December 11 th Gail, our manager, was retiring and our statistician had gone on maternity leave All the money was gone but Anne Marie had come as a visiting fellow so I asked her to plot out the cancers covering the whole period as we now had data up to 10 years since commencement. I jokingly predicted that the lines would separate at 4 yeasr just after our last analysis
Per protocol analysis showing proportion who developed CRC We decided to use tablet counts. 2 per day for 2 years equals 1460 tabs We rounded down to 1400 (blind to the data) to allow for those who missed some tablets Note there are few cases at 10 years out so confidence limits expand Note also that the aspirin placebo group who dropped out early are the same as the aspirin treated groups suggesting they may have dropped into self medication IRR is equivalent to HR using Poisson to allow for multiple primary events eg a person gets endometrial cancer at 3 years and crc at 6 years
THIS GRAPH ANALYSES ALL Lynch syndrome cancers rather than just colorectal Note again that the early finish placebo aspirin group are tracking with the main treatment line.
Mismatch repair – corrects bases which are not paired according to the Watson-Crick base pairing A-T and C-G.
CIMP-negative, chromosomal instability, mainly MSS, origin in adenomas (may be sporadic, FAP-associated or MUTYH (formerly MYH) polyposis associated 108) (57%). CIMP-low, KRAS mutation, MGMT methylation, chromosomal instability, MSS or MSI-L, origin in adenomas or serrated polyps (20%). CIMP-high, methylation of MLH1, BRAF mutation, chromosomally stable, MSI-H, origin in serrated polyps, known generally as sporadic MSI-H (12%). CIMP-high, partial methylation of MLH1, BRAF mutation, chromosomally stable, MSS or MSI-L, origin in serrated polyps (8%). Lynch syndrome, CIMP-negative, BRAF mutation negative, chromosomally stable, MSI-H, origin in adenomas (3%) (familial MSI-H CRC) .
Mismatch repair – corrects bases which are not paired according to the Watson-Crick base pairing A-T and C-G.
The first case of what we now term a constitutional epimutation to be identified in humans was reported by Richard Kolodner’s group in 2002 in the mismatch repair gene, MLH1, in a patient diagnosed with Lynch syndrome. Constitutional epimutations of MLH1 are characterised by soma-wide monoallelic methylation of the MLH1 promoter, that is methylation of a single allele. We know this because some epimutation carriers are heterozygous for the common G/A SNP in the MLH1 promoter, and through methylation-sequencing of individual alleles, the methylation is shown to be associated with just one genetic allele. Because this methylation is present in normal somatic tissues, epimutation carriers are typically identified by the detection of methylation in their peripheral blood DNA. The effect of the epimutation is to cause transcriptional silencing of the methylated allele, and once again we know this by tracing the activity of SNPs that are present within the expressed part of the gene. In cases that are heterozygous for an exonic SNP, analysis of their RNA has shown that only one allele is detected in their mRNA, due to loss of transcriptional activity from the other allele. Finally, MLH1 epimutations occur in the context of a normal gene sequence – there are no genetic mutations identified in the gene or its vicinity.
Patient YT, represents the youngest affected case reported to date, but his profile is otherwise typical of other cases with a constitutional MLH1 epimutation. YT was 18 years old when he presented with cancer of the ascending colon, but had no family history of LS. The tumour demonstrated microsatellite instability and immunohistochemical loss of MLH1 and PMS2, but no germline mutation was identified within MLH1 and no evidence for structural alterations of MLH1 or MSH2 was identified by MLPA. Methylation testing revealed hemiallelic soma-wide methylation of the MLH1 promoter, as shown in these epigrams. Approximately half of his alleles were hypermethylated, as seen by the black dots, whilst the other half were unmethylated as indicated by the white dots.
A small number of families in whom the affected members have loss of the MSH2 protein in their tumours, have now been identified with deletions of the EPCAM gene located immediately upstream of MLH1, which give rise to a somatic epimutation of MSH2. In these cases, there is mosaic methylation of the MSH2 promoter, which is predominant in epithelial tissues. Although MSH2 itself in intact, this epigenetic error is caused by a linked deletion of the final exons of EPCAM. These abolish the transcription termination signal and cause elongation of transcription into MSH2 to produce fusion transcripts, ultimately leading to MSH2 methylation. Since this type of epimutation is associated with a cis-acting genetic defect, it demonstrates dominant inheritance, and has been found in large Lynch syndrome kindreds.
