Molecular Biomarkers for the Evaluation of Colorectal Cancer · words for the concepts “colorectal cancer,”“biomarkers,” “treatment,” and “treatment outcomes” and

The Journal of Molecular Diagnostics Vol 19 No 2 March 2017

jmdamjpatholorg

SPECIAL ARTICLE

Molecular Biomarkers for the Evaluation ofColorectal Cancer

Guideline From the American Society for Clinical PathologyCollege of American Pathologists Association for MolecularPathology and American Society of Clinical OncologyAntonia R Sepulveda MD PhD1 Stanley R Hamilton MD PhD2 Carmen J Allegra MD5 Wayne Grody MD PhD6

Allison M Cushman-Vokoun MD PhD7 William K Funkhouser MD PhD8 Scott E Kopetz MD PhD3 Christopher Lieu MD9

Noralane M Lindor MD10 Bruce D Minsky MD4 Federico A Monzon MD11 Daniel J Sargent PhD12 Veena M Singh MD13

Joseph Willis MD14 Jennifer Clark SCT MB (ASCP)CM15 Carol Colasacco MLIS16 R Bryan Rumble MSc17

Robyn Temple-Smolkin PhD18 Christina B Ventura MT(ASCP)16 and Jan A Nowak MD PhD19

From the 1Department of Pathology and Cell Biology Columbia University New York NY Departments of 2Pathology 3Gastrointestinal (GI) MedicalOncology and 4Radiation Oncology University of Texas MD Anderson Cancer Center Houston 5Division of Hematology and Oncology University ofFlorida Medical Center Gainesville 6Departments of Pathology and Laboratory Medicine Pediatrics and Human Genetics UCLA Medical Center LosAngeles CA 7Department of Pathology and Microbiology University of Nebraska Medical Center Omaha 8Department of Pathology and LaboratoryMedicine University of North Carolina School of Medicine Chapel Hill 9Division of Medical Oncology University of Colorado Denver School of MedicineDenver 10Department of Medical Genetics Mayo Clinic Scottsdale AZ 11Castle Biosciences Friendswood TX 12Department of Health Sciences ResearchMayo Clinic Rochester MN 13Biocept San Diego CA 14Department of Pathology Case Western Reserve University Cleveland OH 15ASCP Institute forScience Technology and Policy American Society for Clinical Pathology Washington DC 16Laboratory and Pathology Quality Center College ofAmerican Pathologists Northfield IL 17American Society of Clinical Oncology Alexandria VA 18Association for Molecular Pathology Bethesda MD and19Department of Pathology and Laboratory Medicine Roswell Park Cancer Institute Buffalo NY

Accepted for publication

November 7 2016

Corresponding author AntoniaSepulveda MD PhD Depart-ment of Pathology and CellBiology 630 W 168th St VC14-212 Columbia UniversityNew York NY 10032 E-mailas4400cumccolumbiaedu

opyright ordf 2017 American Society for Clini

merican Society for Investigative Pathology

ttpdxdoiorg101016jjmoldx201611001

Objectives To develop evidence-based guideline recommendations through a systematic review of theliterature to establish standard molecular biomarker testing of colorectal cancer (CRC) tissues to guideepidermal growth factor receptor (EGFR) therapies and conventional chemotherapy regimensMethods The American Society for Clinical Pathology College of American Pathologists Association forMolecular Pathology and American Society of Clinical Oncology convened an expert panel to develop anevidence-based guideline to establish standard molecular biomarker testing and guide therapies forpatients with CRC A comprehensive literature search that included more than 4000 articles wasconductedResults Twenty-one guideline statements were establishedConclusions Evidence supports mutational testing for EGFR signaling pathway genes sincethey provide clinically actionable information as negative predictors of benefit to anti-EGFRmonoclonal antibody therapies for targeted therapy of CRC Mutations in several of the bio-markers have clear prognostic value Laboratory approaches to operationalize CRC molecular testingare presentedKey Words Molecular diagnostics Gastrointestinal Histology Genetics Oncology (J Mol Diagn 201719 187e225 httpdxdoiorg101016jjmoldx201611001)

cal Pathology College of American Pathologists Association for Molecular Pathology American Society for Clinical Oncology and

Sepulveda et al

Molecular testing to select targeted and conventional ther- and sensitivity meet the clinical needs While earlier testing

apies for patients with colorectal cancer (CRC) has been thefocus of a number of recent studies and is becoming stan-dard practice for management of patients with CRC Mo-lecular markers that predict response to a specific therapy ortreatment regimen are known as predictive biomarkers1

