13, 2020 march after · 25 phd laboratory director 2/20/2020 8:03 am 26 geneticist 2/20/2020 2:37...

MMR and MSI Testing in Patients Being Considered for Checkpoint Inhibitor Therapy: DRAFT

Recommendations

1 / 54

63.11% 154

2.87% 7

17.21% 42

2.05% 5

0.82% 2

4.10% 10

10.25% 25

2.87% 7

11.48% 28

Q1 What is your occupation/role? (select all that apply)Answered: 244 Skipped: 0

Total Respondents: 244

Pathologist

Physician(non-patholo...

MedicalDirector

Technologist/Technician

QA/QCCoordinator

LaboratoryManager

Industry

PatientAdvocate

Other (pleasespecify)

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

ANSWER CHOICES RESPONSES

Pathologist

Physician (non-pathologist)

Medical Director

Technologist/Technician

QA/QC Coordinator

Laboratory Manager

Industry

Patient Advocate

Other (please specify)

DisclaimerThe information, data, and draft recommendations provided by the College of American Pathologists are presented for informational and public feedback purposes only. The draft recommendations and supporting documents will be removed on March 20, 2020.The draft recommendations along with the public comments received and completed evidence review will be reassessed by the expert panel in order to formulate the final recommendations. These draft materials should not be stored, adapted, or redistributed in any manner.

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

2 / 54

# OTHER (PLEASE SPECIFY) DATE

1 BloodPAC Consortium Executive Director 3/13/2020 3:08 PM

2 molecular genetics fellow 3/13/2020 2:49 PM

3 Molecular Diagnostic Laboratory Director 3/13/2020 12:04 AM

4 Scientific Director, Molec Pathology 3/12/2020 4:53 PM

5 Scientist 3/11/2020 7:07 AM

6 Statistician 3/10/2020 10:10 AM

7 Scientist 3/9/2020 10:20 PM

8 Patient 3/9/2020 7:16 AM

9 Genetic counselor 3/8/2020 5:04 PM

10 Geneticist 3/8/2020 2:05 PM

11 genetic counselor 3/6/2020 10:37 AM

12 Laboratory Director (non-physician) 3/6/2020 8:43 AM

13 Laboratory Director, Clinical Molecular Geneticist 3/2/2020 10:30 AM

14 Genetic Counselor 3/2/2020 9:12 AM

15 Clinical Reference Laboratory Director 2/28/2020 12:43 PM

16 Med 2/26/2020 11:55 PM

17 Genetic Counselor 2/26/2020 8:56 AM

18 xxx 2/24/2020 6:32 AM

19 s 2/21/2020 7:33 AM

20 Staff Scientist 2/20/2020 2:47 PM

21 ABMG molecular geneticist 2/20/2020 2:18 PM

22 Laboratory scientist 2/20/2020 1:58 PM

23 a 2/20/2020 12:43 PM

24 . 2/20/2020 12:04 PM

25 PhD laboratory director 2/20/2020 8:03 AM

26 Geneticist 2/20/2020 2:37 AM

27 Laboratory Director 2/19/2020 6:59 PM

28 Computational biologist 2/19/2020 4:19 PM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

3 / 54

Q2 Which of the following best describes your practice setting? (selectone)

Answered: 244 Skipped: 0

Universityhospital/aca...

Voluntary,non-profit...

Proprietaryhospital

City/County/State hospital

Veteranshospital

Army/AirForce/Navy...

National/corporate laboratory

Regional/localindependent...

Public Health,non-hospital

Clinic, group,or doctor...

Industry orvendor

PatientAdvocacy...


0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

4 / 54

44.26% 108

8.61% 21

2.46% 6

5.74% 14

0.82% 2

0.41% 1

5.33% 13

4.10% 10

0.41% 1

4.51% 11

18.44% 45

1.64% 4

3.28% 8

TOTAL 244


1 Liquid Biopsy Stakeholder Consortium 3/13/2020 3:08 PM

2 National Institute of Health?National Cancer Institute 2/21/2020 11:12 AM

3 s 2/21/2020 7:33 AM

4 Reference lab for healthcare system (non-profit) 2/21/2020 7:31 AM

5 a 2/20/2020 12:43 PM

6 Health System 2/19/2020 6:59 PM

7 independent pathologist owned lab 2/19/2020 6:58 PM

8 Public health academic university affiliated community hospital 2/19/2020 2:42 PM


University hospital/academic medical center

Voluntary, non-profit hospital

Proprietary hospital

City/County/State hospital

Veterans hospital

Army/Air Force/Navy hospital

National/corporate laboratory

Regional/local independent laboratory (except clinic or group practice and not owned by a national corporation(s))

Public Health, non-hospital

Clinic, group, or doctor office laboratory

Industry or vendor

Patient Advocacy Organization



Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

5 / 54

67.39% 93

18.84% 26

7.97% 11

5.07% 7

Q3 Draft Recommendation Statement 1In CRC patients being consideredfor checkpoint blockade therapy, pathologists should use MMR IHC and/orMSI by PCR for the detection of DNA mismatch repair defects. Although

MMR IHC or MSI by PCR are preferred, pathologists may use a validatedMSI by NGS assay for the detection of DNA mismatch repair defects.

Note: MSI by NGS assay must be validated against MMR IHC or MSI byPCR and must show equivalency.(Quality of Evidence: Moderate;

Strength of Recommendation: Strong)Answered: 138 Skipped: 106

TOTAL 138

Agree aswritten

Agree withsuggested...

Disagree(please incl...

Neutral

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Agree as written

Agree with suggested modifications (please include comments)

Disagree (please include comments)

Neutral


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

6 / 54

# COMMENTS DATE

1 It is not clear why IHC or MSI by PCR are preferred for detection of MMR deficiency. In point offact, an NGS assay that interrogates hundreds of microsatellite loci may in fact yield betterperformance than a PCR kit testing 5 MSI loc by PCR. Furthermore, given the additionalmolecular alterations that often need to be interrogated in high stage CRC, a single NGS assaythat can test KRAS/BRAF/NRAS status, MMR status, ERBB2 amplification and TRKrearrangement status may be preferable to individual assays testing for all of these alterationsseparately.

3/13/2020 8:14 PM

2 While good quality IHC has shown a high degree of concordance with PCR for MSI, theimplementation of IHC in the real world setting has shown a higher degree of variability inseveral studies. I suggest that NGS should be validated against PCR to avoid misinterpretationagainst IHC cases with low levels of staining being interpreted incorrectly as negative.

3/13/2020 3:47 PM

3 Currently, tissue-based biomarker testing has been utilized for the diagnosis and classificationof CRC and remains the gold standard for MMR/MSI testing. However, tissue testing has itslimitations. Difficulty obtaining sufficient quantity of sample for biomarker testing, samplingbiases due to tumor heterogeneity, and tumor quality due to necrosis can impair testingresults(1-6). Additionally, patients may choose to not undergo biopsy or be medically unfit. Inorder to not limit patient access to pembrolizumab through tissue only MSI testing, utilizingcfDNA/ctDNA biopsies as an alternative may be an option for many patients. A study hasshown that plasma MSI testing with a limit of detection of 0.1% tumor content can accuratelydetected 87% (71/82) of tissue MSI-H and 99.5% of tissue microsatellite stable (863/867) for anoverall accuracy of 98.4% (934/949) and a positive predictive value of 95% (71/75) whencomparing to solid tumor IHC and PCR methods(7). Additionally, three companies are pursuingFDA approval on their plasma based assays that measure MSI, FoundationOne™ Liquid(Foundation Medicine, MA, USA), PGDx elio™ (Personal Genome Diagnostics, MD, US) andGuardant360® assay (Guardant Health, CA, USA) (8-10). We do not believe that liquid biopsiesshould be used as the primary sources for MSI status, but as an alternative when other optionshave been exhausted. The National Comprehensive Cancer Network states for cfDNA/ctDNAtesting in Non-Small Cell Lung Cancer: “ cell-free/circulating tumor DNA testing should not beused in lieu of a histologic tissue diagnosis” and that “studies have demonstrated cell-freetumor DNA testing to generally have very high specificity, but significantly compromisedsensitivity, with up to a 30% false-negative rate”(11). Additionally, the NCCN guidelines forbreast cancer similarly state, “if liquid biopsy is negative, tumor tissue testing isrecommended”(12). Considering MSI is a pan-caner biomarker, it’s reasonable to extrapolatesimilar guardrails for testing in CRC. We believe that access to checkpoint blockade therapythrough MSI testing, should not be limited for those who have the available tumor for testing.We support the inclusion of utilizing NGS or PCR liquid biopsy assays that have beenorthogonally validated as a potential option for patients that don’t have tumor available fortesting. Comments and Changes: In CRC patients being considered for checkpoint blockadetherapy, pathologists should use MMR IHC and/or MSI by PCR for the detection of DNAmismatch repair defects. Although MMR IHC or MSI by PCR are preferred, pathologists mayuse a validated MSI by NGS assay for the detection of DNA mismatch repair defects. Note: MSIby NGS assay must be validated against MMR IHC or MSI by PCR and must showequivalency. ADD TO GUIDANCE: in cases where patient have limited tumor for testing ormedically unfit for invasive tissue sampling, cell free/circulating tumor DNA assay results thathave been orthogonally validated, may be used. ADD TO NOTE: ctDNA/ctDNA assays havebeen known to have high specificity but comprised sensitivity. False negative results arecommon with cfDNA/ctDNA assays and should be followed up with tissue testing if possible.cfDNA/ctDNA testing should not be used in lieu of tissue testing and should be orthogonallyvalidated against IHC or PCR tissue assays. 1. Vanderlaan PA, Yamaguchi N, Folch E, et al.Success and failure rates of tumor genotyping techniques in routine pathological samples withnon‐small‐cell lung cancer. Lung Cancer. 2014;84:39‐44. [PMC free article] [PubMed] [GoogleScholar] 2. Ellis PM, Shepherd FA, Millward M, et al. Dacomitinib compared with placebo inpretreated patients with advanced or metastatic non‐small‐cell lung cancer (NCIC CTG BR.26):a double‐blind, randomised, phase 3 trial. Lancet Oncol. 2014;15:1379‐1388. [PubMed][Google Scholar] 3. Popper HH. Commentary on tumor heterogeneity. Transl Lung Cancer Res.2016;5:433‐435. [PMC free article] [PubMed] [Google Scholar] 4. De Mattos‐Arruda L, WeigeltB, Cortes J, et al. Capturing intra‐tumor genetic heterogeneity by de novo mutation profiling ofcirculating cell‐free tumor DNA: a proof‐of‐principle. Ann Oncol. 2014;25:1729‐1735. [PMC freearticle] [PubMed] [Google Scholar] 5. Lebofsky R, Decraene C, Bernard V, et al. Circulatingtumor DNA as a non‐invasive substitute to metastasis biopsy for tumor genotyping andpersonalized medicine in a prospective trial across all tumor types. Mol Oncol. 2015;9:783‐790.

3/13/2020 3:15 PM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

7 / 54

6. Murtaza M, Dawson SJ, Pogrebniak K, et al. Multifocal clonal evolution characterized usingcirculating tumour DNA in a case of metastatic breast cancer. Nat Commun. 2015;6:8760 7.Willis, J., & Kopetz, S. (2019). Validation of Microsatellite Instability Detection Using aComprehensive Plasma-Based Genotyping Panel. Precision Medicine and Imaging, 7035–7045. 8. Foundation Medicine Press Release. Foundation Medicine’s new liquid biopsy assaygranted Breakthrough Device Designation by U.S. Food and Drug Administration. April 26,2018. Available online: http:// investors.foundationmedicine.com/news-releases/newsrelease-details/foundation-medicines-new-liquid-biopsyassay-granted 9. 9 . Personal GenomeDiagnostics Press Release. Personal Genome Diagnostics’ PGDx elio™ plasma resolveReceives Breakthrough Device Designation from FDA. July 2018. Available online:http://www.personalgenome. com/wp-content/uploads/2018/07/Personal-GenomeDiagnostics-PGDx-elio-plasma-resolve-ReceivesBreakthrough-Device-Designation-from-FDA.pdf 10.Guardant Health Press Release. The Guardant360® Assay Receives Expedited AccessPathway Designation for Breakthrough Devices from FDA. February 15, 2018. Available online:https://www.prnewswire.com/newsreleases/the-guardant360-assay-receives-expedited-accesspathway-designation-for-breakthrough-devices-fromfda-300599629.html 11. NCCNClinical Practice Guidelines in Oncology, Non-Small Cell Lung Cancer, Version 3.2020-February,2020 12. NCCN Clinical Practice Guidelines in Oncology, Breast, Version 3.2020-March,2020

4 Given the demonstrated high concordance with IHC and PCR based assays, therecommendation should give equal weight to NGS based assays for MSI testing. Given that allvalidated assays must demonstrate equivalency to the gold standard NGS assays it seemsdiscriminatory to not give the full recommendation to NGS assays. Additionally because of thepractice guidelines to also test for specific mutations in CRC, NGS assays would be the bestchoice for limiting the amount of tissue needed.

3/13/2020 2:57 PM

5 Recommend clarification that the lab performing the MSI by NGS needs to be the labconducting the validation study against MMR IHC or MSI by PCR.

3/13/2020 1:56 PM

6 The proposed recommendations do not specify which MSI by PCR Panel should be used. Westrongly suggest that the guidelines align with the revised Bethesda guidelines and recommenda five mononucleotide panel including BAT25, BAT26, MONO27, NR21, and NR24 instead ofthe original Bethesda panel, including BAT-25, BAT-26, D2S123, D5S346, and D17S250, sincethe mononucleotide markers have become the gold standard for MSI by PCR analysis by aseveral large molecular diagnostic laboratories across the world [Svrcek et al. (2019) Bull.Cancer 106(2): 119-128; Baudrin et al. (2018) Front. Onco. 8: 621]. There is significantevidence that shows that the original NCI/Bethesda microsatellite panel for evaluation of MSIunderestimates the number of MSI-H tumors and overestimates MSI-L tumors due to the use ofthree dinucleotide repeats. The use of mononucleotide markers exclusively in MSI testingimproves the sensitivity of the panel [Umar et al. (2004) J Natl Cancer Inst 96(4):261-8]. Testingfor MMR by IHC has been shown to have a false negative rate of 5-10% [Funkhouser et al.(2012) J. Mol Diagn 14:91-203; Dudley et al. (2012) Clin Cancer Res 22 813-20]. We thereforebelieve that due to the complementary nature of MMR IHC and MSI by PCR technologies andthe potential impact of misdiagnosis, there is a growing recognition in the field that these testsshould be performed together for maximal sensitivity when identifying patients for hereditarycancer risk and immunotherapy eligibility [Goodfellow et al. (2015) J Clin Oncol 33:4301-8;Leenen et al.(2012) Gynecologic Oncology 125(2): 414-20; Mann & Cheng (2020) Exp Reviewof Anticancer Therapy 20(1):1-4; Silva et al (2019) Ann Transl Med 7(21):600]. We also havesignificant concerns about recommending MSI by NGS since that technology is still fraught witha number of challenges. MSI by NGS is dependent upon the availability of expensiveequipment and a bioinformatics infrastructure whereas both MMR by IHC and MSI by PCR arereadily accessible for most pathology and molecular laboratories and are sufficient for mostdiagnostic purposes. Lack of a common standard (a set of markers, cut-offs etc.) furtherhampers use of MSI by NGS for routine diagnostic testing. There is very little evidence tosupport the clinical equivalency of MSI by NGS. Studies have shown that up to 36% of LynchSyndrome tumors were missed using thousands of markers by NGS [Latham et al (2019)J ClinOnc 37(4): 286-295]]. Currently there are no FDA-approved molecular tests available. Wewould strongly suggest recommending an FDA-approved molecular test when one is available.

3/13/2020 12:57 PM

7 A recent assessment of guideline aligned biomarker testing in CRC demonstrated suboptimaltesting rates despite the prognostic and predictive implications. Specifically, the current testingrates for the National Comprehensive Cancer Network(NCCN) recommended biomarkers RAS,BRAF, and MSI/MMR were 41%, 43%, and 51% respectively. (11) Of note, the NCCN colonand rectal guidelines currently recommend next-generation sequencing (NGS) as validmethodology for assessing MSI in addition to PCR testing, “especially in patients with

3/13/2020 11:43 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

8 / 54

metastatic disease who require genotyping of RAS and BRAF” (1)(2). The ability tosimultaneously detect multiple biomarkers as part of a comprehensive genomic profiling (CGP)assay has the potential to inform on biomarkers with positive predictive value for availabletargeted therapies and immunotherapies, reduce exposure to expensive and ineffectivetherapies, resulting in improved patient outcomes (11). Insufficient tissue has been identified asa key barrier to MSI/MMR testing, (12). CGP has the potential to mitgate tissue exhaustionassociated with iterative testing. Additionally, the European Society for Medical Oncology(ESMO) guidelines for MSI testing state that, “molecular tests guarantee the highest values ofspecificity and sensitivity in MSI testing” and that “NGS represents another type of moleculartests to assess MSI”, in addition to PCR methods (3). NGS has shown to be equivalent insensitivities and specificities when comparing to IHC and PCR for colorectal cancers. • Instudies comparing performance of NGS to PCR for MSI detection in samples including CRCspecimens, NGS performance was comparable to that of PCR with sensitivity estimates of ≥96.4% and specificity estimates of ≥ 97.2% reported across 3 different sized panels (4). • In apan-cancer study (including 2,315 colon cancer specimens) evaluating MSI across 18 cancertypes, 5930 cancer exomes were examined. Among 2,315 specimens with MSI statusmeasured by NGS method and PCR, accuracy of the NGS method was 96.6% (5). • In a pan-cancer study (including 1193 colorectal cancer specimens) analyzing MSI status, NGSperformance was comparable to that of MSI PCR with sensitivity of the NGS based method as100% and specificity as 99.9% (6). • In a validation study analyzing MSI status, results were97% (70/72) concordant with PCR/IHC-based MSI testing and 95% (70/74) concordant if MSI-intermediate samples were included and considered discordant. (7). • In a pan-cancer studyevaluating MSI statuses of 12,288 advanced solid cancers, validation against MSI PCR and/orMMR IHC performed for 138 CRCs and 40 uterine endometrioid cancers (UECs) showed aconcordance of 99.4% .(8). • In a validation of computational determination of MSI status usingwhole exome sequencing data in CRC patients, found there was 100% agreement betweenNGS and PCR methods in 130 patients (18 patients microsatellite instable patients) .(9). • In avalidation study looking at MSI status concordance between MSI NGS and PCR, of the 100cases being tested (73 CRC cases), there was 98% concordance. (10). NGS is well establishedin CRC literature with extensive pan-cancer studies that have demonstrate that validated NGS-based methods can accurately determine MSI-H/dMMR status and overcome limitations oftumor site or type. Additional therapy selection biomarkers have been approved for use in CRCsuch as BRAF, RAS and NTRK. MSI testing should therefore be considered part of acomprehensive genomic profiling strategy for patients where additional biomarker testing isneeded in late stage patients and tumor sample is limited. Continuity across guidelines isessential for patient management. Both NCCN and ESMO guidelines consider NGS a validtype of molecular test to assess MSI. NGS has demonstrated concordance across multipleplatforms/assays with IHC and PCR techniques and therefore CGP by NGS should be aprimary method recommended for detecting MSI status. We recommend the section bechanged to read as follows: "In CRC patients being considered for checkpoint blockadetherapy, pathologists should use MMR IHC and/or MSI by PCR or NGS for the detection ofDNA mismatch repair defects or microsatellite instability. Note: MSI by NGS assay must bevalidated against MMR IHC or MSI by PCR and must show equivalency." 1. NCCN ClinicalPractice Guidelines in Oncology, Colon Cancer, Version 2.2020-March,2020https://www.nccn.org/professionals/physician_gls/pdf/colon.pdf 2. NCCN Clinical PracticeGuidelines in Oncology, Rectal Cancer, Version 2.2020-March,2020https://www.nccn.org/professionals/physician_gls/pdf/rectal.pdf 3. C. Luchini, F. Bibeau, M. J.Ligtenberg, N. Singh, A. Nottegar, T. Bosse, R. Miller, N. Riaz, J.-Y. Douillard, F. Andre and A.Scarpa, "ESMO recommendations on microsatellite instability testing for immunotherapy incancer, and its relationship with PD-1/PD-L1 expression and tumour mutational burden: asystematic review-based approach," Ann Oncol, vol. 30, pp. 1232-1243, 2019.https://www.annalsofoncology.org/article/S0923-7534(19)31269-4/fulltext 4. S. J. Salipante, S.M. Scroggins, H. L. Hampel, E. H. Turner and C. C. Pritchard, "Microsatellite instabilitydetection by next generation sequencing," Clin Chem, vol. 60, pp. 1192-1199, 2014.https://jmd.amjpathol.org/article/S1525-1578(15)00153-1/fulltext 5. R. J. Hause, C. C. Pritchard,J. Shendure and S. J. Salipante, "Classification and characterization of microsatellite instabilityacross 18 cancer types," Nat Med, vol. 22, pp. 1342-1350, 2016.https://www.nature.com/articles/nm.4191 6. A. Vanderwalde, D. Spetzler, N. Xiao, Z. Gatalicaand J. Marshall, "Microsatellite instability status determined by next-generation sequencing andcompared with PD-L1 and tumor mutational burden in 11,348 patients," Cancer Medicine, vol.7, no. 3, pp. 746-756, 2018. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5852359/ 7. S. E.Trabucco, K. Gowen, S. L. Maund, E. Sanford, D. A. Fabrizio, M. J. Hall, E. Yakirevich and etal, "A novel next-generation sequencing approach to detecting microsatellite instability and pan-tumor characterization of 1000 microsatellite instability-high cases in 67,000 patient samples," J