MCC 2011 - Slide 6
ESO-ESSO Masterclass in colorectal cancer surgery Genetic and epigenetic analysis of sporadic colon cancer-Decision & treatment Professor Sir John Burn MD FRCP FRCPCH FRCOG FMedSci Institute of Human Genetics, Newcastle University, UK Institute of Human Genetics Newcastle University Centre for Life, Newcastle UK
plan <ul><li>Stop press Aspirin treatment </li></ul><ul><li>CIN and MIN </li></ul><ul><li>Kras and BRAF </li></ul><ul><li>CIMP </li></ul><ul><li>Epimutation </li></ul><ul><li>“ trapped” Darwinian selection </li></ul><ul><li>immunoprevention </li></ul>
December 11 th 2008;359:2567-2578 CAPP2: 1009 Lynch syndrome recruits Gail Barker Anne-Marie Gerdes Julie Coaker Lynn Reed John Mathers Tim Bishop 600mg aspirin for up to 4 yrs had no effect
CAPP2: Per protocol analysis (primary endpoint CRC after 2 years treatment) significantly fewer colorectal cancers [submitted] HR 0.45 (0.22-0.93) p = 0.03 IRR 0.41 (0.20-0.84) p= 0.01
CAPP2 per protocol analysis All Lynch syndrome cancers HR 0.48 (0.28-0.83)p= 0.008 IRR 0.44 (0.26-0.75)p=0.002
Normal epithelium Adenoma Carcinoma Metastases APC, K-ras 12p, DCC 18q, p53 17p, …….. 85 % 15 % CIN MSI FAP HNPCC hMLH1, hMSH2, TGF-ß RII, Bax, TCF4, ACVRII, Caspase 5 ………. The Genomic Pathogenesis of Colorectal Cancer
Molecular Classification of Colorectal Cancer Jass, Histopathology 2007, 50, 113–130 FAP Less responsive To 5FU? Gallinger group Ribic CM et al NEJM 2003 ;349:247-57,
Colorectal Cancer Genetics & 5-FU Ribic CM NEJM 2003 MSS MSI Hazard Ratio 0.69 (0.50-0.94) p=0.02 2.17 (0.84-5.55) p=0.10 Courtesy Steve Gallinger
<ul><ul><li>BRAF is a kinase encoding gene from the RAS/RAF/MAPK pathway </li></ul></ul><ul><ul><li>The BRAF-V599E hotspot mutation is found in 40% of sporadic MSI-H tumours and 0% of HNPCC tumours </li></ul></ul><ul><ul><li>Presence of V599E mutations excludes diagnosis of HNPCC </li></ul></ul>MSI and BRAF
SEQUENOM TM MALDI TOF MASS SPECTROMETER Matrix Assisted Laser Desorption/Ionisation Time of Flight mass spectrometry Laser Flight path Detector Time of flight Sequenom chip
KRAS in colorectal cancer <ul><li>23 KRAS activating mutations </li></ul><ul><li>Codons 12, 13, 61 and 146 </li></ul><ul><li>BRAF V600E </li></ul>KRAS c.38G>A; p.13Gly>Asp KRAS c.35G>T; p.12Gly>Val Wild type KRAS c.436G>A; p.146Ala>Thr KRAS c.1799T>A; p.600Val>Glu
T G G A C C G T T C A A C - ve - ve - ve - ve - ve - ve - ve Polymerase - ve - ve - ve - ve - ve - ve - ve C
% resistance change A C T G A C T Time = ACTACT
Epigenetics <ul><li>“ The chemical and physical code superimposed on the genome that influences gene activity without altering the DNA sequence ” . </li></ul>Adapted from Jane Qiu, Nature (2006) 441:143-145 C G CH 3 CH 3 CH 3 C G Cytosine Methylated cytosine Megan Hitchins N N HO NH 2 1 2 3 4 5 6 N N HO NH 2 1 2 3 4 5 6 CH 3
Cytosine methylation at CpG islands CpG m CpG N N HO NH 2 1 2 3 4 5 6 CH 3 5-methylcytosine cytosine Normal biallelic expression Biallelic methylation – eg cancer cells N N HO NH 2 1 2 3 4 5 6 TF TF
A C A T C A C G T G A TT A C A Sodium bisulphite (NaHS0 3) conversion of genomic DNA Genomic DNA Treatment with sodium bisulphite Unmethylated sequence Methylated sequence A T A T T A C G T G A TT A T A A T A T T A T G T G A TT A T A
