1. International Society for Gastrointestinal Hereditary Tumours (InSiGHT): Edited highlights Düsseldorf, 2009

2. InSiGHT’s history In 1985 polyposis experts (Dukes, Bussey, Morton et al … ) met at Leeds Castle, Kent, UK. In 1989 the Leeds Castle Polyposis Group (LCPG) was formally established. The International Collaborative Group on Hereditary Non Polyposis Colorectal Cancer (ICG-HNPCC) was conceived in Jerusalem, in 1989. In 1990 the first formal meeting of the ICG-HNPCC was organized in Amsterdam. In 1997 the first joint meeting of LCPG and ICG-HNPCC took place in Noordwijk. At the meeting in Venice in 2001 it was agreed that the ICG-HNPCC should formally merge with the LCPG to form a new society. InSiGHT’s journal is Familial Cancer www.insight-group.org Members include physicians, geneticists, surgeons, pathologists, scientists, counsellors etc It co-ordinates Registries on an International basis

4. Pre-meeting UVs Feedback from Sian Tavtigian post-Lyon Feb 09 Adding phenotype to InSiGHT mutation database Functional assays Worldwide collaborative project on UVs –International interpretation panel –Feb 2013 InSiGHT meeting in Melbourne will be linked to HVP –Nature Genetics paper of all 10,000 MMR mutations

5. Phenotype No phenotype FHx +/- AC +/- Summary FH based on Barnetson & Chao Pedigree (for segregation)

6. Algorithms Family mutation probability: Wijnen Individual mutation probability: Barnetson R et al NEJM 2006; 354:2751-63. [See also the PREMM 1,2,6 model] Developed from a population 55 https://hnpccpredict.hgu.mrc.ac.uk Pr/(1-Pr) = 1.39 x 0.89Age x 2.57Gender x 4.45Location x 9.53Sync/Met x 46.26CRCFH( 50) x 59.36ECFH Stage 1 and Stage 2 * Age is expressed in years at time of diagnosis * Gender is 1 = male, 0 = female * Location is 1 = proximal, 0 = distal tumours (Proximal colon = any of the following sites Splenic flexure, Transverse colon, Ascending colon/hepatic flexure, Caecum, Appendix) * Synchronous and/or metachronous tumours are: Sync/MetTumour = 1 and no Sync/MetTumour = 0; * There are three possible colorectal cancer family history categories CRCFH( =50): 1 and 0 if the youngest affected first degree relative was aged <50yrs 0 and 1 if the youngest affected first degree relative was aged >=50yrs 0 and 0 if there were no affected first degree relatives; * Family history of endometrial cancer: ECFH = 1 if there are any first degree relatives with endometrial cancer ECFH = 0 if none.

7. InSiGHT summary FH phenotype dataset As these phenotype parameters are used to assess mutation probability, they can be used to inform the Bayesian probability of risk for any unclassified variant. Hence: Stage 1 –Age, gender, location, sync/met, FH, ECFH Stage 2 –IHC –MSI Stage 3 –Additional family data: e.g. # affecteds –Co-segregation

8. UVs likelihood ratios Five classes 5: >0.99 4: 0.95-0.99 3: 0.05-0.95 2: 0.001-0.05 1: <0.001 Need to quantitate “natural variation” in results HGVS @ UNESCO HQ, Paris, Jan [May] 2010

9. Functional assays (IHC) Robert Hofstra, Leiden Couch FJ Hum Mutation 29:1314-26 Assymetrical F-tagged PCR construct with an engineered mismatch In vitro expression of MSH2 etc Add to –MSH2 brew Relative heights of cut/uncut peaks gives measure of in vitro activity Quick, cheap, simple, easy …

10. PMS2 & BRAF status Abnormal PMS2 expression alone in tumours Strong indicator of PMS2 mutations MLH1 + PMS2 absent … but weak MLH1 on re-testing ! iQC & EQA Others find (uninterpretable) complex combinations of IHC abnormality Tumours with BRAF V600E in PMS2 families Probably sporadic Due to ascertainment bias consequent upon clinical referral criteria PMS2 carriers ascertained secondary to homoz children have v little if any FH

11. Houlston, Dunlop, Tomlinson, Gallinger, Peters et al 10 SNPs = 6% of FCRC risk Nature Genetics 40:1426-35 [100 SNPs = 80%] [172 SNPs = 100%] Issues: Population heterogeneity Rare private variants Incomplete SNP capture Copy Number Variants need studying Gene-Environment and GxG interactions Deep re-sequencing needed, to find private variants % of FCRC risk SNPs (n) Genome-Wide Association Studies

12. LS Modifiers (Houlston) CycD1 G242AMSH2, but not MLH1 MTHFR –C677T & age of onset – iffy –A1298C – iffy –but compound hets 15 y younger IGF1(dinuc) <18rpts = younger onset & age of onset α 1/rpt RNASEL R462Q HFE (?) … 4 others

13. Main meeting

14. PMS2-PMS2CL hybrids, Wimmer PMS2-PMS2CL hybrid alleles Inv dup 7p22.1 φ = exons 9, 11-15 Non-homol recomb Allele drop out Hybrid 1-12,φ13-15 PCR by exon specific SNPs - 13: 2253T/C; 14: 2324A/G; 15+ *92dupA 98/300 alleles are hybrid

15. MSI tumour immunology Tumour infiltrating lymphocytes. in tumours with MSI Novel antigens in MSI tumours due to coding microsatellite (MS) mutations