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

9 / 54

Mol Diagn, vol. 21, pp. 1053-1066, 2019. https://jmd.amjpathol.org/article/S1525-1578(19)30358-7/fulltext 8. S. Middha, L. Zhang, K. Nafa, G. Jayakumaran, D. Wong, H. R.Kim, J. Sadowska, M. F. Berger, D. F. Delair, J. Shia, Z. Stadler, D. S. Kimstra, M. Ladanyi, A.Zehir and J. F. Hechtman, "Reliable pan-cancer microsatellite instability assessment by usingtargeted next-generation sequencing data," JCO Precision Oncol, vol. 1, pp. 1-17, 2017.https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6130812/pdf/nihms-984169.pdf 9. Johansen,A.F.B., Kassentoft, C.G., Knudsen, M. et al. Validation of computational determination ofmicrosatellite status using whole exome sequencing data from colorectal cancer patients. BMCCancer 19, 971 (2019). https://doi.org/10.1186/s12885-019-6227-7https://bmccancer.biomedcentral.com/articles/10.1186/s12885-019-6227-7 10. S. Pabla, J.Andreas, F.L. Lenzo, B. Burgher, J. Hagen, V. Giamo, M.K. Nesline, Y. Wang , M. Gardner, J.M.Conroy, A. Papanicolau-Sengos, C. Morrison, S.T. Glenn, “Development and analyticalvalidation of a next-generation sequencing based microsatellite instability (MSI) assay”,Oncotarget. (2019) 10(50):5181-5193 Abstract retrieved fromhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC6718258/https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6718258/pdf/oncotarget-10-5181.pdf 11. M. E.Gutierrez, K.S. Price, R.B. Lanman, R.J. Nagy, I. Shah, S. Mathura, M. Mulcahy, A.D. Norden,S.L. Goldberg, Genomic Profiling for KRAS, NRAS, BRAF, “Microsatellite Instability, andMismatch Repair Deficiency Among Patients With Metastatic Colon Cancer”, JCO PrecisionOncology 2019 :3, 1-9 https://ascopubs.org/doi/pdf/10.1200/PO.19.00274 12. Eriksson, J.,Amonkar, M., Al-Jassar, G., Lambert, J., Malmenäs, M., Chase, M., Sun, L., Kollmar, L., &Vichnin, M. (2019). Mismatch Repair/Microsatellite Instability Testing Practices among USPhysicians Treating Patients with Advanced/Metastatic Colorectal Cancer. Journal of clinicalmedicine, 8(4), 558. https://doi.org/10.3390/jcm8040558https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6518162/

8 Suggested Modification: In CRC patients being considered for checkpoint blockade therapy,pathologists should use MMR IHC and/or MSI by PCR and/or MSI by validated NGS assay forthe detection of DNA mismatch repair defects. Note: MSI by NGS assay must be validatedagainst MMR IHC or MSI by PCR and must show equivalency Additional Comments: Therecommendation for the utilization of IHC or PCR for the detection of mismatch repair defectsover NGS is poorly supported by the evidence utilized by the committee for drafting of theguidelines. Specifically, the literature supports that NGS has high concordance with PCR andIHC, as evidenced by references 25, 38, 45, 55, 59, 60, 75, 81, and 95 used in drafting theseguidelines (https://documents.cap.org/documents/mmrmsi_references_ocp.pdf). Therecommendation recognizes the acceptability of using a validated NGS assay, as such thereshould not be a recommendation to use one validated assay over another. In the setting of highconcordance between all three tests, the recommendation should be that IHC, PCR, and/orNGS may be used for the detection of DNA mismatch repair defects. Current NCCN guidelinesrecommend determination of MMR or MSI for all metastatic CRC patients, if not previouslydone. Specifically, these guidelines state that MSI may be accomplished with a validated NGSpanel, especially in patients with metastatic disease who require genotyping of RAS and BRAF.Patients with mCRC should have RAS, BRAF, and HER2 amplification testing, eitherindividually or as part of an NGS panel. (NCCN Colon Cancer Guidelines v1.2020 published12-19-19 and NCCN Rectal Cancer Guidelines v.1.2020 published 12-19-19).

3/13/2020 11:36 AM

9 1. Recommend the use of a FDA-authorized panel for MMR IHC 2. Remove the vendor-specificreference to Promega in the initial statement

3/13/2020 10:02 AM

10 The recommendation for the utilization of IHC or PCR for the detection of mismatch repairdefects preferably to NGS is not supported by the evidence utilized by the committee for thedrafting of the guidelines. Specifically, the literature supports that NGS has high concordancewith PCR and IHC, as evidenced by references used in drafting these guidelines. Theacceptability of using NGS is in the Note, which states that the assay must be validated. For avalidated assay, there is not strong evidence for the use of one method over the others. In thesetting of high concordance between all three tests, the recommendation should be that IHC,PCR, and/or NGS may be used for the detection of DNA mismatch repair defects. A strongrecommendation to prefer IHC and PCR conflicts with the moderate quality of evidence and thehigh concordance between IHC, PCR, and NGS. The note that NGS testing must be validatedagainst IHC or MSI testing is redundant as any clinical assay requires validation. In addition, forsome samples, using a single test is preferable for both laboratory workflows and tissuepreservation.

3/13/2020 9:22 AM

11 We especially like the "and" portion of MMR IHC and PCR/NGS being allowed. 3/13/2020 9:10 AM

12 The definition of PCR methods to be used for the detection of DNA mismatch repair defects 3/13/2020 5:05 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

should be not limited to the standard NIH panel or Promega MSI panel but broadened toinclude any validated molecular method, e.g. other PCR methods like the Idylla™ MSI Assay.The Idylla™ MSI Assay has shown excellent concordance to other molecular methods as wellas IHC (References: Li et al. (2019) Clin Colorectal Cancer, pii: S1533-0028(19)30086-6;Samaison (2019) Clin Pathol. 72(12):830-835; Maloney et al. (2018) TT054, J. Mol. Diagn.20(6), 895–1039; Maertens et al. (2017) Annals of Oncology 28 (suppl_5): v22-v42; De Craeneet al. (2018) Journal of Clinical Oncology 36:15_suppl,e15639; De Craene B. et al. (2018)Annals of Oncology 29 (suppl_8): viii14-viii57; De Craene et al. (2017) Annals of Oncology 28(suppl_5): v209-v268; Nicka et. al. (2018) J. Mol. Diagn. 20(6), 895–1039; Nafa. et.al. (2018) J.Mol. Diagn. 20(6), 895–1039), is used in a large number of laboratories (> 40 in the US) andhas been validated and approved for clinical use at MSKCC by the New York State Departmentof Health Clinical Laboratory Evaluation Program (project ID: 68841;https://www.wadsworth.org/print/82410). We propose to change the first paragraph as follows:Mismatch repair (MMR) Immunohistochemistry (IHC) and microsatellite instability (MSI)-polymerase chain reaction (PCR; defined as standard NIH panel of 5-7 microsatellites or othervalidated PCR-methods such as Promega MSI panel and Idylla MSI panel; most clinicallaboratories in the US are using one of these panels).

13 I think "preferred" is too strong a word. MSI by NGS may be comparable or even better giventhat in can interrogate thousand of loci. Also likely, it will become more accessible in the future.We should not preference towards one method over the others, but recommend that all arelegitimate if validated.

3/12/2020 7:19 PM

14 Laboratories have been using MMR IHC and MSI PCR techniques for decades. Reviewing thereferences attached to these draft guidelines most do not apply today because: 1. IHC prior to2008 mostly involved the use of MLH1 and MSH2 immunostains alone. 2. References vary intheir definition of negative stains and most do not include a definition. 3. Many references useddinucleotide biomarkers in their PCR assays. I would like to see a list of references, which arerestricted to those interrogating mismatch repair using 4 IHC MMR stains and MSI-PCR withmononucleotide markers. With the publication of these guidelines, I would like the inclusion oftesting standards: IHC: MLH1, PMS2, MSH2 and MSH6, with definition of loss MSI-PCR: Panelof at least 5 mononucleotide biomarkers, with >30% indicating the MSI-H phenotype. Currentpublications with optimal IHC and MSI methodologies indicate concordance to be 93-97%.Each methodology has different false negatives and as a result, I would like to see emphasisplaced on co-testing. Given the remarkable patient outcomes seen with checkpoint inhibitortherapy, each patient should have comprehensive co-testing available. MSI by NGS:Recommending NGS testing by NGS is premature because as reflected in the list ofpublications submitted there is no standard approach. At the Promega workshop AMP 2019,NGS MSI data showed sensitivities of 86% compared to MSI-PCR 97%. If NGS is to beincluded, it should be in a co-testing setting with MSI-PCR.

3/12/2020 4:58 PM

15 IHC should be mainstay for most community type practices due to little tissue required, sameday answers, availability. PCR can be added as supplement if available. NGS is generallyirrelevant in this setting. See attached letter.

3/10/2020 3:36 PM

16 In CRC patients being considered for checkpoint blockade therapy, pathologists should useMMR IHC and/or MSI by PCR for the detection of DNA mismatch repair defects. Although MMRIHC and/or MSI by PCR are preferred, pathologists may use a validated MSI by NGS assay forthe detection of DNA mismatch repair defects as second option.

3/9/2020 9:31 PM

17 Should the MMR IHC and MSI PCR be done simultaneously or sequentially? 3/9/2020 8:01 PM

18 The definition of PCR should be expanded. To lock down a set of standard NIH panel targets orthe Promega MSI panel will prevent incorporation of novel targets that may prove to beclinically valuable. Further, technological innovations such as ddPCR may add additional value.

3/9/2020 4:40 PM

19 Would recommend preferring MMR IHC over MSI by PCR/NGS for two reasons - MMR IHCcan use substantially less tissue for biopsies/small samples and also the pattern of MMR IHCdeficiency can provide information with regard to next steps for lynch syndrome testing (i.e.which genes should be sequenced first, should hypermethylation be definitively excluded viaseparate assay, etc.)

3/9/2020 12:23 PM

20 Agree in general, especially the "MMR IHC and/or MSI by PCR" (emphasis on "and/or" asopposed to "or"); however, it might be useful to clarify that since neither assay has a 100%sensitivity, labs that perform only one assay consider testing with an orthogonal assay when atest is negative.

3/6/2020 8:33 AM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

11 / 54

21 In CRC patients being considered for checkpoint blockade therapy, pathologists should useMMR IHC, MSI by PCR and/or validated MSI by NGS assayfor the detection of DNA mismatchrepair defects. Note: MSI by NGS assay must be validated against MMR IHC or MSI by PCRand must show equivalency.

3/2/2020 10:47 AM

22 Should offer PD-L1 testing as an alternative and bypass MMR MSI testing. 2/29/2020 12:54 AM

23 MSI-NSG is depended on the availability of the NGS platform for each lab. MMR-IHC and MSI-PCR should sufficient for diagnostic purposes for regular labs.

2/28/2020 1:13 PM

24 NGS assay might need a common standard, such as a set of markers and set the cut off. 2/28/2020 12:48 PM

25 Should follow EU recommendations and read MMR IHC and MSI by PCR. The assays measuretwo different aspects of the same biological phenomenon and have a~10% differential detectionrate. Also MSI by NGS has not been clinically proven to be equivalent. Only real correlativedata is Foundation Medicine PMA submission. Assay need a high level of validation as TMB isnot equivalent clinically!

2/24/2020 10:25 AM

26 dfasfjlj kldasjfdsal flsajf' ljfdlk sjflkjsal; jdsflk jfllkasdjfl dsllkjk fdljfldsaj llkjalkdf dalllkj dfsajlklkjfdjsaljll;lkdlfjlsadjlkajsfljflkjsdlkfl lajsdlkfjdklsaj ljd f

2/24/2020 10:16 AM

27 Validation of MMR IHC using the CAP Guidelines can be very challenging. Will this beaddressed in the final manuscript?.

2/22/2020 9:32 AM

28 There is robust data supporting the performance of validated NGS testing for detection of MSIin CRC and other tumor types (27863258, 29955144, 30211344, 29596542, 29277635,30389464, 31530574, and more) and it should be considered an equal of IHC and PCR. In theabsence of seeing the data used to assemble this recommendation, it is difficult to know if thetwo tiered recommendation is an artifact of the way GRADE criteria are applied or if the timingof the literature review meant that much of the supporting literature was missing.

2/22/2020 8:01 AM

29 CT guided biopsies performed for these assessments have limited tumoral tissue and MMR isthe most appropriate study in these siturations. MSI and NGS require substantially more tissueand are more appropriate for resection specimens. This should be incorporated into therecommendation.

2/21/2020 11:22 AM

30 I believe the data supports that MSI detection in a validated NGS assay is equivalent to MMRby IHC and MSI by PCR. By stating that MMR by IHC and MSI by PCR are preferred you areimplying that they are superior. According to our labs validation of MSI detection by NGS theyare equivalent.

2/21/2020 9:05 AM

31 In CRC patients being considered for checkpoint blockade therapy, pathologists should useMMR IHC and/or MSI by PCR or a validated NGS MSI technique for the detection of DNAmismatch repair defects.

2/21/2020 7:05 AM

32 I completely disagree with this recommendation, for the following reasons. 1) MSI detection byNGS is more sensitive than the traditional 5-7 marker PCR panel. Our lab has been using NGSfor MSI for the past 2 years and we have seen several examples where it has identified MSI-high tumors that were missed by PCR (and in one case missed by MMR IHC). 2) NGS panelsnot only can provide MSI coverage, but can simultaneously provide broad coverage forclinically informative mutations in KRAS, NRAS, BRAF and the MMR genes. Why run othertests when NGS is a single solution? 3) Several NGS tests on the market, including the oneoffered by our (non-profit) academic laboratory, now have approval for coverage by Medicare,so insurance coverage is no longer a good reason to bias testing recommendations againstNGS-based assays. I strongly urge you to consider changing the recommendation to put NGSassays on equal footing with MMR IHC and MSI PCR. There is no good reason to prefer thesetests over NGS.

2/20/2020 3:29 PM

33 MSI is change in nucleotide hence has to be a sequencing technology to detect accurately.With 11 steps of cumulative errors NGS is not accurate MSI by PCR is subjective read smallpeaks and not accurate. MMR IHC is non verifiable.

2/20/2020 2:35 PM

34 To suggest that PCR MMR is superior to NGS MSI is misleading. The fact that PCR MMR hasbeen around longer does not make it the superior test.There are many advantages of NGS.There is the ability to know the quality of material that is being sequenced. NGS assess 100s-1000s of microsatelittes compared to <6. If NGS preceded PCR we would never consider usingit in clinical practice. I don't think guidelines should enshrine technology that is on its way out.

2/20/2020 2:11 PM

35 'cancer' is a broad term and includes e.g. lymphoma. Here and throughout the document, 2/20/2020 10:54 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

12 / 54

specify adenocarcinoma or other specific cancer types.

36 This suggestion applies to other tumor types as well. While MMR by IHC or NGS can be donewith tumor tissue only, MSI by PCR requires submitting normal tissue, which causesinconvenience and delay, especially when dealing with small biopsy specimens. I suggest MMRshould be first line. When equivocal, NGS or PCR may be attempted.

2/20/2020 8:16 AM

37 As many laboratories are moving to next generation sequencing for mutation detection,expanded panels are more often including the MMR genes. We now have the capability ofidentifying the specific mutation in the MMR gene, something we used to use IHC for. I believeit would be fraudulent to bill for MMR IHC and NGS when the NGS panel includes these genes.It would also allow for better tissue stewardship if we did not need to do MMR IHC.

2/20/2020 8:10 AM

38 We use the Idylla of Biocartis and is a very robust method 2/20/2020 7:49 AM

39 MMR IHC AND MSI BY PCR, PERFORMED SIMULTANEOUSLY ARE MORE LIKELY TOIDENTIFY PATIENTS ELIGIBLE FOR CHECKPONT INHIBITOR THERAPY

2/20/2020 12:25 AM

40 Testing is being recommended for patients being considered for target therapy. ? Testing shouldbe done on all patients to investigate for Lynch syndrome, irrespective of check point therapy?

2/19/2020 11:49 PM

41 Consider spelling out abbreviations in the final document, at least when first mentioned. 2/19/2020 11:23 PM

42 I have no way of actually knowing how these NGS labs have validated their MSI by NGS assay.Getting detailed validation studies out of these commercial laboratories can be challenging. It isalso impractical for community pathologists to be able to assess the robustness of an NGSvalidation. I would recommend rephrasing this so that the impetus for assessing validationdocuments is not placed upon community pathologists who will be providing this service as asend out test.

2/19/2020 2:42 PM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

13 / 54

56.52% 78

13.77% 19

15.22% 21

14.49% 20

Q4 Draft Recommendation Statement 2In gastroesophageal and smallbowel cancer patients being considered for checkpoint blockade therapy,pathologists should use MMR IHC and/or MSI by PCR over MSI by NGS

for the detection of DNA mismatch repair defects.Note: Thisrecommendation does not include esophageal squamous cell carcinoma.

(Quality of Evidence: Low; Strength of Recommendation: Strong)Answered: 138 Skipped: 106

TOTAL 138

Agree aswritten



Neutral

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Agree as written



Neutral


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

14 / 54

# COMMENTS DATE

1 It is unclear why IHC or MSI by PCR should be preffered in this context. Historically, MSI testingby PCR has been optimized to identify Lynch syndrome patients by screening CRC patients. It'snot at all clear why these methods would be better for GEJ and SB adenocarcinoma MMRtesting than NGS.

