Transcript Bethesda Guidelines and MSI Testing

1
MICROSATILLITE INSTABILITY
IN CRC
HENRY T. LYNCH, MD
JANE F. LYNCH, BSN
Creighton University
School of Medicine
Omaha, Nebraska
2
Colorectal Cancer
Worldwide estimates for colorectal cancer during 2008*:
Incidence – 1,233,711
Mortality –
608,644
Worldwide estimates for familial/hereditary CRC during
2008*:
Lynch syndrome 3-5% of all CRC
37,011-61,686
FAP
<1% of all CRC
<12,337
Familial
20% of all CRC
246,742
*GLOBOCAN. The International Agency for Research
on Cancer web site. URL: http://www.iarc.fr/
3
Familial/Hereditary CRC in US
Annual CRC incidence in US: 142,570
Lynch syndrome 3-5% of all CRC
4,277 - 7,129
FAP
<1% of all CRC
<1,426
Familial
20% of all CRC
28,514
Jemal et al. CA Cancer J Clin 60:277-300,2010.
3
4
Magnitude of the Problem
Question: Why are these figures of such
significant public health impact?
Answer: Each hereditary cancer comes from a
family that could benefit immensely from genetic
counseling.
DNA testing, surveillance, and highly-targeted
management are the key!
Problem: Significance of family frequently missed!
5
Should we test all
colorectal cancer for
Lynch Syndrome?
YES! Test everybody.
6
Search for LS Among CRC Affecteds*
Evidence:
Among 500 CRC patients, 18 (3.6%) had LS.
Of these 18:
 18 (100%) had MSI-H CRCs;
 17 (94%) were correctly predicted by IHC;
 only 8 (44%) were dx < 50 years;
 only 13 (72%) met the revised Bethesda guidelines;
 1/35 cases of CRC show LS.
*Hampel et al. J Clin Oncol 26:5783-5788, 2008.
7
Molecular Genetic Screening for LS
Recommendation*:
All incident CRC and EC cases should be molecularly
screened for LS.
MSI highly sensitive (89.3%).
IHC equally sensitive (91.2%), is inexpensive, is more
readily available, and predicts the nonworking gene.
IHC is preferred method to screen for LS*.
*Hampel et al. J Clin Oncol 26:5783-5788, 2008.
8
Familial CRC
Familial clustering of CRC, like that for carcinoma of
the breast and stomach, has been discussed for
more than 100 years.
What does it mean from the standpoint of risk?
Best answer – First- degree relative of CRC
affected has 2-3 fold excess risk for CRC
compared to population expectations.
9
Cardinal Features of Lynch Syndrome
•
•
•
•
•
•
•
AD – MMR mutations
Proximal
Earlier age of onset
Accelerated carcinogenesis
Extra colonic cancers
Pathology – distinctive?
↑ survival
10
Increased risk for certain
extracolonic malignancies










Endometrial
Ovary
Stomach
Small bowel
Pancreas
Liver and biliary tree
Muir-Torre cutaneous features
Brain, (glioblastoma) – Turcot’s syndrome
Possible Prostate cancer and others
Breast cancer - controversial.
Cardinal Features of Lynch Syndrome
• Differentiating pathology features of LS CRCs:
- more often poorly differentiated;
- excess of mucoid and signet-cell features;
- Crohn’s-like reaction;
- medullary features;
- significant excess of infiltrating lymphocytes
within the tumor.
• Increased survival from CRC.
• Sine qua non for diagnosis is identification of
germline mutation in MMR gene (most commonly
MLH1, MSH2, MSH6) segregating in the family.
11
A
C
B
D
14
Bethesda Guidelines and MSI Testing
MSI-H findings are associated with a moderate
degree of uncertainty.
Such patients with MSI-H will merit referral for
further genetic assessment.
Centers of expertise recommend informed consent
for MSI testing.*
BRAF mutation, if present, will negate an LS
diagnosis.
*Palomaki et al. Genet Med 11:42-65, 2009.
15
Cost-effectiveness of DNA Testing
Estimate of the cost-effectiveness of
genetic testing strategies to identify LS
among newly dx CRC patients using MSI
and IHC is cost-effective.*
*Mvundura et al. Genet Med 12:93-104, 2010.
16
Amsterdam Criteria
AC-I (24)
 At least three relatives with histologically verified
colorectal cancer;
1. One is a first-degree relative of the other two;
2. At least two successive generations affected;
3. At least one of the relatives with colorectal
cancer diagnosed at <50 years of age;
4. Familial adenomatous polyposis (FAP) has
been excluded.
17
Amsterdam Criteria
~ 60% of Amsterdam + Families have
MSI-H and MMR mutations.
