MMR / MSI Testing Guidelines

· indications for MMR / MSI testing

· CCO guideline Sep. 2015:

· Adjuvant therapy is recommended for a high-risk subset of patients with Stage II colon cancer, which includes:

· Nodes inadequately sampled

· T4 lesions

· Perforation at tumour site

· Poorly-differentiated histology in the absence of MSI

· When treated with adjuvant therapy, high-risk stage II patients should receive a fluoropyrimidine

· It would be reasonable to consider FOLFOX

· MMR/MSI testing should be performed for all stage II patients for who adjuvant chemotherapy is being considered

· MSI testing is not recommended for stage II disease in the absence of high-risk features

· Adjuvant chemotherapy with fluoropyrimidine monotherapy regimen following surgery in patients who have MSI is not recommended.

· For patients with high-risk stage II colon cancer and high MSI status, the choice of treatment is between observation and FOLFOX

· Completely resected stage III colon cancer should be offered adjuvant chemotherapy. The available treatment options are:

· FOLFOX or FLOX or XELOX

· Capecitabine

· 5-fluorouracil (5-FU) + leucovorin (LV)

· [DRAFT CAP / ASCO guidelines - 2015]:

· (see also colorectal carcinoma notes for the rest of the draft recommendations)

· [Draft recommendation]: dMMR/MSI testing must be performed in all colorectal cancers for prognostic stratification and identification of Lynch syndrome patients. [73% responses / comments agree, 11% disagree]

· [Draft recommendation ] : BRAF V600 mutational analysis in conjunction with dMMR/MSI testing must be performed in carcinoma tissue of patients with metastatic colorectal carcinoma for prognostic stratification

· Testing Algorithms:

· Lynch screening algorithm proposed by CCO:

· CRC patient < 70 y à IHC screen

· If intact, no further action required, if negative family history

· If deficient for PMS2, MSH2, or MSH-6 à genetic counselling

· If deficient for MLH1 à BRAF V600E testing

· If BRAF mutant, and negative family history, no further action required

· If BRAF wild-type à MLH-1 promoter methylation testing

1. If hyper-methylated, and negative family history, no further action required

2. If not hyper-methylated à genetic counselling

· While combined IHC and MSI testing is ideal, IHC testing alone may be more feasible for large-scale screening programs (geneReviews 2014)

· Testing for HNPCC – CCO criteria:

· (If a tumour sample is unavailable, germline testing may proceed on the youngest, living, affected individual from families meeting criteria 1 & 2 ONLY.)

· 1. Affected and unaffected individuals from families with a known HNPCC causing mutation.

· 2. Affected individuals whose families meet the Amsterdam criteria. The family must meet all of the following criteria:

· Three affected relatives with any combination of colorectal, endometrial, small bowel, ureter, transitional cell kidney cancer (urothelial), sebaceous adenoma/carcinoma and/or keratoacanthoma.

· One should be a first-degree relative of the other two.

· At least two successive generations should be affected.

· At least one diagnosis must be before age 50 years.

· Tumour type should be confirmed by review of pathology or other medical records.

· 3. Affected individuals from families with: Three affected individuals, one with colorectal cancer, and the other two with any combination of: colorectal, endometrial, small bowel, ureter, sebaceous adenoma/carcinoma, ovarian, pancreatic, kidney (transitional cell cancer only), gastric, primary brain or primary hepatobiliary cancer. ·

· Two of the three family members must be in a first-degree relationship. ·

· At least one diagnosis younger than 50 years of age. ·

· Familial adenomatous polyposis should be excluded. ·

· Tumours should be verified by pathological examination.

· 4. Individual affected with colorectal cancer (CRC) and a second primary HNPCC-associated cancer (as listed in #3). This includes synchronous and metachronous CRCs. At least one primary cancer must be diagnosed before 55 years of age. Families are eligible with or without family history of HNPCC-associated cancer, and tumours should be verified by pathological examination.

· 5. Individual diagnosed with CRC under the age of 35. Families are eligible with or without family history of HNPCC-associated cancer, and tumours should be verified by pathological examination.

· 6. One case of CRC before 50 years of age, with a first- or second-degree relative with one of the following HNPCC-related cancers diagnosed before 50 years of age: colorectal, endometrial, small bowel, ureter, urothelial, sebaceous adenoma/carcinoma or keratoacanthoma.

· 7. Individuals with immunodeficient tumours (regardless of family history) as follows MOAC – 3 Evidence Summary Evidence Summary – September 28, 2015 Page 27 ·

· MSH2-deficient tumour ± MSH6 deficiency (sequence and multiplex ligation-dependent probe-amplification technique [MLPA] of MSH2 gene only) ·

· MSH6 (only)-deficient tumour (sequence and MLPA of MSH6 gene only) ·

· MLH1-deficient tumour in individual younger than 60 years of age (sequence and MLPA of MLH1 gene only)

· Amsterdam Criteria:

· Three affected relatives with any combination of colorectal, endometrial, small bowel, ureter, transitional cell kidney cancer (urothelial), sebaceous adenoma/carcinoma and/or keratoacanthoma.

