLX4, SMARCB1, SUFU, TERT, TSHR, TYR, WRN ,WT1, XPA, XPC, XRCC3

25. Revolution of genetic testing Assess patient Test for most likely gene(s) Test for most likely gene(s) Disclose result and reassess

26. New approach? Assess patient Send multigene panel Disclose result and reassess

27. Why do this? • More cost effective (for the testing) to do multigene rather than serial testing • Patients (and providers!) can get testing fatigue • The same cancer can be seen with different genes mutations • Ovarian cancer in both BRCA1/2 and Lynch • Uterine cancer in Lynch and Cowden • Breast in Li-Fraumeni and BRCA1/2 • Isn’t more better?

28. Potential Issues • High penetrance and moderate penetrance genes are on one panel • Implications for counseling • Keeping track of it all • Don’t we recognize clinical syndromes? • (And if we don’t – what does it mean?) • Variants of uncertain significance • Clinical utility: order tests you will act on • At least actionability Domchek et al, JCO 2013

29. What will we find? BRCA1/2 negative patients with BC <40 N=278 Patients with Class 4 VUS & Class 5 Mutations N=31 (11%) Class 3 VUS(s) only N=49 (18%) MUTYH Heterozygotes N=6 (2.2%)* No Class 3-5 Variants N=192 (69%) Bin A Genes TP53, PTEN, STK11, CDH1, CDKN2A, MLH1, MSH2, MSH6, PMS2, MUTYH (AR) Risk established for breast or other cancers Guidelines available *Clinically actionable* N=7 (2.5%) Bin B Genes ATM, BARD1, BRIP1, CHEK2, FAM175A, MRE11A, NBN, PALB2, RAD50, RAD51C Risk established for breast and some other cancers Less clear actionability N=24 (8.6%) 6% 13% 13% 31% 74% 63% White Non-white Class 3 VUS Class 4/5 Mutation No Class 3-5 Variants MUTYH N=1 CDKN2A N=1 ATM & CHEK2 N=18 Other genes N=6 TP53 N=4 MSH2 N=1 Maxwell et al GIM, 2014

30. What do we do? ACCE Framework

31. What is actionable? • Something that potentially could be acted upon • It does not mean that it is acted upon • It does not mean that such action benefits a patient • Actionability = clinical utility • Critically important that all this be studied

32. Summary of Clinical Validity Easton et al, NEJM 2015

33. Clinical utility • Risk assessment • Value of the true negative • Risk of breast and as well as risk of second primary cancer • Risk of other cancers (Ovarian cancer risk for BRCA1/2 was a major reason for rapid uptake of testing)

34. Clinical utility • Screening and prevention • Need to understand risks and benefits • What age to start screening? • What screening? • What age to have preventative surgery? • What to do with “unexpected” high penetrance mutation • When we find things we don’t expect, what should we do?

35. Clinical utility • Therapeutics • Prognosis: may impact administration of adjuvant therapy • Drug development/selection • Will tumors with mutations in these other genes be sensitive to specific types of drugs?

36. Conclusions • Genetic testing can be very useful to patients and their family members • Both the prevent and to treat cancer • Genetic testing is continuously evolving • BRCA1 and BRCA2 mutations are the most commonly found and we have reasonable data on how to manage • New genetics tests are often less clear in terms of how to change patients care – and improve patient outcome • Variants of unknown significance should NOT be managed as mutations • In the face of rising prophylactic mastectomies, we need to emphasize to patients how mutations in these genes are different from those in BRCA1/2