18
Familial CRC Type “X”
Amsterdam Criteria positive but lacking MSI and MMR
mutations will constitute ~ 40% of those AC-I without
MMR mutations and therein referred to as familial CRC
type X.*
1) CRC > left side
2)  CRC and extra colonic CRC
3) Later age CRC onset
4) Molecular genetics (MSI and IHC or MMR
mutation) ABSENT!
*Lindor et al. JAMA 293:1979-1985, 2005.
19
Epithelial Cell Adhesion Molecule (EPCAM)
Gene and Its Lynch Syndrome Connection*
A portion of this ~40% lacking MMR
mutations is caused by a mutation
mechanism in the gene known as
EPCAM.
*Kovacs et al. Hum Mutat 30:197-203, 2009.
5’ EPCAM deletion
Exons 8 and 9 and polyadenylation sequence
Polyadenylation Sequence
Transcriptional read through
Hypermethylation of the MSH2 promoter
Ligtenberg MJ, Nature Genetics 2009.
20
21
Why LS with Site-Specific CRC?
Deletion in EPCAM results in hypermethylation
and incomplete silencing of MSH2.
EPCAM mutation carriers may have phenotypic
features that differ from carriers of MSH2
mutations – namely, an almost exclusive
expression of site-specific CRC, thereby lacking
extracolonic cancers.
22
American and Dutch families have the same
deletion in the EPCAM gene
MSH2
EPCAM
Deletion
c.859-1462_*1999del
(4.9 kb, starting in intron 7
and including exons 8 & 9)
Lightenberg, Nature Genetics 2009.
23
American and Dutch EPCAM mutations
originate from a common ancestor
Family R and the Dutch families share a 6.1 MB region surrounding
the same EPCAM deletion indicating a common ancestor. Based on the size of the
shared region it is estimated the deletion occurred 10 generations ago.
Chromosome 2
Chromosome 2
Deletion and Region
inherited from
common ancestor
Dutch Families
Family R
24
History of Family R*
Ascertained by us in 1970 and followed continuously.
700 blood line relatives
327 individuals age ≥ 18, ≥ 25% pedigree risk
Phenotype strikingly similar to LS but integral extracolonic
cancers absent (site-specific CRCs)
*Lynch et al. Cancer 56:934-938, 1985.
Lynch et al. Cancer 56:939-951, 1985.
25
First patient identified with
EPCAM mutation
CRC affecteds
EPCAM results
26
MSI in CRC*
Microsatellite instability (MSI):
a) is hypermutable phenotype due to loss of DNA
MMR activity;
b) is present in ~ 15% of all CRCs;
c) ~ 3% associated with LS, remaining 12%
sporadic acquired hypermethylation of promoter
of MLH1 gene;
d) occurs in tumors manifesting the CpG island
methylator phenotype.
* Boland and Goel. Gastroenterology 138:2073-2087,
2010.
27
What Is MSI?*
Nearly all LS tumors differ from their sporadic
counterparts and most are characterized by MSI.
MSI mutations occur at short repetitive sequences
called microsatellites. Most are insertion or
deletion mutations and, as a consequence of
loss of DNA MMR activity.
*Boland & Goel. Gastroenterology 138:2073-2087,
2010.
28
Definitions of MSI’s Dx Criteria
These comprise BAT-25 and BAT-26
mononucleotides and D5S346, DS123,
and D12S250 dinucleotide repeats.
> 2 or more markers → MSI-H;
those with one unstable marker MSI-L;
those absent instability referred to as
MSS.
29
BRAFV600E Mutations in MSI*
The BRAFV600E mutation occurs exclusively in
sporadic forms of MSI CRC.
Combined analysis of MSI and BRAFV600E
mutation is included in current protocols of LS
since it is a reliable, fast, and low-cost strategy.
Helps identify sporadic cases and avoids timeconsuming and expensive screening of MMR
germline mutation analysis.
*Seruca et al. Expert Rev Gastroenterol Hepatol
3:5-9, 2009.
30
Algorithm
1. IHC on all colorectal patients on
tumor block;
2. If positive, BRAF (if positive, then
sporadic);
3. Only do full MMR genetic tests on
patient IHC +. BRAF neg.
MORPHOLOGY
SUSPICIOUS
FOR MSI-H
Run PCR test
for MSI status
Is there
MSI-H?
NO
FAMILIAL CRC
TYPE “X”
NO
YES
Is there loss
of staining
with any of
the Abs?
Run mutation analysis
for BRAF V600E
Is there
BRAF V600E
mutation?