3/13/2020 8:14 PM

2 Currently, tissue-based biomarker testing has been utilized for the diagnosis and classificationof gastroesophageal and small bowel cancer and remains the gold standard for MMR/MSItesting. However, tissue testing has its limitations. Difficulty obtaining sufficient quantity ofsample for biomarker testing, sampling biases due to tumor heterogeneity, and tumor qualitydue to necrosis can impair testing results(1-6). Additionally, patients may choose to not undergobiopsy or be medically unfit. In order to not limit patient access to pembrolizumab throughtissue only MSI testing, utilizing cfDNA/ctDNA biopsies as an alternative may be an option formany patients. A study has shown that plasma MSI testing with a limit of detection of 0.1%tumor content can accurately detected 87% (71/82) of tissue MSI-H and 99.5% of tissuemicrosatellite stable (863/867) for an overall accuracy of 98.4% (934/949) and a positivepredictive value of 95% (71/75) when comparing to solid tumor IHC and PCR methods(7).Additionally, three companies are pursuing FDA approval on their plasma based assays thatmeasure MSI, FoundationOne™ Liquid (Foundation Medicine, MA, USA), PGDx elio™(Personal Genome Diagnostics, MD, US) and Guardant360® assay (Guardant Health, CA,USA) (8-10). We do not believe that liquid biopsies should be used as the primary sources forMSI status, but as an alternative when other options have been exhausted. The NationalComprehensive Cancer Network states for cfDNA/ctDNA testing in Non-Small Cell LungCancer: “ cell-free/circulating tumor DNA testing should not be used in lieu of a histologic tissuediagnosis” and that “studies have demonstrated cell-free tumor DNA testing to generally havevery high specificity, but significantly compromised sensitivity, with up to a 30% false-negativerate”(11). Additionally, the NCCN guidelines for breast cancer similarly state, “if liquid biopsy isnegative, tumor tissue testing is recommended”(12). Considering MSI is a pan-caner biomarker,it’s reasonable to extrapolate similar guardrails for testing in gastroesophageal and small bowelcancer. We believe that access to checkpoint blockade therapy through MSI testing, should notbe limited for those who have the available tumor for testing. We support the inclusion ofutilizing NGS or PCR liquid biopsy assays that have been orthogonally validated as a potentialoption for patients that don’t have tumor available for testing. Comments and Changes: DraftRecommendation Statement 2 In gastroesophageal and small bowel cancer patients beingconsidered for checkpoint blockade therapy, pathologists should use MMR IHC and/or MSI byPCR over MSI by NGS for the detection of DNA mismatch repair defects. Note: Thisrecommendation does not include esophageal squamous cell carcinoma. ADD TO GUIDANCE:in cases where patient have limited tumor for testing or medically unfit for invasive tissuesampling, cell free/circulating tumor DNA assay results that have been orthogonally validated,may be used. ADD TO NOTE: ctDNA/ctDNA assays have been known to have high specificitybut comprised sensitivity. False negative results are common with cfDNA/ctDNA assays andshould be followed up with tissue testing if possible. cfDNA/ctDNA testing should not be used inlieu of tissue testing and should be orthogonally validated against IHC or PCR tissue assays. 1.Vanderlaan PA, Yamaguchi N, Folch E, et al. Success and failure rates of tumor genotypingtechniques in routine pathological samples with non‐small‐cell lung cancer. Lung Cancer.2014;84:39‐44. [PMC free article] [PubMed] [Google Scholar] 2. Ellis PM, Shepherd FA,Millward M, et al. Dacomitinib compared with placebo in pretreated patients with advanced ormetastatic non‐small‐cell lung cancer (NCIC CTG BR.26): a double‐blind, randomised, phase 3trial. Lancet Oncol. 2014;15:1379‐1388. [PubMed] [Google Scholar] 3. Popper HH.Commentary on tumor heterogeneity. Transl Lung Cancer Res. 2016;5:433‐435. [PMC freearticle] [PubMed] [Google Scholar] 4. De Mattos‐Arruda L, Weigelt B, Cortes J, et al. Capturingintra‐tumor genetic heterogeneity by de novo mutation profiling of circulating cell‐free tumorDNA: a proof‐of‐principle. Ann Oncol. 2014;25:1729‐1735. [PMC free article] [PubMed] [GoogleScholar] 5. Lebofsky R, Decraene C, Bernard V, et al. Circulating tumor DNA as a non‐invasivesubstitute to metastasis biopsy for tumor genotyping and personalized medicine in aprospective trial across all tumor types. Mol Oncol. 2015;9:783‐790. 6. Murtaza M, Dawson SJ,Pogrebniak K, et al. Multifocal clonal evolution characterized using circulating tumour DNA in acase of metastatic breast cancer. Nat Commun. 2015;6:8760 7. Willis, J., & Kopetz, S. (2019).Validation of Microsatellite Instability Detection Using a Comprehensive Plasma-BasedGenotyping Panel. Precision Medicine and Imaging, 7035–7045. 8. Foundation Medicine PressRelease. Foundation Medicine’s new liquid biopsy assay granted Breakthrough DeviceDesignation by U.S. Food and Drug Administration. April 26, 2018. Available online: http://investors.foundationmedicine.com/news-releases/newsrelease-details/foundation-medicines-new-liquid-biopsyassay-granted 9. 9 . Personal Genome Diagnostics Press Release. Personal

3/13/2020 3:15 PM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

15 / 54

Genome Diagnostics’ PGDx elio™ plasma resolve Receives Breakthrough Device Designationfrom FDA. July 2018. Available online: http://www.personalgenome. com/wp-content/uploads/2018/07/Personal-GenomeDiagnostics-PGDx-elio-plasma-resolve-ReceivesBreakthrough-Device-Designation-from-FDA.pdf 10. Guardant Health Press Release.The Guardant360® Assay Receives Expedited Access Pathway Designation for BreakthroughDevices from FDA. February 15, 2018. Available online:https://www.prnewswire.com/newsreleases/the-guardant360-assay-receives-expedited-accesspathway-designation-for-breakthrough-devices-fromfda-300599629.html 11. NCCNClinical Practice Guidelines in Oncology, Non-Small Cell Lung Cancer, Version 3.2020-February,2020 12. NCCN Clinical Practice Guidelines in Oncology, Breast, Version 3.2020-March,2020

3 Again NGS based assays have demonstrated high concordance with IHC and PCR andtherefore should not be discriminated against. Pathologists should be able to best determinetesting methodologies based on tissue availability and available resources. By preferring oneequivalent method over another these recommendations are not advocating for good tissuestewardship.

3/13/2020 2:57 PM

4 The "Note" is unclear. Is the Note implying that Esophageal Squamous Cell Carcinoma shouldnot be tested at all for MMR/MSI (if so a declarative statement to that effect is requested) or isthe Note implying that MMR/MSI PCR testing is equivalent to MSH by NGS for squamous cellcarcinoma (but not adenocarcinoma)? Request further clarification on why MMR IHC/MSI byPCR is preferred over MSI by NGS for GEJ and small bowel cancer but MSI by NGS isconsidered equivalent for Colorectal Ca. If a GEJ adenocarcinoma comes to a reference lab fortesting and MSI by NGS is ordered, should the specimen be stopped and the orderingphysician advised to order MMR IHC or MSI by PCR testing instead? If we make therecommendation as it stands, then exposure to legal issues if the less favored technology isperformed based on ordering physician request. Guideline statement could have untowardeffect.

3/13/2020 1:56 PM

5 The proposed recommendations do not specify which MSI by PCR Panel should be used. Westrongly suggest that the guidelines align with the revised Bethesda guidelines and recommenda five mononucleotide panel including BAT25, BAT26, MONO27, NR21, and NR24 instead ofthe original Bethesda panel, including BAT-25, BAT-26, D2S123, D5S346, and D17S250, sincethe mononucleotide markers have become the gold standard for MSI by PCR analysis by aseveral large molecular diagnostic laboratories across the world [Svrcek et al. (2019) Bull.Cancer 106(2): 119-128; Baudrin et al. (2018) Front. Onco. 8: 621]. There is significantevidence that shows that the original NCI/Bethesda microsatellite panel for evaluation of MSIunderestimates the number of MSI-H tumors and overestimates MSI-L tumors due to the use ofthree dinucleotide repeats. The use of mononucleotide markers exclusively in MSI testingimproves the sensitivity of the panel [Umar et al. (2004) J Natl Cancer Inst 96(4):261-8]. Testingfor MMR by IHC has been shown to have a false negative rate of 5-10% [Funkhouser et al.(2012) J. Mol Diagn 14:91-203; Dudley et al. (2012) Clin Cancer Res 22 813-20]. We thereforebelieve that due to the complementary nature of MMR IHC and MSI by PCR technologies andthe potential impact of misdiagnosis, there is a growing recognition in the field that these testsshould be performed together for maximal sensitivity when identifying patients for hereditarycancer risk and immunotherapy eligibility [Goodfellow et al. (2015) J Clin Oncol 33:4301-8;Leenen et al.(2012) Gynecologic Oncology 125(2): 414-20; Mann & Cheng (2020) Exp Reviewof Anticancer Therapy 20(1):1-4; Silva et al (2019) Ann Transl Med 7(21):600].

3/13/2020 12:57 PM

6 Currently, the National Comprehensive Cancer Network (NCCN) small bowel adenocarcinoma(SBA) guidelines specifically identify next-generation sequencing (NGS) as the validmethodology for assessing MSI and do not mention PCR testing (1). As stated previously, theEuropean Society for Medical Oncology (ESMO) guidelines for MSI testing state that,“molecular tests guarantee the highest values of specificity and sensitivity in MSI testing” andthat “NGS represents another type of molecular tests to assess MSI”, in addition to PCRmethods (2). The recommended Bethesda 5 MSI loci sites have been optimized for CRCcancer types and claims in gastroesophageal and SBA cancers are typically extrapolated fromCRC data. The ability to detect 100s to 1000s of MSI sites in addition to the Bethesda 5 loci,has the potential to increase the sensitivity and specificity for detecting MSI status ingastroesophageal and SBA cancers (3-5). Additionally, utilizing molecular techniques thatisolate the tumor DNA overcomes tissue specific limitations that IHC testing may encounter dueto tumor heterogeneity. NGS has shown to be equivalent in sensitivities and specificities whencompared to IHC and PCR. • In a pan-cancer study evaluating MSI across 18 cancer types,5930 cancer exomes were examined (including 118 stomach cancer specimens). Among 2,315specimens with MSI status by the NGS method and PCR, accuracy of the NGS method was

3/13/2020 11:43 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

16 / 54

96.6% (4). • In a pan-cancer study comparing PCR and NGS MSI status of 2189 patients (12 ofwhich were gastric cancers), NGS performance was comparable to that of MSI PCR withsensitivity of the NGS based method as 100% and specificity as 99.9% (5). • In a validationstudy analyzing MSI status, results were 97% (70/72) concordant with PCR/IHC-based MSItesting and 95% (70/74) concordant if MSI-intermediate samples were included and considereddiscordant. (6). • In a pan-cancer study evaluating MSI statuses of 12,288 advanced solidcancers, validation against MSI PCR and/or MMR IHC showed a concordance of 99.4% .(7).Considering the FDA has approved MMR/MSI status as a pan-cancer biomarker, the CAPorganization should consider both gastroesophageal/SBA studies and pan-cancer studies intheir entirety when evaluating the data landscape. There are extensive pan-cancer studies thathave demonstrated appropriately validated NGS-based methods can accurately determineMSI-H/dMMR status and overcome limitations of tumor site or type. Also, additional therapyselection biomarkers have been approved for use in gastroesophageal and SBA cancers suchas NTRK gene fusions for TRK inhibitors. MSI testing should be considered part of acomprehensive genomic profiling strategy for patients where additional biomarker testing isneeded and tumor sample is limited. As NGS has demonstrated concordance across multipleplatforms/assays with IHC and PCR techniques, it therefore should be a primary methodrecommended for detecting MSI status. Continuity across guidelines is essential for patientmanagement. Both NCCN and ESMO guidelines I consider NGS a valid type of molecular testto assess MSI. CAP’s current recommendation would not be aligned with existing guidelines.We recommend the section be changed to read as follows: "In gastroesophageal and smallbowel cancer patients being considered for checkpoint blockade therapy, pathologists shoulduse MMR IHC and/or MSI by PCR or NGS for the detection of DNA mismatch repair defects ormicrosatellite instability. Note: This recommendation does not include esophageal squamouscell carcinoma. Note: MSI by NGS assay must be validated against MMR IHC or MSI by PCRand must show equivalency." 1. NCCN Clinical Practice Guidelines in Oncology, Small BowelAdenocarcinoma, Version 1.2020https://www.nccn.org/professionals/physician_gls/pdf/small_bowel.pdf 2. C. Luchini, F. Bibeau,M. J. Ligtenberg, N. Singh, A. Nottegar, T. Bosse, R. Miller, N. Riaz, J.-Y. Douillard, F. Andreand A. Scarpa, "ESMO recommendations on microsatellite instability testing for immunotherapyin cancer, and its relationship with PD-1/PD-L1 expression and tumour mutational burden: asystematic review-based approach," Ann Oncol, vol. 30, pp. 1232-1243, 2019.https://www.annalsofoncology.org/article/S0923-7534(19)31269-4/fulltext 3. S. J. Salipante, S.M. Scroggins, H. L. Hampel, E. H. Turner and C. C. Pritchard, "Microsatellite instabilitydetection by next generation sequencing," Clin Chem, vol. 60, pp. 1192-1199, 2014.https://jmd.amjpathol.org/article/S1525-1578(15)00153-1/fulltext 4. R. J. Hause, C. C. Pritchard,J. Shendure and S. J. Salipante, "Classification and characterization of microsatellite instabilityacross 18 cancer types," Nat Med, vol. 22, pp. 1342-1350, 2016.https://www.nature.com/articles/nm.4191 5. A. Vanderwalde, D. Spetzler, N. Xiao, Z. Gatalicaand J. Marshall, "Microsatellite instability status determined by next-generation sequencing andcompared with PD-L1 and tumor mutational burden in 11,348 patients," Cancer Medicine, vol.7, no. 3, pp. 746-756, 2018. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5852359/ 6. S. E.Trabucco, K. Gowen, S. L. Maund, E. Sanford, D. A. Fabrizio, M. J. Hall, E. Yakirevich and etal, "A novel next-generation sequencing approach to detecting microsatellite instability and pan-tumor characterization of 1000 microsatellite instability-high cases in 67,000 patient samples," JMol Diagn, vol. 21, pp. 1053-1066, 2019. https://jmd.amjpathol.org/article/S1525-1578(19)30358-7/fulltext 7. S. Middha, L. Zhang, K. Nafa, G. Jayakumaran, D. Wong, H. R.Kim, J. Sadowska, M. F. Berger, D. F. Delair, J. Shia, Z. Stadler, D. S. Kimstra, M. Ladanyi, A.Zehir and J. F. Hechtman, "Reliable pan-cancer microsatellite instability assessment by usingtargeted next-generation sequencing data," JCO Precision Oncol, vol. 1, pp. 1-17, 2017.https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6130812/pdf/nihms-984169.pdf

7 Suggested Recommendation: In gastroesophageal and small bowel cancer patients beingconsidered for checkpoint blockade therapy, pathologists should use MMR IHC and/or MSI byPCR and/or MSI by validated NGS assay for the detection of DNA mismatch repair defects.Note: MSI by NGS assay must be validated against MMR IHC or MSI by PCR and must showequivalency. Additional Comments: The recommendation that IHC or PCR should be used overNGS is poorly supported by the literature or practice guidelines. In the references listed todevelop these guidelines there is ample evidence that IHC, PCR, and NGS have highconcordance (references 25, 38, 45, 55, 59, 60, 75, 81, and 95)(https://documents.cap.org/documents/mmrmsi_references_ocp.pdf). All three testingmodalities, if appropriately validated, should be recommended for use in the guidelines.Pathologists should not be constrained to specific tests based on low levels of evidence and inthe setting of multiple testing options with high concordance. Furthermore, current NCCNGuidelines for both gastric and esophageal/esophagogastic junction cancers state that “when

3/13/2020 11:36 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

17 / 54

limited tissue is available for testing, sequential testing of single biomarkers or use of limitedmolecular panels may quickly exhaust the sample. In these scenarios, comprehensive genomicprofiling via a validated NGS assay performed in a CLIA-approved laboratory may be used forthe identification of HER2 amplification, MSI, and NTRK gene fusions (NCCN GuidelinesGastric Cancer v4.2019; NCCN Guidelines Esophageal/ESJ Cancer v.4.2019). NCCNGuidelines for Small Bowel Adenocarcinoma (v.1.2010) recommend universal MMR or MSItesting in all patients with a personal history of small bowel adenocarcinoma. MMR or MSItesting should be performed in a CLIA-approved laboratory. Testing for MSI may be performedby validated NGS panels.

8 1. Use the term 'rather than' and not 'over' in the Draft Recommendation 2. Remove the vendor-specific reference to Promega in the initial statement

3/13/2020 10:02 AM

9 The recommendation that IHC or PCR should be used over NGS is not supported by theliterature. It’s admitted that the evidence level is low, which is starkly at odds with the highstrength of recommendation. Just in the references for these guidelines, there is ampleevidence that IHC, PCR, and NGS have high concordance. As noted in other comments, thereis evidence in this specific setting for the use of NGS to detect mismatch repair defects. Allthree testing modalities should be recommended for use in the guidelines. Pathologists shouldnot be constrained to specific tests based on low levels of evidence and in the setting ofmultiple testing options with high concordance. In addition, for some samples, using a singletest is preferable for both laboratory workflows and tissue preservation.

3/13/2020 9:22 AM

10 The definition of PCR methods to be used for the detection of DNA mismatch repair defectsshould be not limited to the standard NIH panel or Promega MSI panel but broadened toinclude any validated molecular method, e.g. other PCR methods like the Idylla™ MSI Assay.The Idylla™ MSI Assay has shown excellent concordance to other molecular methods as wellas IHC (References: Li et al. (2019) Clin Colorectal Cancer, pii: S1533-0028(19)30086-6;Samaison (2019) Clin Pathol. 72(12):830-835; Maloney et al. (2018) TT054, J. Mol. Diagn.20(6), 895–1039; Maertens et al. (2017) Annals of Oncology 28 (suppl_5): v22-v42; De Craeneet al. (2018) Journal of Clinical Oncology 36:15_suppl,e15639; De Craene B. et al. (2018)Annals of Oncology 29 (suppl_8): viii14-viii57; De Craene et al. (2017) Annals of Oncology 28(suppl_5): v209-v268; Nicka et. al. (2018) J. Mol. Diagn. 20(6), 895–1039; Nafa. et.al. (2018) J.Mol. Diagn. 20(6), 895–1039), is used in a large number of laboratories (> 40 in the US) andhas been validated and approved for clinical use at MSKCC by the New York State Departmentof Health Clinical Laboratory Evaluation Program (project ID: 68841;https://www.wadsworth.org/print/82410). We propose to change the first paragraph as follows:Mismatch repair (MMR) Immunohistochemistry (IHC) and microsatellite instability (MSI)-polymerase chain reaction (PCR; defined as standard NIH panel of 5-7 microsatellites or othervalidated PCR-methods such as Promega MSI panel and Idylla MSI panel; most clinicallaboratories in the US are using one of these panels).

3/13/2020 5:05 AM

11 Since MSI by NGS can measure thousands of loci and may be become cheaper in the future,we should not necessarily recommend MMR by IHC or MSI by PCR over it. I think it should beworded more similar to the "all other cancers" recommendation (#4), which seems to be moreneutral and unbiased.

3/12/2020 7:19 PM

12 Our understanding of detecting microsatellite instability comes primarily from colorectal cancerand secondarily, from endometrial carcinoma. With sparse MMR IHC and comparative MSI-PCR data from gastroesophageal and small bowel cancers, I would like to see theimplementation of co-testing with the use of optimal methodologies. Solid data should directguidelines.

3/12/2020 4:58 PM

13 IHC should be mainstay for most community type practices due to little tissue required, sameday answers, availability. PCR can be added as supplement if available. NGS is generallyirrelevant in this setting. See attached letter.

3/10/2020 3:36 PM

14 I think that it should read the same as recommendation number 1, with the "pathologists mayuse a validated MSI by NGS assay for the detection of DNA mismatch repair defects"

3/10/2020 8:10 AM

15 In gastroesophageal and small bowel cancer patients being considered for checkpoint blockadetherapy, pathologists should use MMR IHC and/or MSI by PCR as first option and MSI by NGSfor the detection of DNA mismatch repair defects as the second option.

