1. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. EGAPP Recommendations for Lynch Syndrome Genetic Testing: Impact on Colorectal Cancer Care Heather Hampel, MS, CGC Professor, Division of Human Genetics The Ohio State University Department of Internal Medicine Columbus, OH

2. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Outline Why determine which cases of CRC have defective mismatch repair? Screening for Lynch syndrome among newly diagnosed CRC patients EGAPP recommendations OSU clinical experience doing IHC on all newly diagnosed CRC patients

3. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. 13% <1%85% FAPSporadic MIN (MSI+) (Microsatellite Instability) CIN (Chromosome Instability) Lynch Syn Sporadic MSI(+) Germline Mutation MMR genes MLH1, MSH2, MSH6 & PMS2 15% 2-3% Epigenetic silencing of MLH1 by hypermethylation of its promoter region 85% Colorectal Cancer Acquired APC, p53, DCC, kras, LOH,... Germline Mutation APC

4. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Why Determine which CRC Cases are MSI+ and which have LS? All MSI+ CRC patients have a better prognosis MSI+ CRC patients MAY need different treatment in future LS patients at high risk for second primary cancers (CRC and others) LS patients have at-risk relatives who could benefit from genetic testing

5. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Prognostic Implications Pooled data analysis of 32 studies with 7,642 cases found: –HR for Overall Survival with MSI = 0.65 –Restricting analysis to clinical trial patients (HR=0.69) did not alter benefit –Restricting to those with locally advanced disease (HR=0.67) did not alter benefit Popat S, et al. JCO. 2005;23:609-618.

6. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Treatment Implications MSI-H stage II and III patients did not have a significant difference in RFS whether or not they received 5-FU (HR, 0.96; 95% CI, 0.62 to 1.49; P=.86) MSI-H patients did not have a significant difference in OS whether or not they were treated with 5-FU (HR, 0.70; 95% CI, 0.44 to 1.09; P=.12) MSS patients do benefit from 5-FU (HR, 0.77; 95% CI 0.68 to 0.87; P<.001) Des Guetz G, et al. Euro J Cancer 45:1890-6,2009.

7. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Patient & Family Implications: Lynch Syndrome MLH1 MSH2 MSH6 PMS2

8. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Carrier Parent Non-carrier Parent Autosomal Dominant Inheritance Aa aa Aa aa Carrier Non-carrier 1/2

9. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Lynch Syndrome: Lifetime Risks for Cancer CancerLynch syndrome General Public Colon cancer56%22% Endometrial cancer35%26% Gastric cancer13%1% Ovarian cancer12%1.5% Small bowel, bladder, ureter, renal pelvis, brain <4% each<1% each

10. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Lynch Syndrome Surveillance Options Lindor N et al. JAMA 2006;296:1507-17. & Vasen HFA et al. J Med Genet 2007;44:353-62. InterventionRecommendation ColonoscopyEvery 1-2 y beginning at age 20 (MLH1), 25 (MSH2), or 30 (MSH6 & PMS2) Endometrial sampling Every 1 y beginning at age 30-35 Transvaginal U/SEvery 1 y beginning at age 30-35 Urinalysis with cytology Every 1-2 y beginning at age 25-35 History & Exam w/ review of systems Every 1 y beginning at age 21

11. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. 15-year prophylactic colonoscopic screening Järvinen et al. 1995 and 2000 ScreenedNot screened n=133n=119 Colorectal cancer819 n=0.014 Death from colorectal cancer0 9 p<0.001 Overall deaths10 26 p<0.001

12. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Lynch Syndrome Prophylactic Surgery Options Options include subtotal colectomy, hysterectomy, and oophorectomy Subtotal colectomy does not eliminate cancer risk Hysterectomy eliminates risk of endometrial and ovarian cancer Expert panels made no recommendation for or against surgery due to unproven efficacy Schmeler et al. NEJM 2006;354:261-9.

13. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Lynch Syndrome Implications for Patient 16-30% chance of second primary CRC in the 10 years after their first diagnosis NCCN guidelines differ for CRC patients with LS and without LS –With LS, colonoscopy every 1-2 years for life –Without LS, colonoscopy 1 yr after dx, repeat in 2-3 yrs, then every 3-5 years based on findings Management also changes due to the risk for other cancers

14. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Lynch Syndrome Implications for Family 6 relatives tested on average per proband identified with LS 50% with LS need increased cancer surveillance –High Compliance (96% CRC & 97% Gyn) –Cancer risk ratio of relatives with LS compared to relatives without LS is 5.8 –No significant difference in cancer mortality (RR, 2.28) or overall death rates (RR, 1.26) 50% without LS follow the ACS guidelines Jarvinen HJ et al. J Clin Oncol 2009;27:4793-7.

15. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Tumor Tests to Screen for Lynch Syndrome Microsatellite Instability (MSI) testing –Performed on DNA extracted from tumor and normal tissue – requires laboratory –Test is positive in 15% of CRC cases –Test is positive in 77-89% of LS cases Immunohistochemistry staining –Performed on thin slide of tumor – can be done in pathology department –1-2 proteins are absent in 20% of CRC cases –1-2 proteins are absent in 83% of LS cases

16. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. MSI testing on Genotyper

17. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Five Possible IHC Results: 1. Normal – All 4 Stains Present 80% of the time you will get this result CRC is probably not MSI+ Prognosis worse than if MSI+ Refer to Genetics ONLY if you suspect polyposis, patient dx <45, patient has had multiple CRC primaries, or the patient has an first degree relative (FDR) with CRC at any age

18. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. 2. Abnormal – MLH1 & PMS2 Absent 15% of the time CRC is MSI+ Better prognosis 80% acquired methylation of MLH1 20% will be LS MLH1 MSH2 MSH6 PMS2

19. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. 2. Abnormal – MLH1 & PMS2 Absent Either refer all cases to Genetics OR Refer those diagnosed under age 60, those with multiple primary LS cancers, and those with an first or second degree relative (SDR) with a LS cancer at any age OR Reflex to BRAF or MLH1 methylation testing & refer those without BRAF mutation or without methylation

20. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. BRAF and CRC V600E (1799T>A) mutation strongly associated with MSI+ and CpG island methylator phenotype (CIMP) Not yet reported in a patient with a germline MLH1 gene mutation MLH1 promoter methylation –MLH1 absent on IHC, no MMR gene mutation; 68% with V600E in BRAF

21. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. 3. Abnormal – MSH2 & MSH6 Absent 3% of the time CRC is MSI+ Better prognosis Most likely LS due to either MSH2 or MSH6 gene mutation Always refer to Genetics MLH1 MSH2 PMS2 MSH6

22. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. 4. Abnormal – MSH6 Absent 1% of the time CRC is MSI+ Better prognosis Most likely LS due to an MSH6 gene mutation Always refer to Genetics MLH-1 MSH-2 MSH-6 PMS-2

23. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. 5. Abnormal – PMS2 Absent 1% of the time CRC is MSI+ Better prognosis Most likely LS due to an PMS2 gene mutation Always refer to Genetics MLH1MSH2 MSH6 PMS2

24. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. The Family History Is Key to Diagnosing LS – or is it? CRC dx 50s CRC dx 45 CRC dx 61 CRC dx 75 Ovarian Ca, dx 64 CRC dx 48 CRC dx 52 Endometrial Ca, dx 59 CRC dx 42 45

25. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Amsterdam II criteria 3 or more relatives with verified Lynch- associated cancer in family Two or more generations One case a first-degree relative of the other two One CRC dx <50 FAP excluded Vasen HFA et al. Gastroenterology. 116:1453, 1999 Does not include ovarian, gastric, brain, biliary tract or pancreatic cancer

26. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Bethesda Guidelines Individual with CRC dx <50 Individual with synchronous or metachronous CRC, or other Lynch-associated tumors regardless of age Individual with CRC with MSI-H histology dx <60 Individual with CRC with >1 FDR with an Lynch- associated tumor, with one cancer dx <50 Individual with CRC with >2 FDRs or SDRs with an Lynch-associated tumor, regardless of age Umar A, et al. JNCI. 2004;96(4):261-268.

27. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Warning: Family Histories can be Deceiving Family size is getting smaller Wider use of colonoscopy likely to prevent many colon cancers MSH6 & PMS2 may have lower cancer risks

28. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Can predict who is more likely to have LS using family history criteria (Amsterdam & Bethesda) Can predict the likelihood of a MMR gene mutation using three new programs –PREMM1,2,6 http://www.dana-farber.org/pat/cancer/gastrointestinal/crc- calculator/ –MMRpro http://www4.utsouthwestern.edu/breasthealth/cagene/ –MMRpredict http://www1.hgu.mrc.ac.uk/Softdata/MMRpredict.php Can order MSI and/or IHC on tumor to screen for LS Can diagnose Lynch syndrome with genetic testing Identification of Lynch syndrome in the Genetics Clinic

29. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Identification of Lynch syndrome among all Newly Diagnosed CRC Patients Unlikely to have good family history High volume Must rely on screening tests for LS (MSI/IHC) Pathologists will know age at dx, synchronous primaries, but not likely to know all metachronous primary or family history of patients

30. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Columbus-area HNPCC study (1999-2005) Hampel et al. New Engl J Med 2005; 352:1851-60 Hampel et al. J Clin Oncol 2008; 26:5783-88 MSI positive (high & low) n=307 (19.6%) Deleterious mutation n=44* (2.8%) *2 had MSI- tumors Variant of uncertain significance n=55 (3.5%) Sequence Immunohistochemistry Methylation of MLH1 promoter Polymorphism or no mutation n=209 (13.4%) Colorectal cancer Total accrued (n=1600) Testing completed (n=1566) MSI negative n=1259 (80.4%)

31. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. CRC probands with deleterious mutations (n=44) Age at diagnosis – 51.4 (range 23-87) 50% diagnosed over age 50 25% did not meet either Amsterdam or Bethesda criteria Mutations –20.5% MLH1 –52.3% MSH2 –13.6% MSH6 –13.6% PMS2 Hampel et al. New Engl J Med 2005; 352:1851-60 Hampel et al. J Clin Oncol 2008; 26:5783-88

32. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. 35 CRC probands have had genetic counseling Family Studies of 35/44 CRC Probands Hampel et al. NEJM 2005;352:1851-60.; Hampel et al. JCO 2008. Degree of KinshipTestedPositive First9952 Second6428 > Second8629 Total249109

33. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Amsterdam: Yes Lynch: Yes 942+3 a>t MSH2 mutation

34. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Amsterdam: No Lynch: Yes 3155delAG MSH6 mutation

35. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Amsterdam: Yes Lynch: No Tumors MSI- with intact IHC

36. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. EGAPP Recommendations Evaluation of Genomic Applications in Practice and Prevention Established in 2005 to assess evidence regarding the validity & utility of rapidly emerging genetic tests for clinical practice Independent, multidisciplinary panel prioritizes and selects tests, reviews CDC- commissioned evidence reports, finds gaps, and provides guidance

37. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Steps in the EGAPP Working Group Review Process

38. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Seven Evidence Reports Available to Date 1.October 2006 - Genomic Tests for Ovarian Cancer Detection and Management 2.January 2007 – Testing for CYP450 Polymorphisms in adults with depression before trtmnt with SSRIs 3.May 2007 – Lynch diagnostic strategies 4.January 2008 – Gene Expression Profiling and Breast Cancer Outcomes 5.January 2009 – DNA strategies aimed at reducing morbidity and mortality from Lynch syndrome 6.January 2009 – Can UGT1A1 genotyping reduce M&M in pts with metastatic CRC treated w/Irinotecan 7.June 2009 – Outcomes of genetic testing in adults with a history of venous thromboembolism

39. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Four EGAPP Working Group Recommendations Insufficient Evidence to recommend for or against 1.Tumor profiling to improve outcomes in patients with breast cancer 2.UGT1A1 genotyping to reduce morbidity and mortality in patients with metastatic CRC treated with Irinotecan 3.Use of CYP450 testing to predict response to SSRis in adults with depression Sufficient Evidence to recommend for 4. Screening newly diagnosed CRC patients for LS with either MSI or IHC

40. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. EGAPP Recommendations Moderate certainty that testing patients with CRC for LS and then testing their relatives would provide moderate population benefit. Adequate evidence to conclude that the analytic sensitivity and specificity of the preliminary and diagnostic tests were high. Adequate evidence to describe the clinical sensitivity and specificity of three preliminary tests and four testing strategies. Adequate evidence for testing uptake, compliance with surveillance, relatives approachable, harms associated with f/u and effectiveness of routine c-scope supporting the use of genetic testing strategies to reduce morbidity and mortality in relatives with LS. No one test strategy was clearly superior. Inadequate evidence that screening for LS will reduce EC morbidity or mortality EGAPP Genet Med 2009;11:35-41; Palomaki G, Genet Med 2010;11:42-65.

41. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Potential Impact 146,970 new cases of CRC in the US in 2009 4,115 have Lynch syndrome (2.8%) 12,345 of their relatives have LS (~3 per proband) Total of 16,460 individuals who could be diagnosed with LS this year with universal screening American Cancer Society Facts & Figures

42. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Choosing the Screening Test: MSI vs. IHC IHC is available in virtually all hospitals MSI requires molecular diagnostics and normal for comparison IHC with 4 antibodies is similar in cost to MSI with 5 markers IHC directs gene testing saving money Ethical issues surrounding IHC IHC and MSI have limitations

43. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Cost-effectiveness Study Follow-up to EGAPP evidence review Modeling used the statistics from the EGAPP review My role on the project was to: –Explain the various strategies one might use to screen for LS –Provide Medicare reimbursement rates & list prices for various tests

44. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Strategies Compared

45. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Cost-effectiveness Results

46. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Incremental Cost-Effectiveness Ratios per LYS compared to no testing at all StrategyMedicare rates List prices from labs 12 relatives IHC, BRAF testing & sequencing $22,552$30,331$12,332 IHC testing & sequencing $23,321$30,740$12,663 MSI testing & sequencing $41,511$49,272$20,470 Genetic sequencing for 4 genes $142,289$200,037$63,773

47. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Cost-Effectiveness Evaluation Universal screening detects nearly twice as many cases of LS as targeting younger patients Strategy 1 is the most cost effective strategy Cost-effectiveness ratio of universal screening is < $25,000 per life-year saved relative to no testing ICER comparable with other preventive services (colonoscopy every 10 years has ICER of $25,000)

48. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Universal IHC screening for CRC: OSU experience Genetics notified by pathology of all abnormal CRC results Permission from ordering physician to contact patient Patient contacted –Take limited family history –Make recommendation for genetic consultation Letter sent If contact cannot be made, letter is sent explaining results with our contact information Gyn/Onc’s notify their own patients regarding their IHC results

49. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Universal IHC screening for CRC: OSU experience Began March 1, 2006 270 cases of CRC in first 2 years –57 (21.1%) absent for one or two MMR proteins –54 contacted by genetics with physician consent 5 deceased, reported to next of kin 7 prisoners –34 appropriate for consultation –18 scheduled appointment –9 completed appointment –7 tested –2 confirmed Lynch, 3 with MLH1 methylation South et al, Genet Med 2009; 11:812-817

50. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC.

51. All proteins present (80%) MSH2 and/or MSH6 absent; PMS2 only absent (5%) How to Follow-up on IHC Results MLH1 and PMS2 absent (15% ) STOP Sequence and large rearrangements for absent one(s) No germline mutation in MLH1, MSH2, MSH6, PMS2 Consider family history, MSI analysis BRAF mutation analysis BRAF mutation present (10-12%) BRAF mutation absent (3-5%) Sequence and large rearrangements for MLH1

52. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Universal IHC - Challenges These patients are not as motivated to seek genetic counseling and testing –Many who likely have Lynch syndrome declined further counseling/testing –Prisoners?? Many do NOT have Lynch syndrome but we cannot rule these out without further testing - not easy to order and cost –BRAF testing has helped with this tremendously

53. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. OSU Successes and Pitfalls Successes –Proven need for tumor testing rather than family history reliance –Proven equivalence of MSI vs IHC –Institutional buy-in for universal screening –IHC plus BRAF to optimize efforts Pitfalls –Need for multi-provider communication of tumor results to increase patient follow through –IHC only routine on primary CRC resections Uninformative on many polyps IHC should be done on initial biopsy for rectal cancers since neoadjuvant radiation reduces available cancer cells Can be ordered on any specimen –Each institution requires adherence to pathology standards to assure equivalence of results

54. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Conclusions 1 out of every 35 CRC patients has LS Family history criteria will miss 25% of CRC patients with LS Lives can be saved by diagnosing LS early Universal Screening for LS among all newly diagnosed CRC patients –Is feasible –Is recommended –Is cost-effective

55. © CoC 2011—Content cannot be reproduced or repurposed without written permission of the CoC. Acknowledgements Albert de la ChapelleJenny Panescu Judith WestmanJan Lockman Ilene ComerasJennifer LaJeunesse Wendy FrankelDan Fix Julie StephensLeigha Senter Thomas PriorMark Clendenning Jeffrey FowlerKaisa Sotamaa David CohnYange Zhang Edward MartinHidewaki Nakagawa Mark ArnoldMartha Yearsley