NO EVIDENCE OF
LYNCH
SYNDROME
NO
IHC for MLH1,
MSH2, MSH6, PMS2
YES
PUTATIVE
LYNCH
SYNDROME
MMR GENES MUTATION
ANALYSIS
YES
SPORADIC CRC
WITH MSI-H
NO
Is there
a mutation in MMR
gene?
YES
Gatalica Z, Torlakovic E. Fam Cancer 2008;7:15-26
LYNCH
SYNDROME
31
MSI Analysis
A functional assay for
the MMR proteins
MSI Analysis
MSI High Data
New – Focus on Mononcleotides
37
MSI and Therapeutic Implications
Pharmacogenetics:
5-FU-based chemotherapy refractory in MSI
CRCs;
possible advantage of irinotecan-based therapy;
the latter “not ready for prime time” but 5-FU
approaching clinical acceptability.
Boland and Goel. Gastroenterology 138:2073-2087, 2010.
38
MSI and Therapeutic Implications*
Where did this knowledge originate?
Through DNA MMR genes first identified in
bacteria through exposure to cytotoxic mutagens
(alkylating agents) and selecting for strains
resistant to DNA damage.
*Carethers et al. J Clin Invest 98:199-206, 1996.
Ribic et al. N Engl J Med 349:247-257, 2003.
39
MSI and Therapeutic Implications*
Resulting bacteria were hypermutable and
resistant to DNA alkylation.
Resistant to cytotoxic agents that acted by
damaging DNA.
Raised possibility that DNA MMR deficient cells
might be relatively resistant to some types of
cytotoxic chemotherapy.
*Carethers et al. J Clin Invest 98:199-206, 1996.
Ribic et al. N Engl J Med 349:247-257, 2003.
40
MSI and Therapeutic Implications
Virtually all studies show either no benefit or
adverse effects in response to 5-FU-based
adjuvant chemotherapy (reviewed by Boland
and Goel.*)
In vitro responses suggest that chemoresistance is
seen for many chemotherapeutic agents.**
*Gastroenterology 138:2073-2087, 2010.
**Aebi et al. Cancer Res 56:3087-3090, 1996.
41
MSI and Therapeutic Implications*
Currently, guidelines do not recommend using MSI
status to determine whether or not to use
chemotherapy.
This recommendation merits a second look, given
the wealth of data showing the inadequacy of 5FU for CRC with MSI.
Should be tested only in the context of a
randomized clinical trial.
*Boland & Goel. Gastroenterology 138:2073-2087,
2010.
42
Immune Response in CRCs with MSI
CRCs with MSI demonstrate a host-immune
response characterized by tumor infiltrating
lymphocytes (TILs):
• may exert effects on tumor cell apoptosis and cell
proliferation;
• associated with improved disease-free survival;
*Sinicrope et al. Int J Cancer 120:1232-1238, 2007.
43
Immune Response in CRCs with MSI
(continued:)
• Apoptosis/proliferation ratios and reduced cell
proliferation;
• Phenotypic features of MSI-H tumors, associated
with ↑ TILs;
• Indicate an activated immune response that may
contribute to their favorable survival rates.
*Sinicrope et al. Int J Cancer 120:1232-1238, 2007.
44
Host Immune Response Characterized by
Tumor Infiltrating Lymphocytes (TILs)*
Summary:
Higher apoptosis/proliferation ratios and reduced
cell proliferation are phenotypic features of MSI-H
tumors that are associated with ↑ TILs;
Indicating an activated immune response, and may
contribute to their favorable survival rates.
*Sinicrope et al. Int J Cancer 120:1232-1238, 2007.
45
IHC and LS
IHC is an acceptable substitute for MSI, given
strong concordance rates of 94% between MSI
and IHC in CRC and EC.*
Allows gene-specific DNA analysis based on
staining pattern.
MMR protein deficiency occurs in 12-20% of
CRCs.**
*Palomaki et al. Genet Med 11:42-65, 2009.
**Hampel et al. J Natl Comp Cancer Net 8:597601, 2010.
46
Targeted CRC Screening
Screening is melded to LS’s natural history:
Proximal location  colonoscopy
Early age of onset  beginning at age 25
Accelerated carcinogenesis  every 1-2 yrs < age 40,
then annually
Pattern of extra-colonic cancers  targeted screening
47
MSI/IHC in CRC*
Conclusion:
CRCs with MSI may show distinctive clinical,
pathologic features:
a) predominance of CRC in proximal colon;
b) lymphocyte infiltration within tumor;
c) poorly differentiated mucinous or signet cell
appearance;
d) better prognosis;
e) differing response to chemotherapeutics;
f) molecular screening all CRC cases.
*Boland and Goel. Gastroenterology 138:2073-2087, 2010.