3/9/2020 9:31 PM

16 If equivalency of NGS can be demonstrated why would a preference to PCR or IHC whenuptake in NGS is rapidly expanding.

3/9/2020 1:13 PM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

18 / 54

17 in case of non-crc we prefer to use only ihc because of msi validation only for crc 3/6/2020 10:30 AM

18 see comments above. Labs that routinely use only one assay, may consider an orthogonalassay when a result is "negative" for MMR deficiency.

3/6/2020 8:33 AM

19 Not sure I understand why this should be different to the recommendations for CRC. Is thereevidence that MSI by NGS is less accurate for GE/SI tumours than for CRC? A priori thisdoesnt make a lot of sense. Is this due to a lack of published evidence in non-CRCmalignancies for MSI by NGS?

3/3/2020 9:40 PM

20 In gastroesophageal and small bowel cancerpatients being considered for checkpoint blockadetherapy, pathologists should use MMR IHC, MSI by PCR and/or validated MSI by NGS assayforthe detection of DNA mismatch repair defects. Note: MSI by NGS assay must be validatedagainst MMR IHC or MSI by PCR and must show equivalency.

3/2/2020 10:47 AM

21 Please remove phrase: being considered for checkpoint blockade therapy. We need these testsalso to evaluate for LS, not just therapy evaluation. Should offer PD-L1 testing as an alternativeand bypass MMR MSI testing.

2/29/2020 12:54 AM

22 MSI-PCR should be used when MMR-IHC display heteromerous staining due to the htergeinityof tissue.

2/28/2020 1:13 PM

23 Should follow EU recommendations and read MMR IHC and MSI by PCR. The assays measuretwo different aspects of the same biological phenomenon and have an ~10% differentialdetection rate

2/24/2020 10:25 AM

24 For patients undergoing NGS for clinical care, NGS can serve as a screen test for MMR/MSI.The provider should consider the sensitivity and specificity for MSI detection of any NGS assaywhen used for patient care.

2/22/2020 10:05 AM

25 Again, there is data to support use of NGS in this context, albeit less than for CRC (31028081;30211344; 31048490) but there is evolving data to suggest that the additional data provided byNGS (TMB, relative number of indels) stratifies response rates to immunotherapy among MSI-Hpatients; this is data that is not apparent from MMR or MSI testing.

2/22/2020 8:01 AM

26 If there is adequate validation of MSI by NGS against IHC or PCR for GE & Small bowelcancers, NGS should be an equal option to IHC & PCR.

2/21/2020 2:21 PM

27 CT guided biopsies performed for these assessments have limited tumoral tissue and MMR isthe most appropriate study in these siturations. MSI and NGS require substantially more tissueand are more appropriate for resection specimens. This should be incorporated into therecommendation.

2/21/2020 11:22 AM

28 I believe the data supports that MSI detection in a validated NGS assay is equivalent to MMRby IHC and MSI by PCR. By stating that MMR by IHC and MSI by PCR are preferred you areimplying that they are superior. According to our labs validation of MSI detection by NGS theyare equivalent.

2/21/2020 9:05 AM

29 Why not also use the potential of a validated NGS assay? Why have different wording toabove? Good and ambitious labs will have the capacity to deliver NGS MSI and MSI by PCR.

2/21/2020 7:05 AM

30 NGS statement should reflect that of CRC. 2/20/2020 9:30 PM

31 The recommendation is for "patients being considered for checkpoint blockade therapy". Thatinformation is generally not available to the pathologists. The clinicians often demand reflexMMR testing for all cases, which is not a cost-effective approach. Please further elaborate andguide the pathologists to identify "patients being considered for checkpoint blockade therapy".

2/20/2020 6:38 PM

32 Please see comments under Statement 1. 2/20/2020 3:29 PM

33 See statement 1 2/20/2020 2:35 PM

34 same as above 2/20/2020 2:11 PM

35 1. In GEA, MLH1 methylation is very closely associated with MSI status, and Lynch syndromeis rare and is unrelated to the IO therapy goal of this guideline. Especially in low resource areasbut even generally, discuss if it is reasonable to do MLH1 IHC alone (skipping the IHCs forother rarely-lost MMR proteins) or MLH1 methylation test? 2. It is unseemly to recommend anon-FDA approved commercial product, but if you insist on doing so then use the specific kitname that you recommend, e.g. 'Microsatellite Instability Analysis System, v1.2'. Explain why

2/20/2020 10:54 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

19 / 54

you do not recommend alternative commercial products such as the kit athttps://www.thermofisher.com named 'TrueMark MSI Assay for Microsatellite InstabilityAnalysis'

36 Labs should not be discouraged from using technology available to them if adequately validatedfor purpose. Include Note: MSI by NGS assay must be validated against MMR IHC or MSI byPCR and must show equivalency.

2/20/2020 10:35 AM

37 Newer MSI assays like the one from Biocartis use a SNP based system and requires only onesection of tumor tissue. This is much easier, cheaper, faster and better tissue stewardship.

2/20/2020 8:10 AM

38 Should be written as for CRC. If MSI by NGS is compared with IHC or MSI-PCR then it can besubstituted

2/20/2020 2:42 AM

39 MMR IHC AND MSI BY PCR, PERFORMED SIMULTANEOUSLY ARE MORE LIKELY TOIDENTIFY PATIENTS ELIGIBLE FOR CHECKPONT INHIBITOR THERAPY

2/20/2020 12:25 AM


2/19/2020 11:49 PM

41 I think I have an issue with the phrase “patients being considered for checkpoint therapy”because pathologists are not the physicians considering the checkpoint therapy. How is acommunity pathologist suppose to know what the oncologist is considering? I think theseshould be revised to state best practices for determining MMR/MSI status regardless of theindication.

2/19/2020 2:42 PM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

20 / 54

50.75% 68

12.69% 17

17.91% 24

17.16% 23

Q5 Draft Recommendation Statement 3In endometrial cancer patientsbeing considered for checkpoint blockade therapy, pathologists should

use MMR IHC over MSI by PCR or NGS for the detection of DNAmismatch repair defects. (Quality of Evidence: Low; Strength of

Recommendation: Strong)Answered: 134 Skipped: 110

TOTAL 134

Agree aswritten



Neutral

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Agree as written



Neutral


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

21 / 54

# COMMENTS DATE

1 While MMR deficiency in endometrial cancer may be more subtle in terms of microsatellitealterations than CRC, it is not clear that NGS especially is an inferior method for detecting MMRdeficiency.

3/13/2020 8:14 PM

2 I don't think the use of "over" reads properly 3/13/2020 5:01 PM

3 Currently, tissue-based biomarker testing has been utilized for the diagnosis and classificationof endometrial cancer and remains the gold standard for MMR/MSI testing. However, tissuetesting has its limitations. Difficulty obtaining sufficient quantity of sample for biomarker testing,sampling biases due to tumor heterogeneity, and tumor quality due to necrosis can impairtesting results(1-6). Additionally, patients may choose to not undergo biopsy or be medicallyunfit. In order to not limit patient access to pembrolizumab through tissue only MSI testing,utilizing cfDNA/ctDNA biopsies as an alternative may be an option for many patients. A studyhas shown that plasma MSI testing with a limit of detection of 0.1% tumor content canaccurately detected 87% (71/82) of tissue MSI-H and 99.5% of tissue microsatellite stable(863/867) for an overall accuracy of 98.4% (934/949) and a positive predictive value of 95%(71/75) when comparing to solid tumor IHC and PCR methods(7). Additionally, three companiesare pursuing FDA approval on their plasma based assays that measure MSI, FoundationOne™Liquid (Foundation Medicine, MA, USA), PGDx elio™ (Personal Genome Diagnostics, MD, US)and Guardant360® assay (Guardant Health, CA, USA) (8-10). We do not believe that liquidbiopsies should be used as the primary sources for MSI status, but as an alternative whenother options have been exhausted. The National Comprehensive Cancer Network states forcfDNA/ctDNA testing in Non-Small Cell Lung Cancer: “ cell-free/circulating tumor DNA testingshould not be used in lieu of a histologic tissue diagnosis” and that “studies have demonstratedcell-free tumor DNA testing to generally have very high specificity, but significantlycompromised sensitivity, with up to a 30% false-negative rate”(11). Additionally, the NCCNguidelines for breast cancer similarly state, “if liquid biopsy is negative, tumor tissue testing isrecommended”(12). Considering MSI is a pan-caner biomarker, it’s reasonable to extrapolatesimilar guardrails for testing in endometrial cancer. We believe that access to checkpointblockade therapy through MSI testing, should not be limited for those who have the availabletumor for testing. We support the inclusion of utilizing NGS or PCR liquid biopsy assays thathave been orthogonally validated as a potential option for patients that don’t have tumoravailable for testing. Comments and Changes: In endometrial cancer patients being consideredfor checkpoint blockade therapy, pathologists should use MMR IHC over MSI by PCR or NGSfor the detection of DNA mismatch repair defects. (Quality of Evidence: Low; Strength ofRecommendation: Strong) ADD TO GUIDANCE: in cases where patient have limited tumor fortesting or medically unfit for invasive tissue sampling, cell free/circulating tumor DNA assayresults that have been orthogonally validated, may be used. ADD TO NOTE: ctDNA/ctDNAassays have been known to have high specificity but comprised sensitivity. False negativeresults are common with cfDNA/ctDNA assays and should be followed up with tissue testing ifpossible. cfDNA/ctDNA testing should not be used in lieu of tissue testing and should beorthogonally validated against IHC or PCR tissue assays. 1. Vanderlaan PA, Yamaguchi N,Folch E, et al. Success and failure rates of tumor genotyping techniques in routine pathologicalsamples with non‐small‐cell lung cancer. Lung Cancer. 2014;84:39‐44. [PMC free article][PubMed] [Google Scholar] 2. Ellis PM, Shepherd FA, Millward M, et al. Dacomitinib comparedwith placebo in pretreated patients with advanced or metastatic non‐small‐cell lung cancer(NCIC CTG BR.26): a double‐blind, randomised, phase 3 trial. Lancet Oncol. 2014;15:1379‐1388. [PubMed] [Google Scholar] 3. Popper HH. Commentary on tumor heterogeneity. TranslLung Cancer Res. 2016;5:433‐435. [PMC free article] [PubMed] [Google Scholar] 4. De Mattos‐Arruda L, Weigelt B, Cortes J, et al. Capturing intra‐tumor genetic heterogeneity by de novomutation profiling of circulating cell‐free tumor DNA: a proof‐of‐principle. Ann Oncol.2014;25:1729‐1735. [PMC free article] [PubMed] [Google Scholar] 5. Lebofsky R, Decraene C,Bernard V, et al. Circulating tumor DNA as a non‐invasive substitute to metastasis biopsy fortumor genotyping and personalized medicine in a prospective trial across all tumor types. MolOncol. 2015;9:783‐790. 6. Murtaza M, Dawson SJ, Pogrebniak K, et al. Multifocal clonalevolution characterized using circulating tumour DNA in a case of metastatic breast cancer. NatCommun. 2015;6:8760 7. Willis, J., & Kopetz, S. (2019). Validation of Microsatellite InstabilityDetection Using a Comprehensive Plasma-Based Genotyping Panel. Precision Medicine andImaging, 7035–7045. 8. Foundation Medicine Press Release. Foundation Medicine’s new liquidbiopsy assay granted Breakthrough Device Designation by U.S. Food and Drug Administration.April 26, 2018. Available online: http:// investors.foundationmedicine.com/news-releases/newsrelease-details/foundation-medicines-new-liquid-biopsyassay-granted 9. 9 .Personal Genome Diagnostics Press Release. Personal Genome Diagnostics’ PGDx elio™

3/13/2020 3:15 PM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

22 / 54

plasma resolve Receives Breakthrough Device Designation from FDA. July 2018. Availableonline: http://www.personalgenome. com/wp-content/uploads/2018/07/Personal-GenomeDiagnostics-PGDx-elio-plasma-resolve-ReceivesBreakthrough-Device-Designation-from-FDA.pdf 10. Guardant Health Press Release. The Guardant360® Assay ReceivesExpedited Access Pathway Designation for Breakthrough Devices from FDA. February 15,2018. Available online: https://www.prnewswire.com/newsreleases/the-guardant360-assay-receives-expedited-accesspathway-designation-for-breakthrough-devices-fromfda-300599629.html 11. NCCN Clinical Practice Guidelines in Oncology, Non-Small Cell LungCancer, Version 3.2020- February,2020 12. NCCN Clinical Practice Guidelines in Oncology,Breast, Version 3.2020-March,2020

4 Given the presence of molecular subtypes in endometrial cancer, greater support for NGSbased MSI testing should be made in order to take advantage the the additional clinical utilitythat NGS testing provides.

3/13/2020 2:57 PM

5 We suggest adding to Statement 3 the same Note as per Statement 4: Note: Assays must beadequately validated for the specific cancer type being tested with careful consideration ofperformance characteristics of MMR IHC and MSI by NGS for the detection of DNA mismatchrepair defects.

3/13/2020 2:51 PM

6 Recommending use of different technologies for different body sites will be difficult foroncologists (who request the test) to remember. Will it be Pathologists responsibility to "stop"the test and consult with the Oncologist if the less favored technology is ordered? Request thatCAP think globally about the operational effect in addition to the scientific data in terms of real-world deployment. Consider standardizing and recommending one approach for ALL solidtumor types.

3/13/2020 1:56 PM

7 Testing for endometrial cancer with IHC alone is likely to miss up to 5% of MSI-H tumors. Oneexplanation for this well documented observation is that missense mutations in MMR genes canresult in fully expressed but non-functional proteins detected by IHC, resulting in a false pMMRresult [Hechtman et al (2019) Mod Path; Shia (2008) J Mol Diag 10(4); Funkhouser et al. (2012)J. Mol Diagn 14:91-203]. Endometrial and other cancers from patients with a germline mutationin MSH6 can sometimes present an MSI-L phenotype when using the NIH panel. This isbecause the MSH6 protein is not involved in the repair of mismatches of two or morenucleotides in length and consequently the three dinucleotide repeats of the NIH panel oftenshow stability in MSH6-deficient tumors. A study using a panel of 5 mononocleotide repeatscorrectly identified MSI-H status in 15 out of 15 (100%) dMMR cases where matching normalDNA was available and in 28 out of 29 (97%) cases where matching DNA was not available[You et al (2010) Brit J Cancer 103(12):1840-45]. These results indicate that a panel of onlymononucleotide repeats effectively determines the correct MSI status of MSH6 deficienttumors. We also recommend including guidance on the use of a specific (newer) MSI by PCRpanel that show improved sensitivity over the older NIH/Bethesda panel. There is a strong biasin favor of IHC for endometrial cancers because of the misconception that MSI is not sensitiveto detecting MSI in this cancer. The primary reason for this is the subtle MSI changes (incomparison to CRC) may be overlooked using current MSI by PCR systems in someendometrial cancers. Studies have shown that when subtle MSI changes are included, MSI isas sensitive at detecting MSI in endometrial cancer as is MMR by IHC [Wang et al (2017) J MolDiag 19(1): 57-64). Further, new loci being developed for MSI testing may allow for morestraightforward detection of MSI in endometrial cancer [Bacher et al (2018) Development of aPan-Cancer Biomarker Panel for Improved Detection of MSI Across all Cancer Types. Abstract2960 ESMO].

3/13/2020 12:57 PM

8 Clinical data in the Keytruda package insert does not separate dMMR status from MSI-H status.The recommendation of using MMR status instead of MSI status is not aligned with theKeytruda indication or guidelines and may limit identifying eligible patients for checkpointinhibitors (1). Both markers are clinically validated and should be assessed. As statedpreviously, ESMO recommendations state, “Molecular tests guarantee the highest values ofspecificity and sensitivity in MSI testing”, with the molecular test options being both PCR andNGS methods (2). Utilizing molecular tests that isolate the tumor DNA overcomes tissuespecific limitations that IHC testing may encounter due to tumor heterogeneity. Furthermore, theNational Comprehensive Cancer Network uterine cancer guideline does not make anypreferred distinction for dMMR status over MSI-H status (3). MSI-H status should be consideredan equally relevant biomarker to dMMR status as it is measuring the same biological effect.Both PCR and NGS should be methods of testing MSI-H status. Additionally, NGS has shownto be equivalent in sensitivities and specificities when comparing to IHC and PCR forendometrial cancer. • In a pan-cancer study (including 260 uterine endometrioid cancer (UEC)

3/13/2020 11:43 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

specimens) comparing an NGS-based method to PCR and IHC for the detection of MSI-H/dMMR, 16% (42/260) UEC specimens had MSIH by the NGS-based method (4). Among 40UEC patients with MSI status and PCR/IHC data, concordance was 97.5% (39/40) between theNGS-based method and IHC and/or PCR. • In a pan-cancer study (including 261 endometrialcancer (EC) specimens) evaluating MSI across 18 cancer types, 5930 cancer exomes wereexamined (5). Among 617 specimens with MSI status by the NGS method and PCR, accuracyof the NGS method was 96.6%. Most of the samples with discordant results were ECspecimens; evidence suggests that many of these were improperly classified by PCR. • In apan-cancer study (including 709 endometrial cancer specimens) analyzing MSI status, NGSperformance was comparable to that of MSI PCR. In EC specimens, sensitivity of the NGSbased method was 93.9% (123/131) and specificity of 98.8% (570/577) (6). • In studiescomparing performance of NGS to PCR for MSI detection in samples including EC specimens,NGS performance was comparable to that of PCR with sensitivity estimates of ≥ 96.4% andspecificity estimates of ≥ 97.2% reported across 3 panels (7). • In a pan-cancer studyinvestigating the genomic landscape of MSI-H in clinically advanced cancers, uterineendometrial cancers had the highest rate of MSI-H (16.5%, 238/1439) of the samples studied(8). Among 72 samples with conclusive NGS and IHC and/or PCR results, concordance was97.0% (70/72) between the NGS-based method and IHC and/or PCR. NGS is well establishedin the endometrial cancer literature with extensive pan-cancer studies that have demonstratedappropriately validated NGS-based methods can accurately determine MSI-H/dMMR statusand overcome limitations of tumor site or type. Also, additional therapy selection biomarkershave also been approved for use in endometrial cancers such as NTRK gene fusions for TRKinhibitors. MSI testing should be considered part of a comprehensive genomic profiling strategyfor patients where additional biomarker testing is needed and tumor sample is limited. NGS hasdemonstrated concordance across multiple platforms/assays with IHC and PCR techniques andtherefore should be a primary method recommended for detecting MSI status. Furthermore,considering both Keytruda’s package insert and the NCCN endometrial guidelines do not makeany distinctions between dMMR/MSI-H status, and the ESMO guidelines for MSI consider NGSa valid type of molecular test to assess MSI, continuity across guidelines is essential for patientmanagement. CAP’s current recommendation would be in conflict with both FDA approvals andexisting guidelines. We recommend the section be changed to read as follows: "In endometrialcancer patients being considered for checkpoint blockade therapy, pathologists should useMMR IHC or MSI by PCR or NGS for the detection of DNA mismatch repair defects ormicrosatellite instability. Note: MSI by NGS assay must be validated against MMR IHC or MSIby PCR and must show equivalency." 1. Keytruda KEYTRUDA® Package Insert(Pembrolizumab) MERK 2019https://www.accessdata.fda.gov/drugsatfda_docs/label/2019/125514Orig1s054lbl.pdf 2. C.Luchini, F. Bibeau, M. J. Ligtenberg, N. Singh, A. Nottegar, T. Bosse, R. Miller, N. Riaz, J.-Y.Douillard, F. Andre and A. Scarpa, "ESMO recommendations on microsatellite instability testingfor immunotherapy in cancer, and its relationship with PD-1/PD-L1 expression and tumourmutational burden: a systematic review-based approach," Ann Oncol, vol. 30, pp. 1232-1243,2019. https://www.annalsofoncology.org/article/S0923-7534(19)31269-4/fulltext 3. NCCNClinical Practice Guidelines in Oncology, Uterine Neoplasms, Version 1.2020https://www.nccn.org/professionals/physician_gls/pdf/uterine.pdf 4. S. Middha, L. Zhang, K.Nafa, G. Jayakumaran, D. Wong, H. R. Kim, J. Sadowska, M. F. Berger, D. F. Delair, J. Shia, Z.Stadler, D. S. Kimstra, M. Ladanyi, A. Zehir and J. F. Hechtman, "Reliable pan-cancermicrosatellite instability assessment by using targeted next-generation sequencing data," JCOPrecision Oncol, vol. 1, pp. 1-17, 2017.https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6130812/pdf/nihms-984169.pdf 5. R. J. Hause,C. C. Pritchard, J. Shendure and S. J. Salipante, "Classification and characterization ofmicrosatellite instability across 18 cancer types," Nat Med, vol. 22, pp. 1342-1350, 2016.https://www.nature.com/articles/nm.4191 6. Johansen, A.F.B., Kassentoft, C.G., Knudsen, M. etal. Validation of computational determination of microsatellite status using whole exomesequencing data from colorectal cancer patients. BMC Cancer 19, 971 (2019).https://doi.org/10.1186/s12885-019-6227-7https://bmccancer.biomedcentral.com/articles/10.1186/s12885-019-6227-7 7. S. J. Salipante, S.M. Scroggins, H. L. Hampel, E. H. Turner and C. C. Pritchard, "Microsatellite instabilitydetection by next generation sequencing," Clin Chem, vol. 60, pp. 1192-1199, 2014.https://jmd.amjpathol.org/article/S1525-1578(15)00153-1/fulltext 8. S. E. Trabucco, K. Gowen,S. L. Maund, E. Sanford, D. A. Fabrizio, M. J. Hall, E. Yakirevich and et al, "A novel next-generation sequencing approach to detecting microsatellite instability and pan-tumorcharacterization of 1000 microsatellite instability-high cases in 67,000 patient samples," J MolDiagn, vol. 21, pp. 1053-1066, 2019. https://jmd.amjpathol.org/article/S1525-1578(19)30358-7/fulltext Disclaimer

The information, data, and draft recommendations provided by the College of American Pathologists are presented for informational and public feedback purposes only. The draft recommendations and supporting documents will be removed on March 20, 2020.The draft recommendations along with the public comments received and completed evidence review will be reassessed by the expert panel in order to formulate the final recommendations. These draft materials should not be stored, adapted, or redistributed in any manner.