· One should be a first-degree relative of the other two.

· At least two successive generations should be affected.

· At least one diagnosis must be before age 50 years.

· Tumour type should be confirmed by review of pathology or other medical records.

· Bethesda:

· age < 50 y

· presence of synchronous, metachronous colorectal CA

· presence of other HNPCC-associated tumours

· MSI-H histology (see above) in a patient younger than 60 y

· Colorectal CA in 1 or more 1^st degree relatives with an HNPCC related tumour, one of the CA’s being diagnosed at < 50 y

· Colorectal CA in 2 or more 1^st degree relatives with an HNPCC-related tumour, regardless of age

· Tissue for testing:

· ?CCO:

· Adenocarcinoma

· Adenoma (less reliably identified in Lynch syndrome-rleated polyps – 79%)

· HG dysplasia is more likely to detect MMR deficient

· May be less reliable when performed on small tissue samples

· [Draft CAP / ASCO expert OPINION – 2015]:

· Primary carcinoma is acceptable for molecular marker testing (KRAS, eRAS, BRAF, dMMR/MSI)

· Metastatic tumour is also acceptable (preferable in patients with metastatic disease)

· FFPE is an acceptable specimen for molecular marker testing. Use of cytology specimens and other specimen types will require additional adequate validation

· MMR IHC (92% sensitivity for individuals with Lynch – Shia 2008) (range of 74-100% - CCO evidence review)

· Recommended methods:

· Correlation with MSI (PCR) testing:

· Effective testing method ofr identifying tumours that are MSI-H by PCR (CCO evidence review)

· Kappas ranged from 0.73 to 0.83 for studies examining selected populations. For studies examining unselected populations, with the exception of one study showing a 0.58 result, Kappas suggested excellent agreement between the two tests, with scores ranging from 0.81 to 0.95

· Using IHC for MLH1 and MSH2, MMR-D is highly concordant (>95%) with MSI-H using PCR (Genomic Health, referencing Bertagnolli MM et al. J Clin Oncol 2009)

· Effective testing method ofr identifying tumours that are MSI-H by PCR (CCO evidence review)

· Ability to detect LS and comparison to MSI (PCR):

· For studies examining unselected populations, with the exception of one study showing a 0.58 result, Kappas suggested excellent agreement between the two tests, with scores ranging from 0.81 to 0.95

· Overall, IHC was comparable to MSI in its ability to identify tumours that were LS for CRC patients, ranging from 74% to 100% for IHC versus 91% to 100% for MSI in selected populations, and 79% to 94% for IHC versus 84% to 100% for MSI in unselected populations. (CCO evidence review)

· For the most part, PPVs were comparable for IHC and MSI in both selected and unselected populations, with the exception of two studies having MSI PPVs much higher than those of IHC (MSI 75% versus IHC 59%); MSI 79% versus IHC 48% (CCO evidence review)

· IHC can identify MSH6 cases that may not show high MSI and, thus, can be missed by MSI testing

· In Ontario, experts at the 2011 symposium on hereditary gastrointestinal cancer, held at the Zane Cohen Centre at Mount Sinai Hospital in Toronto, focused on optimal approaches to screening for LS and reached unanimous agreement that MMR reflex IHC testing (MMR-IHC) is a viable screening option to detect LS. They concluded that testing by IHC for MLH1, MSH2, MSH6, and PMS2, should be performed on tumours from all patients with CRC or EC cancer who are younger than 70 years of age. (CCO evidence review)

· False negative results:

· It is possible that some missense germline variants will not result in the absence of a detectable protein product

· Not all relevant proteins are tested

· Reporting of IHC results:

· Any positive reaction in the nuclei of tumour cells is considered as intact expression (normal) (CAP protocol 2016)

· It is common for intact staining to be somewhat patchy (CAP protocol 2016)

· An interpretation of expression loss in tumour cells should be made only if a positive reaction is seen in internal control cells (e.g. nuclei of stromal, inflammatory, or non-neoplastic epithelial cells) (CAP protocol 2016)

· No MMR deficiency:

· Sporadic cancer (note: Lynch is not completely ruled out)

· MLH1 and (?often or always) PMS2 lost:

· Sporadic cancer (most likely) or germline MLH1 mutation

· Consider BRAF or methylation studies

· In absence of both BRAF and MLH1 methylation, germline MLH1 testing may be indicated

· PMS2 lost only:

· Germline PMS2 mutation likely, or germline MLH1 mutation

· MSH2 and MSH6 lost:

· Germline MSH2 mutation likely, or germline EPCAM mutation, or rarely germline MSH6 mutation

· MSH6 lost:

· Germline MSH6 mutation likely, or germline MSH2 mutation

· Bear in mind that nucleolar staining or complete loss of MSH6 staining has been described in colorectal cancer cases with prior radiation or chemotherapy, and a significant reduction in MSH6 staining has been described in a small percentage of colorectal carcinomas with somatic mutations of the coding region microsatellites of the MSH6 gene in MLH1/PMS2-deficient carcinomas (CAP protocol 2016)

· MLH1 lost only:

· Germline MLH1 mutation likely (?double check this)

· MSH2 lost only:

· Germline MSH2 mutation likely

· Validation of MMR IHC:

· Number of samples:

· The number of test samples that is required for test validation is determined by power analyses based on the proposed concordance rate and known characteristics of the calculation to be used and the expected “pass rate.” (CAP IHC best practices.pdf)

· (Class II Ab) As many as 50 to 100 samples may be required when validating a new antibody for a class II test.

· preferably by using 50% cases that are unequivocally positive and 50% cases that are the mixture of weakly positive and unequivocally negative.

· A much smaller number of samples may be sufficient for some class II tests (eg, CD117).

· Successful validation requirements:

· Concordance of 95% or more for positive and negative results (CAP IHC best practices.pdf)

· Concordance at this level usually parallels a κ value of 0.80 or more, or “perfect or near perfect” agreement with a reference laboratory or method.2 (CAP IHC best practices.pdf)

· [Draft CAP / ASCO strong recommendation – 2015]:

· Validated testing methods with sufficient performance characteristics for the intended clinical use. Molecular marker testing validation should follow accepted standards for clinical molecular diagnostics tests.

· Performance of molecular marker testing (and IHC testing) must be validated in accordance with best laboratory practices

·

· Ongoing QC of MMR IHC:

· [Draft strong recommendation CAP/ASCO 2015]:

· Labs must incorporate colorectal carcinoma molecular marker testing methods into their overall laboratory quality improvement program, establishing appropriate quality improvement monitors as needed to assure consistent performance in all steps of the testing and reporting process. In particular, labs performing colorectal carcinoma molecular markers testing must participate in a formal proficiency testing programs, if available, or an alternative proficiency assurance activity.

· MSI (PCR) testing (93% sensitivity for identifying individuals with a germline MMR gene variant – Shia 2008) (range of 75-100% - CCO evidence review)

· Test tumour tissue and normal tissue concurrently

· Panel of 5 microsatellite markers (recommended BAT25, BAT26, D2S123, D5S346, and D17S250)

· MSI-high: > 30 % of markers show instability

· MSI-low: < 30% of markers show instability

· MSI-stable: 0% of markers show instability

· Good to excellent agreement with MSI-H by PCR (CCO evidence review)

· although MSI testing maybe an effective alternative, IHC is less expensive and directs gene-specific germline mutation testing, thereby offering an overall cost savings

· IHC can identify MSH6 cases that may not show high MSI and, thus, can be missed by MSI testing

· Recommended methods

· Validation of MSI PCR:

· [Draft CAP / ASCO strong recommendation – 2015]:

· Validated testing methods with sufficient performance characteristics for the intended clinical use. Molecular marker testing validation should follow accepted standards for clinical molecular diagnostics tests.

· Performance of molecular marker testing must be validated in accordance with best laboratory practices

· Number of samples:

· Successful validation requirements:

· Ongoing QC of MSI PCR:

·

· MLH1 methylation analysis:

· MLH1 promoter methylation can help eliminate the diagnosis of Lynch

· However, it can occur as a “second hit” in an individual with Lynch

· Until more studies include germline testing on both negative and positive MLH1 promoter and/or BRAF V600E tumours we cannot, with confidence, determine their ability to triage potential LS patients for germline testing. (CCO evidence review)

· BRAF mutation

· Rare in Lynch syndrome-related cancers

· Good indicator for triaging MLH1-negative patients to germline MMR mutation testing (CCO evidence review)

· However, until more studies include germline testing on both negative and positive MLH1 promoter and/or BRAF V600E tumours we cannot, with confidence, determine their ability to triage potential LS patients for germline testing.

· rare cases of BRAF mutations can occur in patients with LS; thus, the committee recommended that all patients with an MLH1-deficient tumour should be referred for genetic counselling, regardless of BRAF results, if they are younger than 50 years or have a family history meeting the Ministry of Health and Long-term Care (MOHLTC) clinical testing criteria

· Not helpful in Endometrial cancers due to low incidence

· TAT:

· See [Draft CAP / ASCO guidelines] for TAT and send-out timelines.

References:

· Pollett A et al. [CCO Evidence Summary MOAC-3] Screening for Lynch Syndrome by Immunohistochemistry, BRAF mutations analysis, and MLH1 promoter methylation analysis for patients in Ontario with colorectal or endometrial cancers.

· [CCO Guideline] Adjuvant Systemic Chemotherapy for Stage II and III Colon Cancer Following Complete Resection
Guideline 2-29 Version 2: September 2015

· Genetests.org

· GeneReviews (May 2014)

· Genetics Home Reference (May 2013)

·