Monoclonal antibody therapies that target the epidermalgrowth factor receptor (EGFR) bind the EGFR extracellulardomain blocking EGFR signaling pathways Anti-EGFRmonoclonal antibodies have been the main targeted thera-pies for CRC that require knowledge of the mutationalstatus of genes in the pathway as predictive biomarkers ofresponse to these therapies2-4 Initial clinical trial datademonstrated that patients with CRC carrying activatingmutations of KRAS affecting exon 2 codons 12 and 13did not benefit from anti-EGFR monoclonal antibodytherapy2-4 Subsequent studies described other mutations ingenes of the EGFR signaling pathways involving otherexons of KRAS and in NRAS BRAF PIK3CA and PTENthat may affect response of CRC to anti-EGFR antibodytherapies Guidelines addressing the molecular testing ofEGFR pathway genes beyond KRAS have not been estab-lished and are needed in clinical practice

The DNA mismatch repair (MMR) status of CRC mayhave predictive value in some clinical settings While testingof CRC for MMR has been recommended for all patientswith CRC as a workup test to evaluate for possible Lynchsyndrome5 guidelines for the use of MMR as a predictivebiomarker of response to therapy have not been reportedRecent molecular biomarker data have shown the impor-tance of microsatellite instability (MSI) testing a marker ofdeficient mismatch repair (dMMR) for the selection ofpatients for immunotherapy (see section on emerging bio-markers below)

Alterations of a number of critical genes in CRC devel-opment and progression such as dMMR and BRAF acti-vating mutations have been shown to affect prognosis asmeasured by several metrics of tumor progression or sur-vival6-8 The utility of incorporating prognostic biomarkersin the management of patients with CRC has not been welldefined in clinical practice Defining the utility of informa-tion gathered from prognostic molecular biomarkers forclinical management of patients with CRC is warranted

The postgenome era and the emphasis on precisiongenomic-based medicine are providing enormous amountsof new data and many promising new molecular cancerbiomarkers that may emerge as molecular diagnostic toolsthat can be used to enhance successful treatment of patientswith CRC and other cancers Laboratories and regulatoryagencies are faced with challenges to rapidly and efficientlyprovide new test results for the management of patients withcancer Laboratory testing of molecular biomarkers involvesthe selection of assays type of specimens to be testedtiming of ordering of tests and turnaround time for testingresults Recent years have shown that a plethora of technicalapproaches can effectively be used as long as test specificity

approaches were focused on one or a few testing targets thecurrent need for multiple molecular markers from poten-tially minute tumor samples is leading to greater use of genepanels such as targeted next-generation sequencing (NGS)cancer panels which can assay from a few to hundreds ofgenes and amplicons with known mutational hotspots incancerThere is a need for current evidence-based recommen-

dations for the molecular testing of CRC tissues to guideEGFR-targeted therapies and conventional chemotherapyregimens Therefore the current recommendations weredeveloped through collaboration of four societies AmericanSociety for Clinical Pathology (ASCP) College of Amer-ican Pathologists (CAP) Association for Molecular Pa-thology (AMP) and American Society of Clinical Oncology(ASCO) This guideline follows well-established methodsused in their development as well as for regular updatessuch that new advances in the molecular testing for clinicalmanagement of CRC can be integrated in future updates ofthe guideline in a timely manner

Panel Composition

The ASCP the CAP Pathology and Laboratory QualityCenter (the Center) the AMP and the ASCO convened anexpert panel consisting of practicing pathologists oncolo-gists geneticists and a biostatistician with expertise andexperience in molecular biomarker testing and targetedtherapies for CRC The ASCP CAP AMP and ASCOjointly approved the appointment of the project cochairsand expert panel members In addition a methodologistexperienced in systematic review and guideline develop-ment consulted with the panel throughout the project

Conflict of Interest Policy

Prior to acceptance on the expert or advisory panel potentialmembers completed a joint guideline conflict of interest(COI) disclosure process whose policy and form (in effectJuly 2011) require disclosure of material financial interestin or potential for benefit of significant value from theguidelinersquos development or its recommendations 12 monthsprior through the time of publication The potential mem-bers completed the COI disclosure form listing any rela-tionship that could be interpreted as constituting an actualpotential or apparent conflict All project participants wererequired to disclose conflicts prior to beginning andcontinuously throughout the projectrsquos timeline Disclosedconflicts of the expert panel members are listed in Appendix 1and Appendix 2The ASCP CAP AMP and ASCO provided funding for

the administration of the project no industry funds wereused in the development of the guideline All panel mem-bers volunteered their time and were not compensated for

jmdamjpatholorg - The Journal of Molecular Diagnostics

ASCPCAPAMPASCO CRC Biomarker Guideline

their involvement except for the contracted methodologistPlease see the Supplemental Digital Content (SDC) at httpdxdoiorg101016jjmoldx201611001 for full details onthe COI policy