16. MSI tumour immunology T cell Immune responses FrameShiftPolypeptide-specific T cells are present but tumours grow FSP-specific T cells are present in LS patients, without tumours … Patients are probably being immunised by tumours which never present clinically Humoural IR Highest after resection or adv tumours Low level ab in healthy LS pts Immune evasion: HLA Class I expression is lost because of beta2M/HLA A2 – coding MS muts Known for a long time New finding: HLA Class II is lost through CIITA 1962_1963insC, RFX5 56delC, RFXANK, RFXAP muts

17. MLH1 epimutations, Megan Hitchins MLH1 epimutation segregates in a 3 gen fam Soma-wide, allele specific, mosaic T allele methylated Loss of T allele in mRNA + loss of C allele in tumour 3 affecteds and 5 unaffecteds had soma methylation (14-49%) AD transmission Father’s sperm demethylated Son has soma methylation So, now reversible non-Mendelian & Mendelian MLH1 epimutation ?distinct mechanisms

18. EPCAM/TACSTD1 Different deletions, vary in size, cause same effect Methylation of MSH2 promotor Common Dutch founder EPCAM c.859-1462_*1999del 5 patients –1 AC+ –4 FH –CRCa x31, DUCa x1, SBCa x2, BRCa x1, PRCa x3, Leuk x2, ENCa x0 Homoz mutn of EPCAM causes congenital tufting enteropathy ?hemizygosity contributes to cancer EPCAMMSH2 AluJo

19. CAPP2, Burn LS pts treated with Aspirin 600 mg/d Resistant starch 30 g/d Both Neither (placebo) for 29 months (7-74 months) No apparent initial effect on tumour incidence, but delayed action, and significant effect now being seen (as in other aspirin studies) Fodde: aspirin attenuation of betaCATenin/TCF4 pathway, ^stability of betaCAT Ca stem cells produce IL4, but aspirin inhibits IL4

20. Novel CRCa genes, Katachalam 32 YOCC with MSS tumours aCGH Novel CNVs in 5 Dup PTPRJ/DEP-1 11p11.2 110-160 Kb, contains 1-2 genes Candidate human TSG & mouse susceptibility locus So, 3 lines of evidence suggesting significance 1500 Fam CRCa + controls tested for the CNV 2 dups and 4 dels Various sizes ? Why both del and dup associated with same phenotype

21. MMR mut prob model, Kastrinos MMRpredict MMRpro PREMM 1,2,6 http://www.dfci.org/premm

22. MSH6 risks, Jenkins 70y80yHR CRCaM22%44%7.6 F10%20% ENCa26%44%26 Others*M 3% 6%0.8 F11%22%6 *Gastric, Small bowel, Kidney, Ureter, Ovary, Brain Differences between 70 and 80 are due to large rise in population risks at this age

23. MSH2 extracolonic risks (Standardised Incidence Ratios), Engel (German Registry) MFSIR Bladder x75y10.115.47.6 5.8% 4.8% Breast1.84 (1.4 – 2.4) 15.1% Ovary13.4MSH2/6 highest 8.0%MLH1 lower MLH1MSH2MSH6 Gastric, SBHighHighLow BladderLowHighLow Ovary(low)HighHigh

24. HFE and CRCa risk, Scott HFE C282Y and H63D x3 CRCa risk H63DHR 2.93 (1.3 – 6.4) p 0.007 If H63D homoz then ~6y earlier onset Shiu et al Int J Ca 125:78-83 However: HFE is within HLA on chr 6, so is this an association with HLA and not HFE per se … Another group reported an association with C282Y but not H63D, so could well be either a ‘spoof’ association or very population specific, implying HLA not HFE.

25. LS Modifiers, Wijnen 14 SNPs associated with CRCa Are they associated with LS risks? On testing Leiden’s mega set of LS families: AllMF 8q23.3(x2)x3rs16982766 C allele (related to MYC) 11q23.1x3rs3802842 4q13.1x2rs4355419 >3 risk alleles increase risks further ?warrants more intensive surveillance

26. Prostate Ca risk in LS, Pål Møller MSH2, MLH1, MSH6 Obs 9 cases of PRCa, expected 1.52, p <0.01 MSH2MLH1MSH6PMS2 6021 7/8 tumours abnormal IHC Age 60.4, cf. 66.6, p <0.01 5/8 had high Gleason scores (8-10), v signif.

27. IHC in LS, Klarskov & Frayling Interobserver variability –Variability –Tyros and experts no different UK NEQAS ICC –Technical scheme –~50% of 55-74 labs score <12

28. MutYH Elke Holinski-Feder MutYH mutations now observed in a wide variety of phenotypes: FAP, AFAP, ATFAP Christina Thirlwell, London E466X diagnosed 10y earlier than Caucasians with 165/382 Later stage Ca ?more adenomas Café-au-lait spots ?! Sampson CRCa risk in carriers SIR 2.12 (1.3-3.3)

29. MutYH Elke Holinski-Feder MutYH mutations now observed in a wide variety of phenotypes: FAP, AFAP, ATFAP Christina Thirlwell, London E466X diagnosed 10y earlier than Caucasians with 165/382 Later stage Ca, ?more adenomas, Café-au-lait spots ?! Sampson CRCa risk in carriers SIR 2.12 (1.3-3.3)

30. Hereditary Pancreatic Ca, Bartsch Testing BRCA2 if 2 FDR with PACa, or any 3 rels with PACa p16 if malignant melanoma in F/SDR of PACa case PRSS1, MMR, LKB1, ATM, APC as per FH