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

24 / 54

9 Suggested Recommendation: In endometrial cancer patients being considered for checkpointblockade therapy, pathologists should use MMR IHC and/or MSI by PCR and/or MSI by NGSvia a validated assay for the detection of DNA mismatch repair defects. Note: MSI by NGSassay must be validated against MMR IHC or MSI by PCR and must show equivalency.Additional Comments: The recommendation that IHC or PCR should be used over NGS ispoorly supported by the literature or current practice guidelines. In the list of references forwhich the recommendations are based, there is ample evidence that IHC, PCR, and NGS havehigh concordance (references 25, 38, 45, 55, 59, 60, 75, 81, and 95)https://documents.cap.org/documents/mmrmsi_references_ocp.pdf). Furthermore, NCCNGuidelines Uterine Neoplasms v.1.2020 recommends universal testing of endometrialcarcinomas for MMR proteins/MSI as well as molecular testing to complement histologic tumorsubtyping as the four clinically significant molecular subgroups have differing clinicalprognoses; therefore, it is recommended that POLE mutations, MSI-H, copy number low andcopy number high status be determined. Pathologists should not be constrained to specifictests based on low levels of evidence and in the setting of multiple testing options with highconcordance, especially when multiple molecular recommendations are made for clinicalpractice.

3/13/2020 11:36 AM

10 1. Remove the vendor-specific reference to Promega in the initial statement 3/13/2020 10:02 AM

11 The recommendation that IHC or PCR should be used over NGS is not supported by theliterature. It’s admitted that the evidence level is low, which is starkly at odds with the highstrength of recommendation. Just in the references for these guidelines, there is ampleevidence that IHC, PCR, and NGS have high concordance. Pathologists should not beconstrained to specific tests based on low levels of evidence and in the setting of multipletesting options with high concordance. In addition, for some samples, using a single test ispreferable for both laboratory workflows and tissue preservation.

3/13/2020 9:22 AM

12 1. The definition of PCR methods to be used for the detection of DNA mismatch repair defectsshould be not limited to the standard NIH panel or Promega MSI panel but broadened toinclude any validated molecular method, e.g. other PCR methods like the Idylla™ MSI Assay.The Idylla™ MSI Assay has shown excellent concordance to other molecular methods as wellas IHC (References: Li et al. (2019) Clin Colorectal Cancer, pii: S1533-0028(19)30086-6;Samaison (2019) Clin Pathol. 72(12):830-835; Maloney et al. (2018) TT054, J. Mol. Diagn.20(6), 895–1039; Maertens et al. (2017) Annals of Oncology 28 (suppl_5): v22-v42; De Craeneet al. (2018) Journal of Clinical Oncology 36:15_suppl,e15639; De Craene B. et al. (2018)Annals of Oncology 29 (suppl_8): viii14-viii57; De Craene et al. (2017) Annals of Oncology 28(suppl_5): v209-v268; Nicka et. al. (2018) J. Mol. Diagn. 20(6), 895–1039; Nafa. et.al. (2018) J.Mol. Diagn. 20(6), 895–1039), is used in a large number of laboratories (> 40 in the US) andhas been validated and approved for clinical use at MSKCC by the New York State Departmentof Health Clinical Laboratory Evaluation Program (project ID: 68841;https://www.wadsworth.org/print/82410). We propose to change the first paragraph as follows:Mismatch repair (MMR) Immunohistochemistry (IHC) and microsatellite instability (MSI)-polymerase chain reaction (PCR; defined as standard NIH panel of 5-7 microsatellites or othervalidated PCR-methods such as Promega MSI panel and Idylla MSI panel; most clinicallaboratories in the US are using one of these panels). 2. IHC has known limitations/risk withsome of them even playing a more prominent role in EC compared to CRC: False negativeresults where protein function impaired but still present (e.g. some MLH1 promoter methylationcases which may show false-positive nuclear staining for MLH1; pathogenic missensemutations, dominant negative mutations) (Latham et al, J Clin Oncol. 2018;37(4):286-95) Falsepositive dMMR IHC results. In mCRC patients, it was recently shown that misdiagnosis ofdMMR status was the main reason for primary resistance to immune checkpoint inhibitors(Cohen et al. JAMA Oncol. 2019;5(4):551-555). Our experience from the field is thatassessment of dMMR status by IHC is challenging in EC, and false positive dMMR status mayoccur even more frequently in EC as compared to mCRC due to over-fixation (related to thethickness of the sample) leading to false negative staining which leads to false positive dMMRstatus. Yet we have not been able to find published evidence for this experience in EC.Subclonal loss / tumor heterogeneity especially in endometrial cancer (Stelloo et al., Ann Oncol.2017;28(1):96-102) Less reliable on small samples (Zhang et al, g. J Mol Diagn. 2008;10:301-7) Based on recent publications, main reasons why IHC is preferred (or rather „appearsadequate“ (Stelloo et al. 2017) for assessment of MMR protein in EC seems to be lower costsand wide availability of IHC, together with the controversial discussion on whether EC MSI-Lcases (obtained with some molcualr methods) are best considered as MSS or MSI (Stelloo etal). The appearance of MSI-L or indeterminate cases is specific for SOME but not ALLmolecular methods, therefore this disadvantage of specific methodologies should not be

3/13/2020 5:05 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

25 / 54

generalized to all MSI by PCR methods. The Idylla MSI assay clearly separates betwenn MSI-Hand MSS cases without any indeterminate results. Moreover, Idylla MSI uses a completelydifferent sets of biomarkers compared to the Promega MSI panel which is well established inCRC but seems to underperform in EC. The Idylla biomarkers have been originally selectedbased on a dataset of not only of CRC and but also endometrial cases (Zhao et al. (2014) eLife3:e02725, 1-26). In summary, the recommendation should favor molecular testing ofmicrosatellites as direct proof of dMMR, with IHC as an efficient indirect test for dMMR when amolecular laboratory is not available. This would be in line with the recommendation of ESMO(OncologyPRO - Microsatellite Instability - Defective DNA Mismatch Repair: ESMO BiomarkerFactsheet).

13 The wording should give equal weight and legitimacy to any of the three methods, and notpreference towards one of these methods over the others.

3/12/2020 7:19 PM

14 There was ASCO representation on this expert panel to draft these guidelines. I would like tosee representation from SGO. For the reasons stated under the colorectal guidelines, co-testing MMR IHC and MSI-PCR should be strongly recommended for endometrial carcinoma.Instability in endometrial cancers with the Promega v1.2 MSI kit is often presented as subtle,1bp shifts (Wang Y, Shi C, Eisenberg R, Vnencak-Jones CL: Differences in MicrosatelliteInstability Profiles between Endometrioid and Colorectal Cancers: A Potential Cause for False-Negative Results? J Mol Diagn. 2017; 19(1):57-64) which can be missed with sub-optimalexperimental conditions. However, it is also known that MSH6 IHC can be challenging tointerpret (ambiguous/focal) and MSH6 mutations are known to confer a particular risk for Lynchsyndrome in endometrial cancers. Given this landscape, co-testing MMR-IHC and MSI-PCRshould be recommended.

3/12/2020 4:58 PM

15 IHC should be mainstay due to smaller STR shifts. Capillary based PCR will especiallyunderperform in this tumor type but newer methods such as HRM may be practically moreuseful. NGS is generally irrelevant in community setting. See attached letter.

3/10/2020 3:36 PM

16 again should read the same as number 1. 3/10/2020 8:10 AM

17 MSI by PCR is no less accurate or reliable than MMR by IHC for determination of checkpointinhibitor use in endometrial cancer. Both should be used over MSI by NGS.

3/9/2020 7:27 PM

18 Same as above. If demonstrated equivalency for PCR and NGS can be demonstrated thereshouldn't be a preference

3/9/2020 1:13 PM

19 This creates a problems for testing, it woud be easier to use the same modality for all. what isthe evidence of IHC over PCR and is it really significant?

3/9/2020 1:01 PM

20 I feel there needs a clarification on serous vs no serous carcinomas and the testing Clarificationis needed - " TEST All subtypes" of endometrial carcinomas

3/9/2020 12:53 PM

21 Would recommend preferring MMR IHC over MSI by PCR/NGS for two reasons - MMR IHCcan use substantially less tissue for biopsies/small samples and also the pattern of MMR IHCdeficiency can provide information with regard to next steps for lynch syndrome testing (i.e.which genes should be sequenced first, should hypermethylation be definitively excluded viaseparate assay, etc.)

3/9/2020 12:23 PM

22 use endometrial carcinoma instead of cancer or do you intend to include mesenchymalneoplasms. I even think it should be only endometrial endometrioid carcinoma and dediff/undiffcarcinomas. Replace clonal with subclonal because clonal could mean the founder clone.

3/9/2020 12:07 PM

23 Again, while MMR IHC is the preferred method, given its lower than 100% sensitivity, labs mayconsider the use of an orthogonal assay when results are "normal" or "negative" for MMRdeficiency.

3/6/2020 8:33 AM

24 MSI PCR changes are more subtle in endometrial cancers with MMR-D, but may still be reliablydetectable by an NGS approach interrogating multiple loci e.g Foundation's publishedevidence.

3/3/2020 9:40 PM

25 It is not clear why MSI testing should be considered less informative in endometrial cancercompared to CRCA or EGJ cancers.

3/2/2020 9:11 PM

26 Please remove phrase: being considered for checkpoint blockade therapy. We need these testsalso to evaluate for LS, not just therapy evaluation. Also, MMR IHC and/or MSI by PCR and/orNGS testing. Should offer PD-L1 testing as an alternative and bypass MMR MSI testing.

2/29/2020 12:54 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

26 / 54

27 MSI-PCR should be used when MMR-IHC display heteromerous staining due to the htergeinityof tissue.

2/28/2020 1:13 PM

28 MSI by PCR could be used with MMR IHC. 2/28/2020 12:48 PM

29 Rationale for this? Is this saying MMRd hypermethylated MLH1 patients should be consideredfor checkpoint blockade therapy or that the IHC should be used as initial screening? Veryunclear, needs clarification.

2/28/2020 9:38 AM

30 Pathologists should use both MMR IHC and MSI by PCR to increase sensitivity of detection ofdMMR

2/24/2020 1:21 PM

31 Should follow EU recommendations and read MMR IHC and MSI by PCR. Again assayvalidation criteria NGS assays needs to be spelled out in detail - see Foundation Medicine PMAsubmission

2/24/2020 10:25 AM

32 For patients undergoing NGS for clinical care, NGS can serve as a screen test for MMR/MSI.The provider should consider the sensitivity and specificity for MSI detection of any NGS assaywhen used for patient care.

2/22/2020 10:05 AM

33 Use of the word "over" in the two above recommendations seems awkward to me. 2/22/2020 9:32 AM

34 Same issues as above. 2/22/2020 8:01 AM

35 certain MSI by PCR kits have included (and are starting to include) LTR's that are moresensitive to endometrial tumor types and maintain the high level of sensitivity over IHC.

2/22/2020 12:20 AM

36 Is this due to MMR/MSI heterogeneity in endometrial cancers, more subtle changes inmicrosatellite lengths in endometrial cancers, larger proportion of MSH6 -> MSI-Low cases?

2/21/2020 2:21 PM

37 I believe the data supports that MSI detection in a validated NGS assay is equivalent to MMRby IHC and MSI by PCR. By stating that MMR by IHC and MSI by PCR are preferred you areimplying that they are superior. According to our labs extensive validation of MSI detection byNGS they are equivalent.

2/21/2020 9:05 AM

38 MSI by PCR should be used over MMR IHC or NGS for the detection of DNA mismatch repairdefects with simple, fast turn around and reliable results.

2/21/2020 7:39 AM

39 From personal experience, the Promega MSI assay has lower sensitivity for MMR as comparedto the IHC analysis in endometrial cancers

2/21/2020 7:35 AM

40 NGS statement should reflect that of CRC. 2/20/2020 9:30 PM

41 See comments on above statement 2/20/2020 2:35 PM

42 same as above This pattern of Quality of Evidence: Low; Strength of Recommendation: Strongis concerning to me. And I strongly disagree with every one so far.

2/20/2020 2:11 PM

43 Use MMR IHC and/or MSI by PCR 2/20/2020 2:11 PM

44 Evidence quality being low, is it reasonable to do both IHC and MSI PCR or NGS? 2/20/2020 10:54 AM

45 Labs should not be discouraged from using technology available to them if adequately validatedfor purpose. Level of evidence to choose MMR over MSI is low and this recommendation isconsidered premature. Would also add the Note: MSI by NGS assay must be validated againstMMR IHC or MSI by PCR and must show equivalency.

2/20/2020 10:35 AM

46 Newer SNP based MSI assays perform better in endometrial ca. 2/20/2020 8:10 AM

47 We do not apply this test for Endometrial Cancer 2/20/2020 7:49 AM

48 Should be written as for CRC. 2/20/2020 2:42 AM


2/19/2020 11:49 PM

50 See other comments. 2/19/2020 2:42 PM



Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

27 / 54

53.85% 70

16.92% 22

10.77% 14

18.46% 24

Q6 Draft Recommendation Statement 4In patients with cancer types otherthan CRC, GEA, small bowel, and endometrial being considered for

checkpoint blockade therapy, pathologists should test for DNA mismatchrepair, although the optimal approach for the detection of MMR defects

has not been established. Note: Assays must be adequately validated forthe specific cancer type being tested with careful consideration ofperformance characteristics of MMR IHC and MSI by NGS for the

detection of DNA mismatch repair defects.(Quality of Evidence: Very Low;Strength of Recommendation: Conditional)


TOTAL 130

Agree aswritten



Neutral

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Agree as written



Neutral


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

28 / 54

# COMMENTS DATE

1 The optimal approach may well depend upon the specific tumor type and individual specimencharacteristics (tumor content, total amount of tissue / expected DNA yield). In point of fact, asingle approach is unlikely to be optimal for all cases. Laboratories would be best advised tohave several approaches available, with the ability to triage cases amongst them and toperform reflex testing, as necessary, when the initial approach yields equivocal or unexpectedresults.

3/13/2020 8:14 PM

2 This recommendation seems really hard to implement. It doesn't suggest an approach. Italmost seems like the note is more of a recommendation that the main statement.

3/13/2020 5:01 PM

3 Currently, tissue-based biomarker testing has been utilized for the diagnosis and classificationfor all solid tumor cancer types and remains the gold standard for MMR/MSI testing. However,tissue testing has its limitations. Difficulty obtaining sufficient quantity of sample for biomarkertesting, sampling biases due to tumor heterogeneity, and tumor quality due to necrosis canimpair testing results(1-6). Additionally, patients may choose to not undergo biopsy or bemedically unfit. In order to not limit patient access to pembrolizumab through tissue only MSItesting, utilizing cfDNA/ctDNA biopsies as an alternative may be an option for many patients. Astudy has shown that plasma MSI testing with a limit of detection of 0.1% tumor content canaccurately detected 87% (71/82) of tissue MSI-H and 99.5% of tissue microsatellite stable(863/867) for an overall accuracy of 98.4% (934/949) and a positive predictive value of 95%(71/75) when comparing to solid tumor IHC and PCR methods(7). Additionally, three companiesare pursuing FDA approval on their plasma based assays that measure MSI, FoundationOne™Liquid (Foundation Medicine, MA, USA), PGDx elio™ (Personal Genome Diagnostics, MD, US)and Guardant360® assay (Guardant Health, CA, USA) (8-10). We do not believe that liquidbiopsies should be used as the primary sources for MSI status, but as an alternative whenother options have been exhausted. The National Comprehensive Cancer Network states forcfDNA/ctDNA testing in Non-Small Cell Lung Cancer: “ cell-free/circulating tumor DNA testingshould not be used in lieu of a histologic tissue diagnosis” and that “studies have demonstratedcell-free tumor DNA testing to generally have very high specificity, but significantlycompromised sensitivity, with up to a 30% false-negative rate”(11). Additionally, the NCCNguidelines for breast cancer similarly state, “if liquid biopsy is negative, tumor tissue testing isrecommended”(12). Considering MSI is a pan-caner biomarker, it’s reasonable to extrapolatesimilar guardrails for testing to all solid tumor cancers. We believe that access to checkpointblockade therapy through MSI testing, should not be limited for those who have the availabletumor for testing. We support the inclusion of utilizing NGS or PCR liquid biopsy assays thathave been orthogonally validated as a potential option for patients that don’t have tumoravailable for testing. Comments and Changes: Draft Recommendation Statement 4 In patientswith cancer types other than CRC, GEA, small bowel, and endometrial being considered forcheckpoint blockade therapy, pathologists should test for DNA mismatch repair, although theoptimal approach for the detection of MMR defects has not been established. Note: Assaysmust be adequately validated for the specific cancer type being tested with carefulconsideration of performance characteristics of MMR IHC and MSI by NGS for the detection ofDNA mismatch repair defects. (Quality of Evidence: Very Low; Strength of Recommendation:Conditional) ADD TO GUIDANCE: in cases where patient have limited tumor for testing ormedically unfit for invasive tissue sampling, cell free/circulating tumor DNA assay results thathave been orthogonally validated, may be used. ADD TO NOTE: ctDNA/ctDNA assays havebeen known to have high specificity but comprised sensitivity. False negative results arecommon with cfDNA/ctDNA assays and should be followed up with tissue testing if possible.cfDNA/ctDNA testing should not be used in lieu of tissue testing and should be orthogonallyvalidated against IHC or PCR tissue assays. 1. Vanderlaan PA, Yamaguchi N, Folch E, et al.Success and failure rates of tumor genotyping techniques in routine pathological samples withnon‐small‐cell lung cancer. Lung Cancer. 2014;84:39‐44. [PMC free article] [PubMed] [GoogleScholar] 2. Ellis PM, Shepherd FA, Millward M, et al. Dacomitinib compared with placebo inpretreated patients with advanced or metastatic non‐small‐cell lung cancer (NCIC CTG BR.26):a double‐blind, randomised, phase 3 trial. Lancet Oncol. 2014;15:1379‐1388. [PubMed][Google Scholar] 3. Popper HH. Commentary on tumor heterogeneity. Transl Lung Cancer Res.2016;5:433‐435. [PMC free article] [PubMed] [Google Scholar] 4. De Mattos‐Arruda L, WeigeltB, Cortes J, et al. Capturing intra‐tumor genetic heterogeneity by de novo mutation profiling ofcirculating cell‐free tumor DNA: a proof‐of‐principle. Ann Oncol. 2014;25:1729‐1735. [PMC freearticle] [PubMed] [Google Scholar] 5. Lebofsky R, Decraene C, Bernard V, et al. Circulatingtumor DNA as a non‐invasive substitute to metastasis biopsy for tumor genotyping andpersonalized medicine in a prospective trial across all tumor types. Mol Oncol. 2015;9:783‐790.6. Murtaza M, Dawson SJ, Pogrebniak K, et al. Multifocal clonal evolution characterized using