Constitutional epimutation of MLH1 <ul><li>Soma-wide mono-allelic promoter methylation </li></ul><ul><li>Detected by the presence of MLH1 methylation in peripheral blood DNA </li></ul><ul><li>Transcriptional silencing of the affected allele </li></ul><ul><li>Normal gene sequence </li></ul>CH 3 CH 3 CH 3 CH 3 MLH1 -93G>A 655A>G X TF G A G A G G G
Classic example of epimutation carrier <ul><li>Sri Lankan male, Patient YT </li></ul><ul><li>CRC (ascending) 18y </li></ul><ul><li>No family history </li></ul><ul><li>No germline mutation </li></ul><ul><li>MSI-H (all 5 markers) </li></ul><ul><li>IHC loss of MLH1 & PMS2 </li></ul>Soma-wide hemiallelic methylation of MLH1 by clonal bisulphite sequencing Goel et al. Int J Cancer (2010) in press PBL Hair follicles Saliva Buccal Methylated CpG Unmethylated CpG
Tissue-restricted MSH2 epimutation caused by terminal deletion of EPCAM <ul><li>Mosaic methylation of MSH2 promoter - predominant in epithelial tissues (eg colonic mucosa) </li></ul><ul><li>MSH2 is mutation-negative </li></ul><ul><li>Linked deletion of the final exon(s) of EPCAM, which abolish the transcription termination (polyadenylation) signal </li></ul><ul><li>Dominant inheritance: large kindreds with CRC showing MSH2-loss </li></ul>EPCAM MSH2 1 5 3 4 2 6 8 7 9 1 5 3 4 2 6 “ Fusion transcript” Normal allele Deletion of EPCAM 3’ 17kb
TGT . AAA . AAA . AAA . A CG . TGC . TGG . CTA . GCT . GA......... C K K K T C W L A . . . TGT . AAA . AAA . AAA . CGT . GCT . GGC . TAG . CTG . A..... STOP C K K K R A G TGT . AAA . AAA . AAC . GTG . CTG . GCT . AGC . TGA.... frameshift mutations cause (A) 9 (A) 8 repeat length (A) 10 C K K K V L A S STOP <ul><li>loss of function </li></ul><ul><li>generation of cancer specific peptides </li></ul>Slippage in coding microsatellites generates predictable novel peptides
ELISA reactivity reveals immune response to Neopeptides in Lynch syndrome patients with Previous cancer and in MMR mutation carriers patients healthy controls mutation carriers
Distribution of FSP-specific responses in patients and controls Schwitalle et al., Gastroenterology 2008
T-cells infiltrating MSI-H CRC recognize MSI-induced FSP Exemplary results derived from a MSI-H CRC-patient Schwitalle et al. Gastroenterology 2008
Truncating mutations in ß2m are more common with tumour progression but absent from mets Kloor et al Int J Cancer 2007 UICC IV, M1: No truncating mutation in CRC with metastasis ß2m mutation frequency is signif. lower in M1 vs M0 MSI CRC ?Intact ß2m needed for metastasis formation in MSI-H CRC UICC III 9/22 (40.9%) UICC IV 2/26 (7.7%) p=0.01 Only silent mutations or single AA exchanges, w/o functional loss of ß2m
ANVIL <ul><li>A nti N eopeptide V accine I n L ynch syndrome </li></ul><ul><li>Stimulate cell based immunity to a panel of neopeptides predicted in important “tumour” genes using replication deficient chimp adenovirus & MVA as vectors </li></ul><ul><li>Framework 7 proposal with Riccardo Cortese of Okairos ltd </li></ul>
Aspirin inhibits IL-4 making cancer stem cells more vulnerable to chemotherapy
summary <ul><li>CIN and MIN: a valuable basic division </li></ul><ul><li>Kras and BRAF: pivotal to targeting cetuximab and differentiating Lynch syndrome from sporadic MSI high tumours </li></ul><ul><li>CIMP: CpG island methylation phenotype is a marker of gene silencing </li></ul><ul><li>Epimutation: a rare cause of multiple cancers </li></ul><ul><li>“ trapped” Darwinian selection: B2M mutations allow local growth but suppress mets </li></ul><ul><li>Immunoprevention: a new approach to avoidance of MIN cancers </li></ul>