3/13/2020 3:15 PM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

29 / 54

circulating tumour DNA in a case of metastatic breast cancer. Nat Commun. 2015;6:8760 7.Willis, J., & Kopetz, S. (2019). Validation of Microsatellite Instability Detection Using aComprehensive Plasma-Based Genotyping Panel. Precision Medicine and Imaging, 7035–7045. 8. Foundation Medicine Press Release. Foundation Medicine’s new liquid biopsy assaygranted Breakthrough Device Designation by U.S. Food and Drug Administration. April 26,2018. Available online: http:// investors.foundationmedicine.com/news-releases/newsrelease-details/foundation-medicines-new-liquid-biopsyassay-granted 9. 9 . Personal GenomeDiagnostics Press Release. Personal Genome Diagnostics’ PGDx elio™ plasma resolveReceives Breakthrough Device Designation from FDA. July 2018. Available online:http://www.personalgenome. com/wp-content/uploads/2018/07/Personal-GenomeDiagnostics-PGDx-elio-plasma-resolve-ReceivesBreakthrough-Device-Designation-from-FDA.pdf 10.Guardant Health Press Release. The Guardant360® Assay Receives Expedited AccessPathway Designation for Breakthrough Devices from FDA. February 15, 2018. Available online:https://www.prnewswire.com/newsreleases/the-guardant360-assay-receives-expedited-accesspathway-designation-for-breakthrough-devices-fromfda-300599629.html 11. NCCNClinical Practice Guidelines in Oncology, Non-Small Cell Lung Cancer, Version 3.2020-February,2020 12. NCCN Clinical Practice Guidelines in Oncology, Breast, Version 3.2020-March,2020

4 Wording of the Note is ambigous. Is the note mandating that every solid tumor type bevalidated against MSI PCR? Why the need to validate MMR IHC for every single tumor type? Ifthe recommendation is to validate every type of solid tumor cancer, how should this be done?Guidance should be provided.

3/13/2020 1:56 PM

5 This recommendation (in the Note:) omits the use of MSI by PCR. We strongly urge you to addMSI by PCR in addition to the other techniques for detection DNA mismatch repair defects. Theproposed recommendations should specify which MSI by PCR Panel should be used. Westrongly suggest that the guidelines align with the revised Bethesda guidelines and recommenda five mononucleotide panel including BAT25, BAT26, MONO27, NR21, and NR24 instead ofthe original Bethesda panel, including BAT-25, BAT-26, D2S123, D5S346, and D17S250, sincethe mononucleotide markers have become the gold standard for MSI by PCR analysis by aseveral large molecular diagnostic laboratories across the world [Svrcek et al. (2019) Bull.Cancer 106(2): 119-128; Baudrin et al. (2018) Front. Onco. 8: 621]. There is significantevidence that shows that the original NCI/Bethesda microsatellite panel for evaluation of MSIunderestimates the number of MSI-H tumors and overestimates MSI-L tumors due to the use ofthree dinucleotide repeats. The use of mononucleotide markers exclusively in MSI testingimproves the sensitivity of the panel [Umar et al. (2004) J Natl Cancer Inst 96(4):261-8]. Testingfor MMR by IHC has been shown to have a false negative rate of 5-10% [Funkhouser et al.(2012) J. Mol Diagn 14:91-203; Dudley et al. (2012) Clin Cancer Res 22 813-20]. We thereforebelieve that due to the complementary nature of MMR IHC and MSI by PCR technologies andthe potential impact of misdiagnosis, there is a growing recognition in the field that these testsshould be performed together for maximal sensitivity when identifying patients for hereditarycancer risk and immunotherapy eligibility [Goodfellow et al. (2015) J Clin Oncol 33:4301-8;Leenen et al.(2012) Gynecologic Oncology 125(2): 414-20; Mann & Cheng (2020) Exp Reviewof Anticancer Therapy 20(1):1-4; Silva et al (2019) Ann Transl Med 7(21):600].

3/13/2020 12:57 PM

6 We recommend that the section be modified to read as follows: "In patients with cancer typesother than CRC, GEA, small bowel, and endometrial being considered for checkpoint blockadetherapy, pathologists should test for DNA mismatch repair or MSI-H, although the optimalapproach for the detection of these has not been established. Note: Assays must be adequatelyvalidated analytically, including comparison to an orthogonal method, for the specific cancertype being tested". (Please clarify or consider removing the following statement, because itseems redundant to the already stated need for adequate validation: “with careful considerationof performance characteristics of MMR IHC and MSI by NGS for the detection of DNAmismatch repair defects.”)

3/13/2020 11:43 AM

7 Suggested Modification: In patients with cancer types other than CRC, GEA, small bowel, andendometrial being considered for checkpoint blockade therapy, pathologists should test for DNAmismatch repair, although the optimal approach for the detection of MMR defects has not beenestablished. For patients with advanced disease who may require additional genotyping, MSIby validated NGS assay may provide additional utility over MMR by IHC or MSI by PCR.

3/13/2020 11:36 AM

8 1. PCR testing is not mentioned in the Recommendation and should be... it should also becarefully validated.

3/13/2020 10:02 AM

9 1. Could be interpreted as a suggestion for MSI (i.e. DNA-based assay) over MMR testing. Add 3/13/2020 9:15 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

30 / 54

the word “defects” for clarity and consistency with other sections- “pathologists should test forDNA mismatch repair defects” 2. “Conditional recommendation” and forthcoming generaluncertainty within this section may be interpreted as weak evidence/recommendation for use ofMSI/MMR testing in “other cancers” and may not be read as uncertainty specific to testingmodality. To enhance clarity, suggest following the format used in other CAP Guidelines(https://www.cap.org/protocols-and-guidelines/current-cap-guidelines), where there is aseparate column indicating the “strength of recommendation” for testing 3. There is mention ofIHC and NGS, but not PCR. Suggest to include PCR

10 There is sparse data in this testing area for all modalities; MMR-IHC, MSI-PCR and NGS. Iwould like to see comparative studies of MMR-IHC vs MSI-PCR vs NGS-MSI in this setting todirect guidelines.

3/12/2020 4:58 PM

11 In our practice, the oncologists have requested MMR / MSI on every stage 4 solid tumor.Because of this, I take issue with the statement "Assays must be adequately validated for thespecific cancer type being tested..." This statement will essentially preclude most laboratoriesfrom being able to test for MMR / MSI for many cancer types. There are far too many differentcancer types to reasonably expect a full validation for each and every cancer type that may berequested.

3/12/2020 3:07 PM

12 Agree with this but PCR should be added as supplement to IHC where available, as many ofthese cancer types are HNPCC related (possibly more likely than CRC) and would help withscreening.

3/10/2020 3:36 PM

13 Uncommon tumors would be overly burdensome to validate, since mismatch repair deficiencyis so rare (especially for proteins like MSH2), and these tumors are tested so infrequently. Andwhat would be the reference method/gold standard? This recommendation would allow only thelargest few reference labs to do the testing. My recommendation would be for CAP to send outproficiency testing for non-GI/non-endometrial cancers.

3/10/2020 1:22 PM

14 For a clinical trial of basket design, which is open for all types of solid tumor or lymphoma, whatshould be the approach to adequately validated assay? Is the lab supposed to exhaust allpossible tumor types even if the sponsor cannot predict all of them before the trial?

3/9/2020 10:31 PM

15 Separate validation of each cancer type is prohibitive. Testing by MMR IHC and MSI by PCR ofthese cases could be considered.

3/9/2020 7:27 PM

16 We prefer to use only msi because of msi validation is only for crc. Espacialy in consideration ofthe testing costs and workload

3/6/2020 10:30 AM

17 Assays must be adequately validated for the specific cancer type being tested with carefulconsideration of performance characteristics of MMR IHC and MSI by PCR or NGS for thedetection of DNA mismatch repair defects.

3/6/2020 8:55 AM

18 For rare tumor types, this may not be possible, and testing with an orthogonal assay should beconsidered when the primary method is "negative" for MMR deficiency.

3/6/2020 8:33 AM

19 MSI by NGS or PCR 3/5/2020 2:41 PM

20 I do not think we should recommend such blanket testing without better evidence of how wellMMR IHC performs in other tumors. I also disagree with issuing a statement recommendingtesting but not giving some guidance as to testing methodology.

3/2/2020 9:11 PM

21 1. Please remove phrase: being considered for checkpoint blockade therapy. We need thesetests also to evaluate for LS, not just therapy evaluation. 2. Cannot validate for all tumor types.Tests are tumor agnostic. 3. Should offer PD-L1 testing as an alternative and bypass MMR MSItesting.

2/29/2020 12:54 AM

22 NGS assay might need a common standard, such as a set of markers and set the cut off. 2/28/2020 12:48 PM

23 this statement ommitts MSI by PCR - effectively the gold standard. Again I have concernsabout Laboratories submitting TMB results for MSI while using NGS as the detectiontechnology. Validation requirements should be specificfied.

2/24/2020 10:25 AM

24 There is insufficient evidence evaluating clinical validity of IHC vs. PCR vs. NGS to supportStatements 1-4. The performance characteristics of NGS assays vary depending on size ofpanel, informatics tool used, etc.; therefore, clinical decisions should be made depending on thevalidated performance of the individual assay.

2/22/2020 10:05 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

31 / 54

25 Might be worth modifying this to indicate that certain tumor types have other primarybiomarkers for selection of immunotherapy responders (e.g. PD-L1) in the first line setting. thismay render MMR/MSI testing less relevant- and indeed may lead to an additional unnecessarytesting expense if this recommendation is interpreted to mean that all patients should getreflexive MMR testing (because, to be honest, nearly all cancer patients are now beingconsidered for immunotherapy).

2/22/2020 8:01 AM

26 Individual labs will not have adequate numbers of cases to perform their own validations.Published large validation studies of specific cancer types should be used.

2/21/2020 2:21 PM

27 For non CRC/nonendometrial carcinomas the Promega MSI kit includes 5 markers that wereoptimized for analyzing microsatellite instability in CRCs. The sensitivity for detection of MSI-high is lower in endometrial cancers than CRC and lower than MMR IHC. In addition, there isnot sufficient data to recommend MSI over MMR IHC in these other types of tumors.

2/21/2020 11:22 AM

28 I do not think the data favors MMR IHC or MSI testing. The promega MSI in particular seemsoptimized for CRC and its performance is less characterized for these other cancer types.

2/21/2020 7:35 AM

29 The recommendation is for "patients being considered for checkpoint blockade therapy". Thatinformation is generally not available to the pathologists. The clinicians often demand reflexMMR testing for all cases, which is not a cost-effective approach. Please further elaborate andguide the pathologists to identify "patients being considered for checkpoint blockade therapy".

2/20/2020 6:38 PM

30 This recommendation should endorse NGS-based testing rather than leave open the questionof method. NGS panels can provide not only excellent coverage for MSI detection, but will turnup other (more common) alterations that may inform treatment, such as mutations in the DNAdamage repair genes (BRCA1/2, etc.). There is not reason to isolate MSI has a stand-alonetesting target when there are well-validated NGS-based assays that provide more all-aroundinformative coverage.

2/20/2020 3:29 PM

31 as above 2/20/2020 2:35 PM

32 Make it clear that if the patient qualified for therapy by another means, such as PDL1 status,then the pathologist should not be required to also test for mismatch repair status.

2/20/2020 10:54 AM

33 In this instance when "MMR" is being used as a specific IHC strategy, would clarify that MMR,MSI by PCR or MSI by other methods including NGS may be pursued providing appropriatevalidation studies.

2/20/2020 10:35 AM

34 Must include SNP based MSI assays as these may soon be FDA approved. 2/20/2020 8:10 AM

35 Validation in every tumor type is difficult if not impossible. I believe the assays are tumoragnostic. PCR may be preferable since it evaluates the end result of MMR.

2/20/2020 7:59 AM


2/19/2020 11:49 PM

37 It is impractical to validate for each cancer type being tested. 2/19/2020 7:03 PM

38 How can we validate if there is no evidence for the optimal approach? Need guidance on whata validation for uncommon or less common cancer types.

2/19/2020 3:19 PM

DisclaimerThe information, data, and draft recommendations provided by the College of American Pathologists are presented for informational and public feedback purposes only. The draft recommendations and supporting documents will be removed on March 20, 2020.The draft recommendations along with the public comments received and completed evidence review will be reassessed by the expert panel in order to formulate the final recommendations. These draft materials should not be stored, adapted, or redistributed in any manner.Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

32 / 54

66.92% 89

10.53% 14

4.51% 6

17.29% 23

Q7 Draft Recommendation Statement 5For all cancer patients beingconsidered for checkpoint blockade therapy based upon defective

mismatch repair, pathologists should not use TMB as a surrogate for thedetection of DNA mismatch repair defects. If a tumor is identified as TMB-

high, pathologists may perform IHC and/or MSI by PCR to determine ifhigh TMB is secondary to mismatch repair deficiency. Note: In patients

being considered for checkpoint blockade therapy, pathologists should notuse TMB as a surrogate based on DNA mismatch repair defects.(Quality

of Evidence: Low; Strength of Recommendation: Strong)Answered: 133 Skipped: 111

TOTAL 133

Agree aswritten



Neutral

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Agree as written



Neutral


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

33 / 54

# COMMENTS DATE

1 Clearly define TMB-high in the text. 3/13/2020 5:01 PM

2 We suggest rephrasing statement 5 to emphasize that while this is a strong recommendation, itis a strong recommendation NOT to use TMB as a surrogate.

3/13/2020 2:51 PM

3 We suggest the section be modified to read as follows: "For all cancer patients beingconsidered for checkpoint blockade therapy based upon defective mismatch repair or MSI-H,pathologists should not use TMB as a surrogate for the detection of DNA mismatch repairdefects or microsatellite instability. If a tumor is identified as TMB-high, pathologists may usevalidated MMR or MSI detection methods to determine if high TMB is secondary to dMMR orMSI-H. Note: In patients being considered for checkpoint blockade therapy, pathologists shouldnot use TMB as a surrogate for MMR or MSI status. "

3/13/2020 11:43 AM

4 Suggested Modification: For all cancer patients being considered for checkpoint blockadetherapy based upon defective mismatch repair, pathologists should NOT use TMB as asurrogate for the detection of DNA mismatch repair defects. In advanced cancer patients whohave not previously had MMR/MSI testing, NGS assays validated to determine both MSI andTMB simultaneously may be advantageous.

3/13/2020 11:36 AM

5 1. Remove the term 'defective' from he first sentence 2. There is no definition in the literature tosupport use the term "TMB-high" although most understand what it is 3. In the secondsentence, replace the words "pathologists may" with "pathologists must" perform...

3/13/2020 10:02 AM

6 The recommendation that IHC or PCR should be used over NGS for looking at TMB-high casesis not supported by the literature. It’s admitted that the evidence level is low, which is starkly atodds with the high strength of recommendation. Just in the references for these guidelines,there is ample evidence that IHC, PCR, and NGS have high concordance. In this setting, a testthat provides TMB will most likely provide MSI status, which makes another test unnecessary.The guideline should read that pathologists may use NGS results, perform IHC and/or MSI byPCR.

3/13/2020 9:22 AM

7 1. Deviation in consistency in terminology. Suggest including “MMR” prior to “IHC” in thesecond sentence

3/13/2020 9:15 AM

8 The definition of PCR methods to be used for the detection of DNA mismatch repair defectsshould be not limited to the standard NIH panel or Promega MSI panel but broadened toinclude any validated molecular method, e.g. other PCR methods like the Idylla™ MSI Assay.The Idylla™ MSI Assay has shown excellent concordance to other molecular methods as wellas IHC (References: Li et al. (2019) Clin Colorectal Cancer, pii: S1533-0028(19)30086-6;Samaison (2019) Clin Pathol. 72(12):830-835; Maloney et al. (2018) TT054, J. Mol. Diagn.20(6), 895–1039; Maertens et al. (2017) Annals of Oncology 28 (suppl_5): v22-v42; De Craeneet al. (2018) Journal of Clinical Oncology 36:15_suppl,e15639; De Craene B. et al. (2018)Annals of Oncology 29 (suppl_8): viii14-viii57; De Craene et al. (2017) Annals of Oncology 28(suppl_5): v209-v268; Nicka et. al. (2018) J. Mol. Diagn. 20(6), 895–1039; Nafa. et.al. (2018) J.Mol. Diagn. 20(6), 895–1039), is used in a large number of laboratories (> 40 in the US) andhas been validated and approved for clinical use at MSKCC by the New York State Departmentof Health Clinical Laboratory Evaluation Program (project ID: 68841;https://www.wadsworth.org/print/82410). We propose to change the first paragraph as follows:Mismatch repair (MMR) Immunohistochemistry (IHC) and microsatellite instability (MSI)-polymerase chain reaction (PCR; defined as standard NIH panel of 5-7 microsatellites or othervalidated PCR-methods such as Promega MSI panel and Idylla MSI panel; most clinicallaboratories in the US are using one of these panels).

3/13/2020 5:05 AM

9 Include “follow the testing guidelines for organ-specific cancers”. 3/12/2020 4:58 PM

10 Clarification is needed if pathologist is mandated to review NGS reports and reflex high TMBcases for MMR IHC

3/9/2020 12:53 PM

11 If a tumor is identified as TMB-high, pathologists should perform IHC and/or MSI by PCR todetermine if high TMB is secondary to mismatch repair deficiency.

3/6/2020 8:55 AM

12 If TMB-high and test for DNA MMR defects is normal, should consider defects of proof-readingpolymerases.

3/5/2020 2:41 PM

13 Can MSI be performed by a panel other than the NIH 7-microsatellite panel? Some studiessuggest for prostate cancer that a larger number of microsatellites is better. the UW MSINGS

3/2/2020 9:19 AM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

34 / 54

MSI NGS is also more comprehensive.

14 1. Please remove phrase: being considered for checkpoint blockade therapy. We need thesetests also to evaluate for LS, not just therapy evaluation.

2/29/2020 12:54 AM

15 should read IHC and PCR as you have no published data either way! 2/24/2020 10:25 AM

16 Although MSI and TMB are correlated, high TMB can result from multiple mechanisms ofmutagenesis and should not be used as a surrogate for MSI. These are different biomarkers.

2/22/2020 10:05 AM

17 Or may use data from a validated NGS panel to inform likelihood of MSI-G status 2/22/2020 8:01 AM

18 Very low evidence and often high inconsistency with cutoff being considered for positive ornegative results.

2/21/2020 7:39 AM

19 Please spell out "TMB": what does that stand for? How to determine TMB? 2/20/2020 6:38 PM

20 The phrase 'based upon defective mismatch repair' implies that there are many ways to qualifyfor IO therapy, and indeed there are. If the goal of this guideline is to remain focused on use ofMMR tests to qualify pts for IO Rx, then you can skip discussion of TMB or merely state thatTMB status is not a test for MMR deficiency and thus is outside the scope of this guideline.Likewise PDL1 IHC and EBV status and other predictors of IO Rx response are outside thescope of this guideline.

2/20/2020 10:54 AM

21 "TMB-high" may be used to guide additional testing by MMR-IHC or MSI-PCR/NGS but shouldnot be used as a definitive guide to I/O therapy recommendations.

2/20/2020 10:35 AM

22 There is insignificant clinical data to support TMB testing at this time. 2/20/2020 8:10 AM

23 "TMB" is an abbreviation and requires a explanation in the text 2/20/2020 7:57 AM

24 Depends on the PCR and the laboratory should have a external control to check its testingpreformance. The sensitivity should be validated and the testing very well standarized in tissueselection and procedure

2/20/2020 7:49 AM

25 Not always required 2/20/2020 5:06 AM

26 Consider spelling out "Tumor Mutational Burden" before using the abbreviation. 2/19/2020 11:23 PM


d Afte

r Marc

h 13,

2020


Recommendations

35 / 54

67.44% 87

20.16% 26

1.55% 2

10.08% 13

Q8 Draft Recommendation Statement 6For cancer patients beingconsidered for checkpoint blockade therapy, if a MMR deficiency is

identified, pathologists should recommend follow-up evaluation for LynchSyndrome(Quality of Evidence: Low; Strength of Recommendation:

Strong) Answered: 129 Skipped: 115

TOTAL 129

Agree aswritten



Neutral

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Agree as written



Neutral



Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

36 / 54

# COMMENTS DATE

1 For all tumors with MMR deficiency, the cause of such deficiency should be determined if at allfeasible. In most cases, this will be due to purely somatic events. Only when a somatic cause ofMMR defiency cannot be identified should evaluation for Lynch syndrome occur. I think thisrecommendation should be re-phrased to reflect the relative contributions of sporadic versusLS-associated alterations to MMR deficiency across tumors. Furthermore, evaluation for Lynchsyndrome should not be linked to consideration for checkpoint blockade therapy. If a tumor iffound to be MMR deficient, a cause should be sought, regardless of therapeutic considerations.

3/13/2020 8:14 PM

2 This seems reasonable so long as the text includes some commentary on what a suitableevaluation might entail in common contexts

3/13/2020 5:01 PM

3 Absolutely! This is very important for the patient's biological family members. 3/13/2020 3:47 PM

4 We suggest adding (text in capital letters) “… if a MMR deficiency is identified IN ATUMOR/SOMATIC SAMPLE, …” for better understanding/accuracy of Statement 6.

3/13/2020 2:51 PM

5 This is extremely vague and doesn't really provide any clear guidelines/recommendations otherthan follow-up. Would really help if you were to what follow-up should be done.

3/13/2020 12:57 PM

6 We recommend the section be modified to read as follows: "For cancer patients beingconsidered for checkpoint blockade therapy, if a MMR deficiency is identified, pathologistsshould recommend follow-up evaluation for Lynch Syndrome including genetic counseling andgermline testing if indicated."

3/13/2020 11:43 AM

7 The language here is quite vague. A guideline of recommended follow-up should be included. 3/12/2020 4:58 PM

8 Should recommend germline testing for Lynch syndrome, rather than follow-up which is not welldefined.

3/10/2020 4:37 PM

9 For cancer patients being considered for checkpoint blockade therapy, if a MMR deficiency isidentified, pathologists should recommend follow-up evaluation for Lynch Syndrome. But if it isMLH-1 and PMS2 are deficient, evaluating MLH-1 promoter methylation is warranted beforesubjecting patient to Lynch gene genetic sequencing.

3/9/2020 9:31 PM

10 ... in consideration of the age, crc-subtype or other important informations concerning thebethesda criteria.

3/6/2020 10:30 AM

11 If no MLH1 hypermethylation and/or presence of BRAF V600E mutation when loss of MLH1expression on IHC.

3/5/2020 2:41 PM

12 1. Please remove phrase: being considered for checkpoint blockade therapy. We need thesetests also to evaluate for LS, not just therapy evaluation. 2. Need to recommend MLH1 reflex toBRAF testing. If Neg, then perform genetic testing.

2/29/2020 12:54 AM

13 by ordering additional tests such as MLH1 hypermethylation or BRAF mutation testing. 2/28/2020 1:16 PM

14 In endometrial cancer testing should be done for hypermethylation if this was detected on IHC 2/28/2020 9:38 AM

15 If the tumor type has been associated with Lynch syndrome and if there are no other features tosuggest a somatic etiology, eg. MLH1 promoter methylation

2/22/2020 4:51 PM

16 In the Cancer Care Ontario guidelines MLH1/PMS2 deficient tumour patients are only referredfor genetic counselling assessment after negative methylation/BRAF testing.

2/22/2020 9:32 AM

17 Pathologists should take some initiative to rule out sporadic forms of MSI-H in particular MLH-1promoter methylation in endometrial and CRC patients with MLH1 loss. perhaps this needs toreworded as "should follow existing recommendations regarding exclusion of sporadic casesand followup for LS"

2/22/2020 8:01 AM

18 Need to do BRAF testing snd mlh1 methylation testing first 2/22/2020 2:33 AM

19 This recommendation needs some caveats. I think its fairly standard that in the case of PMS2and MLH1 loss that BRAF and MLH1 testing will be performed as most of the tumors with thisstaining pattern are sporadic tumors and likely do not require further workup for Lynchsyndrome.

2/21/2020 9:05 AM

20 Specific recommendations should be made as to the type of follow up, i.e. in CRC vs. other. 2/20/2020 9:30 PM

21 Follow-up evaluation of Lynch syndrome appears to be outside of a pathologist's power. This 2/20/2020 6:38 PM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

37 / 54

should be done by medical geneticist or oncologist. In addition, this recommendation should notbe limited to "cancer patients being considered for checkpoint blockade therapy".

22 Why does this say MMR deficiency? It should be MMR deficiency or micro-satellite instabilityhigh.

2/20/2020 2:11 PM

23 depending on age 2/20/2020 12:03 PM

24 This recommendation should be considered in the context of all clinicopathological informationavailable.

2/20/2020 10:35 AM

25 My understanding (which is probably outdated) is that the cost-benefit of diagnosing LynchSyndrome has not been well established, particularly outside the colon. I wonder whether thestrength of this recommendation should be downgraded to "conditional" or "moderate."Alternatively, the suggestion could be for pathologists to "discuss" (rather than "recommend")evaluation for Lynch Syndrome.

2/19/2020 11:23 PM

26 Follow - up is ideal but comment should state "when such services are available" 2/19/2020 10:34 PM

27 Genetic counseling referral should be offered on these cases 2/19/2020 7:04 PM

28 Lynch Syndrome follow up testing recommendations should be specific and stronglyrecommended.

2/19/2020 7:03 PM

29 This is highly dependent on tumor type. Colon and endometrial have well established algorithmis place that help identified sporadic versus lynch associated MMR deficiency. Other tumortypes we really don’t have a good adjunct test to help make that distinction so I feel making thisrecommendation may lead to over reliance of MLH1 promoter methylation assays outside ofcolon and endometrial cancers. I have recently looked into the performance of MLH1 promotermethylation assays outside of these common lynch related cancers and there doesn’t seem tobe enough evidence to understand what those results may mean. If you are going to make thisrecommendation then I would contest that the statement must include something on the adjunctworkup for sporadic versus lynch associated.

2/19/2020 2:42 PM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

38 / 54

62.31% 81

8.46% 11

11.54% 15

17.69% 23

Q9 Good Practice Statement 1Discordant results:In the event ofdiscordant results, pathologists should interpret any evidence of MMR

deficiency by IHC, MSI by NGS or PCR as a positive result for patients tobe eligible for checkpoint blockade therapy.


TOTAL 130

Agree aswritten



Neutral

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Agree as written



Neutral


d Afte

r Marc

h 13,

2020


Recommendations

39 / 54

# COMMENTS DATE

1 This statement is overly prescriptive and broad. Pathologists should use their best professionaljudgment and their understanding of the technical strengths and limitations of each assay, asperformed in their own laboratories, in conjunction with information about the specific tumortype and sample characteristics to provide an integrated assessment.

3/13/2020 8:20 PM

2 Due to the variability of the quality of implementation of IHC in many laboratories, I would becautious of making a decision regarding checkpoint blockade therapy based on IHC alone.

3/13/2020 3:50 PM

3 MMR by IHC has been shown to have a false negative rate of 5-10% [Funkhouser et al. (2012)J. Mol Diagn 14:91-203; Dudley et al. (2012) Clin Cancer Res 22 813-20]. We therefore believethat due to the complementary nature of dMMR by IHC and MSI by PCR technologies and thepotential impact of misdiagnosis, there is a growing recognition in the field that these testsshould be performed together for maximal sensitivity when identifying patients for hereditarycancer risk and immunotherapy eligibility [Goodfellow et al. (2015) J Clin Oncol 33:4301-8;Leenen et al.(2012) Gynecologic Oncology 125(2): 414-20; Mann & Cheng (2020) Exp Reviewof Anticancer Therapy 20(1):1-4; Silva et al (2019) Ann Transl Med 7(21):600].

3/13/2020 12:59 PM

4 Pathologists should rely on tests that are FDA-approved and have clinical evidence of efficacy 3/13/2020 10:32 AM

5 Evidence of dMMR by gold standard methods should be included. 3/12/2020 5:03 PM

6 But after excluding specimen, technical or analytical error. 3/10/2020 3:38 PM

7 I think that there should be some mention of repeat testing or different block. 3/10/2020 8:16 AM

8 In the event of discordant results, pathologists should interpret any evidence of MMR deficiencyby IHC, MSI by PCR as a positive result for patients to be eligible for checkpoint blockadetherapy. MSI by NGS is always secondary option due to higher cost.

3/9/2020 9:33 PM

9 Seems to indicate that an orthogonal method such as ddPCR is needed to confirm the result.Perhaps combine with Good Practice Statement 2 to clarify.

3/9/2020 4:43 PM

10 I'm not sure about "any evidence". What if it is a difficult to interpret 'patchy' IHC result, reflexedto MSI, which is normal? Do you go with the uncertain IHC? Perhaps "any clear evidence ofMMR deficiency" or "any analytically valid evidence of MMR deficiency" could substitute for"any evidence"

3/3/2020 9:44 PM

11 We should recommend that testing be repeated either on a different sample of the tumor (incase of resections) or possibly at a different lab.

3/2/2020 9:12 PM

12 Need to also include PD-L1 IHC testing for therapy, not just dMMR. 2/29/2020 1:01 AM

13 Is there evidence that normal staining for MMR proteins by IHC is trumped by MSI by NGS?? 2/28/2020 1:17 PM

14 Should read MSI by PCR or NGS. Again NGS assay must be appropriately validated - seeFoundation Medicine PMA subumission

2/24/2020 10:27 AM

15 Pathologist should evaluate the performance characteristics of each assay and assess allevidence in total to make an accurate interpretation of MMR/MSI based on IHC, MSI, and NGS.If the pathologist concludes there is a false positive result in one of the assays, which can bedue to multiple factors, the patient should not receive immunotherapy.

2/22/2020 10:09 AM

16 Not sure about this. IHC is known to suffer from false positive results due to technical issues.PCR can be difficult to interpret. NGS across tumor types that are not well characterized in theMSI spectrum may also be difficult to interpret. These isolated positive results should bereported with a caveat so as not to give cinicians and patients a false hope of possibility ofresponse When NGS data is available, they should be interpreted in the context of othergenomic data such as TMB.

2/22/2020 8:06 AM

17 Has this question really been answered? Any evidence is extremely loose wording. Clinicalinput is needed. Essentially a cinicopathological decision not a definition on paper,

2/21/2020 7:07 AM

18 It seems to be a biased recommendation that every effort be made to get a positive result. 2/20/2020 6:44 PM

19 Poorly worded. "Discordant" should be clarified. 2/20/2020 3:31 PM

20 Reason for discordance must be, at least plausibly, determined. The discordance must bedisclosed to the treating oncologist who can then act accordingly.

2/20/2020 2:27 PM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

40 / 54

21 This statement should explain the recommendation further. It is acknowledged that both MMRIHC and MSI testing have limitations that will lead to false negative results. ~

2/20/2020 2:14 PM

22 There should be no discrepancy, in results . Existing guidelines must be followed 2/20/2020 5:08 AM

23 Technical limitations sometimes preclude unequivocal results. If so, am additional biopsy orrepeat testing should be done

2/19/2020 7:07 PM

24 If results appear valid - don’t you want to do sone double check? 2/19/2020 3:20 PM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

41 / 54

78.46% 102

14.62% 19

0.00% 0

6.15% 8

Q10 Good Practice Statement 2Indeterminate:In the event ofindeterminate result in any method, pathologists should perform an

orthogonal technique or repeat on a different tumor block. Laboratoriesshould have a robust peer review process for indeterminate cases.


TOTAL 130

Agree aswritten



Neutral

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Agree as written



Neutral


d Afte

r Marc

h 13,

2020


Recommendations

42 / 54

# COMMENTS DATE

1 Orthogonal analysis should, in most cases, be preferred over simply performing the sameanalysis on a separate tumor block.

3/13/2020 8:20 PM

2 If you want to say "robust" then I think you should indicate what you would consider "robust" 3/13/2020 5:09 PM

3 I suggest performing a molecular method either by a different method (validated NGS vs. PCR)or by a different laboratory. Using IHC to confirm molecular results is too dependent on thequality of implementation of IHC.

3/13/2020 3:50 PM

4 We suggest providing further details on how the orthogonal technique should be performed. Werecommend to provide further details on how testing methods should be sequentially used in anevent of indeterminate result.

3/13/2020 2:52 PM

5 Orthogonal testing must meet consensus standards. 3/12/2020 5:03 PM

6 remove the word "robust" as it is not defined 3/10/2020 1:25 PM

7 Please include MMR IHC interpretation guidelines. 3/9/2020 8:06 PM

8 Provide an order: Repeat on different block (if available) Consider orthogonal assay Bigquestion: should an indeterminate result move forward or not?

3/9/2020 12:11 PM

9 This probably addresses my concern above 3/3/2020 9:44 PM

10 In the event of indeterminate result in any method in GI cancer pateint samples, pathologistsshould perform an orthogonal technique or repeat on a different tumor block. Laboratoriesshould have a robust peer review process for indeterminate cases.

3/2/2020 10:49 AM

11 Define orthogonal. Also, if repeat is again indeterminate, then what? Need to be more precisein statement.

2/29/2020 1:01 AM

12 ideally this a default or reflex process rather than asking for a new order to be placed by theclinician, which tends to occur only after clinician asks about results and is informed that it wasQNS or indeterminate. This is not timely for patient care

2/28/2020 9:10 AM

13 Sometimes different tests are perfumed in different labs without each lab receiving the otherresults or knowing the test is being carried out. Unfair burden of responsibility on thepathologist. Should be the responsibility of whoever is requesting and coordinating the testing.

2/24/2020 11:38 PM

14 I agree with orthogonal testing, but not necessarily with different tumor block; unless it is thesame tumor in separate blocks.

2/22/2020 12:23 AM

15 Please specify what "orthogonal technique" means. 2/20/2020 6:44 PM

16 Careful not to call actual discordant results as indeterminate... 2/20/2020 2:27 PM

17 Explain what you mean by peer review in the text that will accompany this guideline 2/20/2020 10:54 AM

18 We need to acknowledge there are borderline (indeterminate) cases present. 2/20/2020 8:22 AM

19 I am not sure the term "orthogonal technique" is apt. First of all, many readers will notunderstand it. Secondly, the various tests that can be used are not entirely orthogonal -- theyhave some interdependence. Perhaps the term "alternative technique" or "alternative method"would be a better choice.

2/19/2020 11:25 PM

20 Or additional biopsy obtained. Proper FFPE fixation often overcomes technical problems. 2/19/2020 7:07 PM

21 Please define robust and give examples of acceptable peer review process. 2/19/2020 2:45 PM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

43 / 54

55.04% 71

8.53% 11

13.95% 18

22.48% 29

Q11 Good Practice Statement 3Clonal loss:In the event of a clonal loss byMMR IHC, pathologists should perform MSI by PCR specifically in a

dissected area of tumor that has IHC loss of MMR deficiency.Clonal lossis defined as a complete loss of MMR protein in tumor cells with adjacentnormal cells having intact expression. Adjacent areas of tumor have intact

expression of MMR protein.Answered: 129 Skipped: 115

TOTAL 129

Agree aswritten



Neutral

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Agree as written



Neutral


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

44 / 54

# COMMENTS DATE

1 This IHC pattern is commonly referred to as "subclonal" rather than "clonal". Wording by CAP inthis statement is needlessly confusing. Additionally, the overall value of performing such testingis uncertain. Even if MSI testing shows an MSI-H phenotpye in the dissected area of IHC loss,how should this be interpreted for the tumor overall? This statement should either be revised oreliminated.

3/13/2020 8:20 PM

2 Does this statement add anything beyond Good Practice Statement 1? Wouldn't clonal loss byMMR IHC satisfy "any evidence of MMR deficiency."

3/13/2020 5:09 PM

3 We suggest considering including NGS as an alternative technique when discussing clonalloss.

3/13/2020 2:52 PM

4 Suggested Modification: Clonal loss: In the event of a clonal loss by MMR IHC, pathologistsshould perform MSI by PCR and/or MSI by validated NGS assay specifically in a dissectedarea of tumor that has IHC loss of MMR deficiency. Clonal loss is defined as a complete loss ofMMR protein in tumor cells with adjacent normal cells having intact expression. Adjacent areasof tumor have intact expression of MMR protein. Note: MSI by NGS assay must be validatedagainst MMR IHC or MSI by PCR and must show equivalency

3/13/2020 11:40 AM

5 Only focal, weak equivocal nuclear staining is still acceptable under the designation of Loss.Recommend amending the last phrase with the following: " Focal weak equivocal nuclearstaining in the viable tumor cells in the presence of internal positive controls is acceptablewithin Clonal Loss." Also recommend that the last sentence be removed since therecommendation is to take the tissue sample from the area of IHC loss. Having the lastsentence makes the statement confusing.

3/13/2020 10:32 AM

6 With reference to gold standard testing modalities. 3/12/2020 5:03 PM

7 Few labs have the ability to microdissect tumor and do MSI testing 3/10/2020 1:25 PM

8 I am not sure of the evidence supporting this, if we have a loss of MMR IHC in general, we arenot performing a confirmatory MSI PCR, why would we do it here. It is clonal, the question iswhether or not this person should qualify/receive checkpoint therapy. I do not know theevidence for this, but think that MSI PCR with microdissection of the tumor with IHC loss will notadd anything but additional confirmatory testing of what we already know. It is clonal.

3/10/2020 8:16 AM

9 That isn't necessary. Loss of expression of MMR protein is evidence of deficient mismatchrepair.

3/9/2020 7:33 PM

10 This will be difficult to implement in many places 3/9/2020 1:02 PM

11 Might expand on clonal loss: "...clonal loss (including intraglandular loss, clonal loss, and/orcompartmental loss)..."

3/9/2020 12:25 PM

12 I am confused by the definition: do you mean subclonal with loss in a certain tumor area? 3/9/2020 12:11 PM

13 we prefer to test distant metastasis because of the more likely non heterogenoustumorspecimen compared to the primary tumor. This is because only the metastasis is thebiologic active and prognostic relevant part of the tumor.

3/6/2020 10:34 AM

14 Ridiculous recommendation. Poor fixation frequently results in loss of MMR IHC staining. Also,it is totally impractical to microdissect these areas for MSI.

2/29/2020 1:01 AM

15 Once again, is assuming all platforms for testing are in the same lab. Not so in some centres 2/24/2020 11:38 PM

16 shouldn't this read "subclonal loss"? there is no great data to suggest that a subclonal patternwill predict response. Why is PCR necessary in this setting?

2/22/2020 8:06 AM

17 MSI by PCR in dissected area IF FEASIBLE. 2/21/2020 2:23 PM

18 MSI by PCR or by NGS validated assay 2/21/2020 7:07 AM

19 There is no evidence on how cancers with clonal loss will respond to ICI treatment. 2/20/2020 9:32 PM

20 I am not whether this is necessary. If IHC is negative, that is considered a positive result. Thisrecommendation appears contradictory to Statement 1. In addition, the phrase "adjacentnormal cells" appear to indicate "adjacent tumor cells".

2/20/2020 6:44 PM

21 Would include MSI by NGS as an equally good approach to determining whether clonal loss is 2/20/2020 3:31 PM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

45 / 54

clinically relevant.

22 Seems like a dumb recommendation. Has anyone showed that this is valid or has ever beenvalid, just in one case?

2/20/2020 2:14 PM

23 MSI can already be detected by PCR if normal tissue is also present. 2/20/2020 2:10 PM

24 why do you use the word clonal, do you mean subclonal or focal? What is the evidence that ptsw focal loss of a MMR protein benefit from IO Rx?

2/20/2020 10:54 AM

25 Consider alternative approach for those cases with insufficient tumor tissue for further testing. 2/20/2020 10:36 AM

26 When there is no normal tissue present, performing PCR is a problem 2/20/2020 8:22 AM

27 i don't know if this is practical, especially with the variable sensitivity of some of the MMRantibodies.

2/20/2020 8:04 AM

28 Difficult to perform in all labs 2/20/2020 2:43 AM

29 What are the implications of clonal loss? Generally not Lynch and perhaps not eligibility forimmune drugs. More details and support needed for this statement

2/19/2020 7:07 PM

30 This statement is unnecessary because according to good practice 1, a tumor is to beconsidered MMR deficient if either the IHC, PCR, or NGS are abnormal. In this specificscenario, the IHC is abnormal within the clonal area and thus should be considered MMR-deficient in the clone. I would suggest that clonal loss should be considered MMR-deficient forcheckpoint blockade purposes but should be considered somatic for Lynch Syndromescreening purposes.

2/19/2020 2:52 PM

31 We do not perform the MSI by PCR in house so the extent of any macro dissection is out of myhands. I can not assure that the reference labs will be doing this so I would Like that removefrom the statement. I have ordered MSI by PCR on these clinal loss cases and resulted a MSIhigh based on bulk tumor.

2/19/2020 2:45 PM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

46 / 54

59.02% 72

40.16% 49

0.82% 1

Q12 How feasible is it to implement this guideline?Answered: 122 Skipped: 122

TOTAL 122

All of it isfeasible to...

Parts of itare feasible...

None of it isfeasible to...

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


All of it is feasible to implement.

Parts of it are feasible to implement.

None of it is feasible to implement.


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

47 / 54

# COMMENTS ABOUT THE FEASIBILITY OF IMPLEMENTING THE GUIDELINE: DATE

1 Guidelines on MMR testing are certainly needed. However, these guidelines, as written, seemsignificantly off the mark, and do not appear to reflect testing and assay performance as itcurrently stands in 2019-2020.

3/13/2020 8:27 PM

2 Several fixes are indicated 3/13/2020 10:40 AM

3 Not all validated PCR tests that are used today were mentioned. 3/13/2020 5:08 AM

4 Need to support the guidelines with references exploiting gold standard testing (4 IHCimmunostains and MSI with mononucleotide biomarkers) for colorectal and endometrial tumors.Sparse data is available for other tumor sites for all testing modalities and for NGS approaches.

3/12/2020 5:11 PM

5 NGS is not available at all labs, nor are both MSI and IHC, so may need collaboration or newtest development to accomplish.

3/10/2020 4:40 PM

6 It gives options regarding choice of test; seems feasible to me 3/3/2020 9:46 PM

7 As this becomes essentially SOC, most if not all labs will have the capability to test via IHC andPCR at the very least

2/28/2020 1:18 PM

8 It should be feasible. There will likely remain hindrance to the clinician of noncolorectal/GE/endometrial patients who need this testing done more routinely

2/28/2020 9:11 AM

9 Availability and funding of testing beyond MMR IHC will be problematic in some systems, I amsure.

2/22/2020 9:36 AM

10 Most laboratories have no problem in performing IHC. PCR or NGS based testing can be aproblem - it is costly, not widely available, and has a longer turnaround time. Therecommendations puts all burdens on pathologists. The biggest challenge for a pathologist ishow to identify "patients being considered for checkpoint blockade therapy". This is notaddressed in the recommendation.

2/20/2020 6:52 PM

11 In a discordant there should be a Reference Laboratory, to confirm the results. It is preferablyone per country that could do this test to solve difficult cases. That remains to be defined byeach country. It is necessary to have external quality evaluation surveys and peer comparisonfor this relative new testing

2/20/2020 9:24 AM

12 Depends on easy access to this testing specially in countries like ours( India) 2/20/2020 5:11 AM

13 I do not think that good practice #3 is necessary 2/19/2020 2:53 PM


d Afte

r Marc

h 13,

2020


Recommendations

48 / 54

37.61% 41

11.93% 13

19.27% 21

11.93% 13

19.27% 21

13.76% 15

34.86% 38

7.34% 8

20.18% 22

Q13 What barriers might impede adoption of the final guideline? (Chooseall that apply.)



Disagreementwith the dra...

Disagreementwith how the...

Too burdensome

Lack ofsupport from...



Lack ofresources...

Do not wish togive up...


0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Disagreement with the draft recommendations

Disagreement with how the guideline was developed

Too burdensome

Lack of support from administration

Lack of support from other members of the medical team

Lack of support from the community (others outside your institution e.g., patients, industry)

Lack of resources (funding)

Do not wish to give up personal autonomy to follow the guideline



Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

49 / 54


1 Resources for genetic counseling when referring patients for Lynch syndrome screening can bevery limited/non-existent.

3/13/2020 3:52 PM

2 Any discordant recommendation of the FDA-approved package insert protocol(s) andmanufacturers Interpretation Guide

3/13/2020 10:40 AM

3 Please see comments 3/12/2020 3:08 PM

4 Confidence of the data behind the guideline is critical. Is there evidence based information,particularly on endometrial cancer and TMB.

3/10/2020 4:40 PM

5 Time to adopt the practice guidelines. 3/10/2020 8:19 AM

6 This is a general guideline which is definitively necessary. Some details, like adequatelyvalidated, should be elaborated or clarified if possible.

3/9/2020 10:38 PM

7 Lack of communication at the time of test request so pathologist can assist in appropriate testselection.

3/9/2020 12:29 PM

8 Whatever barriers are identified, they need to be removed to make this a standard practice 3/8/2020 9:38 PM

9 Insufficient tumor amounts in what will likely be a population made up predominantly ofbiopsies.

3/2/2020 9:13 PM

10 Recommendation by NCCN and/or ASCO. 2/25/2020 6:17 PM

11 Lack of resources for NGS based testing of MSI status because of increased cost and TAT 2/24/2020 1:23 PM

12 Insurance coverage policies need to be rewritten 2/24/2020 10:33 AM

13 Lack of literature evidence to support recommendations. 2/22/2020 10:10 AM

14 NA 2/21/2020 7:42 AM

15 Overall, I agree with the guidelines and think they should be implemented. I am not convincedMSI is better than MMR IHC for the non-CRC, endometrial, etc cancer types

2/21/2020 7:38 AM

16 Need clarification on the when and which sample (biopsy vs resection )needs testing 2/20/2020 7:55 PM

17 Insurance (incl. medicare) coverage can be a problem. 2/20/2020 6:52 PM

18 lack of available tissue is an issue 2/20/2020 10:54 AM

19 Standarization is an issue, and not all PCR are equal in preformance. Avalable Referencelaboratories in every region are necessary for some cases

2/20/2020 9:24 AM

20 As pathologists, we must embrace changes and replacement of methods for what we dosometimes.

2/20/2020 8:12 AM

21 Clinicians want MMR IHC performed on cell blocks from pancreatic fine needle aspirations.Evidence for this practice is scant and predominantly presented as abstracts from the 2019American Society of Cytopathology. This topic needs to be addressed.

2/20/2020 8:03 AM

22 The community pathology labs performing MMR IHC will have difficulty ensuring all of thesestatements CNA be enacted. Also, it is not our responsibility to know which patients are beingconsidered for immunotherapy. We really cannot be held responsible for an oncologist orderingan NGS test for a commercial lab and then being expected to work up results in our communitylabs. This burden should fall on the NGS and molecular labs performing the actual assays.

2/19/2020 2:50 PM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

50 / 54

47.06% 56

26.89% 32

25.21% 30

42.86% 51

55.46% 66

21.01% 25

9.24% 11

Q14 What facilitators might assist in your adoption of the final guideline?(Please select your top 3 facilitators.)



If leaders ofthe medical...

If there weretools to hel...

If we areforced to...

If we findthat peer...

If othertrusted...

If we know andtrust the...


0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


If leaders of the medical staff discussed adoption/adaption of the guideline for our practice setting

If there were tools to help implement the guideline

If we are forced to comply with the guideline by administration or an accreditation body

If we find that peer institutions/practices adopt the guideline

If other trusted organizations endorse the guideline

If we know and trust the members of the panel members and/or organizations who developed the guideline



d Afte

r Marc

h 13,

2020


Recommendations

51 / 54


1 If the guidelines were significantly revised. 3/13/2020 8:27 PM

2 Low cost / high benefit ratio, particularly with regard to early detection of Lynch syndromepatients and their family members.

3/13/2020 3:52 PM

3 If the guidelines are brought into alignment with the current standard of care and evidence, thenadoption is more likely.

3/13/2020 9:23 AM

4 Our institutions pretty much follow this practice already. 3/10/2020 8:19 AM

5 already implemented 3/9/2020 12:11 PM

6 Need Medicare approval, otherwise no traction in the market to implement therecommendations.

2/29/2020 1:02 AM

7 Need to get NCCN/ASCO to endorse guideline as improved testing (IHC and PCR according toEU guideline) is focussed on gaining access to life saving treatment for their pateints

2/24/2020 10:33 AM

8 If the panel can provide evidence for their recommendations, or if there is prospective clinicalevidence to show how different analytical methods ultimately impact patient care throughclinical trials.

2/22/2020 10:10 AM

9 Improved method to detect MSI 2/20/2020 2:37 PM

10 Due to clinician demand for MMR testing on cytology specimens, a comment from theAmerican Society of Cytopathology is needed.

2/20/2020 8:03 AM

11 Billable CPT codes for review of outside lab results or QA of outside molecular test results. 2/19/2020 2:50 PM


Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

52 / 54

Q15 Please provide any general comments or concerns in the spacebelow:



Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

53 / 54

# RESPONSES DATE

1 I work in a large academic laboratory that was one of the first in the country to offer MMRanalysis by IHC. We have cumulatively profiled many thousands of tumor by at least twomethods for assessing MMR deficiency. In our published and unpublished experience, NGSperforms at least as well as IHC and MMR testing by PCR. However, all three assays haveseparate individual strengths and weaknesses. In day to day practice, we routinely rely on theavailability of all three approaches, and I think this autonomy and robustness should beavailable to all practicing pathologists. These proposed guidelines are overly focused ontraditional techniques that were initially designed to detect Lynch syndrome in CRC patients. It'snot at all clear that these assays are the optimal choice for screening non-CRC patients foreither LS or immune checkpoint inhibitor therapy.

3/13/2020 8:27 PM

2 There appears to be a bias against NGS based testing for MSI without sound evidence that thismethod is not equivalent in sensitivity or specificity. While PCR and MMR methods have beenused in a greater number of studies, this is solely based on historical reasons not analyticalreasons. The guidelines should embrace current technologies and demonstrate thoughtfulnessin terms of tissue stewardship.

3/13/2020 3:01 PM

3 We would suggest providing specific guidance on somatic MMR/MSI test result reporting. 3/13/2020 2:52 PM

4 As a CRC organization, we chose neutral for the draft recommendation guidelines that were notrelated to CRC.

3/13/2020 1:17 PM

5 1. Please include topical references 2. Please use consistent terminology throughout thedocument 3. Define MSI status (e.g. MSI-H vs. non-MSI-H)

3/13/2020 10:40 AM

6 1. Lack of references and therefore rationale for recommendations. Consider annotating andciting key research supporting recommendations 2. Lack of suggestion for standardizeddefinitions, particularly since several different definitions of MSI are used within the literature.Consider defining MSI-H and MSI-low—in line with the ESMOGuidelines(https://www.ncbi.nlm.nih.gov/pubmed/31056702); consider incorporating MSI-lowwithin MSS tumors and endorsing the terms “MSI-H and non-MSI-H” 3. Detection by NGS ismore sensitive than traditional PCR and can simultaneously capture MMR gene deficiencies aswell as yield information for BRAF, KRAS, and NRAS. If NGS is placed in the recommendationsas currently proposed, NGS testing may encounter reimbursement barriers, despite providingmore robust information. Consider revision to equal recommendation for PCR and NGS for MSIdetection \ 4. There is an overreaching recommendation to use MMR IHC and/or MSI byPCR/NGS, however, this may lead to reimbursement barriers if interpreted as the need for twopositive tests. Consider addressing in footnote

3/13/2020 9:28 AM

7 If the guidelines limit molecular MSI testing to the standard NIH panel or Promega MSI panel,customers who have implemented an alternative molecular MSI testing method like the Idylla™MSI Assay, will not longer be able to use them for clinical testing. The Idylla™ MSI Assay hasshown excellent concordance to other molecular methods as well as IHC (References: Li et al.(2019) Clin Colorectal Cancer, pii: S1533-0028(19)30086-6; Samaison (2019) Clin Pathol.72(12):830-835; Maloney et al. (2018) TT054, J. Mol. Diagn. 20(6), 895–1039; Maertens et al.(2017) Annals of Oncology 28 (suppl_5): v22-v42; De Craene et al. (2018) Journal of ClinicalOncology 36:15_suppl,e15639; De Craene B. et al. (2018) Annals of Oncology 29 (suppl_8):viii14-viii57; De Craene et al. (2017) Annals of Oncology 28 (suppl_5): v209-v268; Nicka et. al.(2018) J. Mol. Diagn. 20(6), 895–1039; Nafa. et.al. (2018) J. Mol. Diagn. 20(6), 895–1039), isused in a large number of laboratories (>40 in the US) and has been validated and approvedfor clinical use at MSKCC by the New York State Department of Health Clinical LaboratoryEvaluation Program (project ID: 68841; https://www.wadsworth.org/print/82410).

3/13/2020 5:08 AM

8 Thank you for initiating this effect. There is a need to standardize IHC and PCR testingapproaches. There are labs still using dinucleotide biomarkers. There is no interpretationstandard of MSI. Likewise, the IHC loss seems to be lacking a consensus definition.

3/12/2020 5:11 PM

9 In review of weekly summaries, I see that comments are overwhelmingly from academia andindustry. Most actual patients are in fact managed in the community. I would like to present acommunity perspective from our substantial experience with biomarker testing at theMinneapolis VA hospital. Limitations of the capillary based MSI tests pertain not to theperformance of those assays under ideal input conditions, but rather the lack of those idealconditions in field practice. For capillary based MSI, high quantity and quality of DNA, highallelic ratios required, furthermore somatic comparisons are needed just where it matters in thecomplementarity with IHC. In our practice we give cancer diagnoses from little ditzels of few

3/10/2020 3:41 PM

Not Vali

d Afte

r Marc

h 13,

2020


Recommendations

54 / 54

hundred cells, and then run PCR from corresponding sub-nanogram effective inputs. In ournon-academic practice experience that involves surgical pathology and MSI testing, IHCdemonstrated itself as the practical mainstay of testing. DNA based is a supplement to that, buta necessary one as some of the HNPCC mutations, as you know express protein. If you aregoing to offer capillary based only, without IHC, the odds are that you are going to be doing alesser service because of delayed, QNS, and false negative results that I would anticipate.Comments above equally apply to NGS fans. NGS, for the same (and other) reasons, remainssimply inferior to conventional testing in clinical oncology practice, whether it is MSI or anyother biomarker, but unless you are practicing in your tertiary cancer center where everyonegoes on to investigational therapy and every tumor is a treasure chest of datapoints. Mostcancer patients here go through their standard lines of therapy and then to hospice. We test fortheir standard line targeted therapies in 2 days, with rare-to-absent QNS results, with littlemoney spent. I have to but agree with the third party payers not wanting to reimburseexpensive tests in the absence of parameters separating effective utilization from bombasticsdriven ordering that is off touch with reality. Lastly, HRM could be the way forward on MSItesting. The commercial cartridge based MSI assay appears HRM based. There is a few otherrecent blips in literature on MSI by HRM. We have ourselves just launched in here our own firstversion of a lab developed HRM assay to substitute the capillary based assay. Performance isvery promising, and HRM without somatic comparison performed much better than capillaryanalysis without somatic comparison. This seemed mostly because HRM works well withlimited tissue. We are accepting 0.1ng effective input per rx in this first version. Deniz AslanStaff Pathologist Minneapolis VAHS 612 467 2508

10 a 3/10/2020 11:56 AM

11 We need try to perform the tests using MMR IHC and MSI-PCR as the first option rather thanperforming expensive NGS for TMB or MSI.

3/9/2020 9:36 PM

12 1: The recommendations are fairly reasonable and can be implemented in practice 2: Clarity isrequired in defining "indeterminate" result is and role of tumor heterogeneity particularly inmetastatic setting if MMR by IHC is negative in metasttaic sitre ..should we be testing primarytumor 3: How many years back can a pathologist go back to test the primary tumor tissue ( forexample <5 yrs or <10yrs)

3/9/2020 1:07 PM

13 Review the format of Notes shown with the Recommendations (Note 1 with recommendation#2, note 2 with recommendation #4, and note 3 with recommendation #5) in the table posted forpublic comment athttps://documents.cap.org/documents/mmrmsi_draft_recommendations_ocp.pdf

3/9/2020 12:29 PM

14 These guidelines are very useful and timely, and will impact my practice. 3/6/2020 8:58 AM

15 I suspect that centers that already rely on a well-validated NGS assay for MSI detection will bereluctant to adopt the recommendations as written, as they feel like a step backwards,particularly for non-CRC indications. NGS should be generally given equal footing as IHC andPCR, with an emphasis that orthogonal testing is warranted in unusual or difficult to interpretcontexts.

2/22/2020 8:09 AM

16 I appreciate all of the time invested by the members in developing these guidelines and believethey will improve patient care. There is confusion currently regarding MMR IHC versus MSItesting for immune checkpoint inhibitors.

2/21/2020 7:38 AM

17 Not sure how the panel members are chosen. The recommendation is mostly for GI tumors, Idon't see many GI pathologists in the panel.

2/20/2020 6:52 PM

18 thank you for drafting this guidance 2/20/2020 10:54 AM

19 Its going to difficult to give IHC and PCR a primary role when the clinicians are clamoring forNGS and accept the results fairly uncritically.

2/20/2020 8:07 AM

20 Please address the use of cell blocks for MMR testing. 2/20/2020 8:03 AM

21 Nicely done! Very clear. 2/19/2020 11:26 PM

22 Nice work ! 2/19/2020 3:21 PM

23 Thanks for doing this 2/19/2020 2:50 PM


Not Vali

d Afte

r Marc

h 13,

2020

13, 2020 march after · 25 phd laboratory director 2/20/2020 8:03 am 26 geneticist 2/20/2020 2:37...

Documents