Objective

The scope of the project was to develop an evidence-basedguideline to help establish standard molecular biomarkertesting guide targeted therapies and advance personalizedcare for patients with CRC The panel addressed thefollowing key questions

1 What biomarkers are useful to select patients with CRCfor targeted and conventional therapies

2 How should tissue specimens be processed for biomarkertesting for CRC management

3 How should biomarker testing for CRC management beperformed

4 How should molecular testing of CRC be implementedand operationalized

5 Are there emerging genesbiomarkers that should beroutinely tested in CRC

Materials and Methods

This evidence-based guideline was developed followingstandards as endorsed by the Institute of Medicine9 Adetailed description of the methods and systematic review(including the quality assessment and complete analysis ofthe evidence) can be found in the SDC

Literature Search and Selection

A comprehensive search for literature was performed inMEDLINE using the OvidSP (August 1 2013) and PubMed(September 17 2013) interfaces The initial MEDLINEsearch encompassed the publication dates of January 1 2008through August 1 2013 (OvidSP) and January 1 2008through September 17 2013 (PubMed) A supplementalliterature search was performed using Scopus (September 252013) to identify relevant articles published between January1 2008 and September 25 2013 in journals not indexed inMEDLINE The literature search of the electronic databasesinvolved two separate searches in each database the firstusing Medical Subject Headings (MeSH) terms and key-words for the concepts ldquocolorectal cancerrdquo ldquobiomarkersrdquoldquotreatmentrdquo and ldquotreatment outcomesrdquo and the second usingterms for the concepts ldquocolorectal cancerrdquo ldquobiomarkersrdquo andldquolaboratory methodsrdquo Limits were set for human studiespublished in English and a publication filter was applied toexclude lower levels of evidence such as letters commen-taries editorials and case reports The Ovid search was rerunon February 12 2015 to identify articles published sinceAugust 1 2013

The Journal of Molecular Diagnostics - jmdamjpatholorg

In addition to the searches of electronic databases anInternet search of international health organizations theNational Guidelines Clearinghouse and Guidelines Interna-tional Network was conducted for existing relevant guide-lines or protocols Guidelines were included if they werepublished since 2008 in English The proceedings of themeetings of the ASCO and ASCO-Gastrointestinal CancersSymposium European Society for Medical Oncology andthe American Association for Cancer Research from 2012and 2013 were also searched for relevant abstracts

A focused examination of all systematic reviews retrievedby the initial literature search and retained after full-textreview was performed to identify primary research studiesnot already included In addition recommendations fromthe expert panel were reviewed and the reference lists of allarticles deemed eligible for inclusion were scanned forrelevant reports The results of all searches were combinedand deduplicated

Detailed information regarding the literature searchstrategy can be found in the SDC

Eligible Study Designs

Practice guidelines consensus documents systematic re-views meta-analyses randomized controlled trials compar-ative studies reviews and evaluation studies were eligible forinclusion In addition to journal articles the search identifiedmeeting abstracts

Inclusion Criteria

Published studies were selected for full-text review if theymet each of the following criteria

1 Patients with colorectal or rectal cancer with a pathologydiagnosis of adenocarcinoma or adenocarcinoma withneuroendocrine differentiation either primary or metastatic

2 Patients of all ages3 Patients with cancer of any invasive stage (T1-T4)4 Biomarker testing such asKRAS (Kirsten rat sarcoma viral

oncogene homolog) DNA MMRMSI BRAF (V-raf mu-rine sarcoma viral oncogene homolog B1) NRAS [neuro-blastoma RAS viral (v-ras) oncogene homolog] PIK3CA(phosphatidylinositol-45-bisphosphate 3-kinase catalyticsubunit alpha) PTEN (phosphatase and tensin homolog)MLH1 (MutL homolog 1) methylation or gene expressionprofiles

5 Comparative studies6 Human studies7 Studies published in English

Exclusion Criteria

1 All other tumor primaries and types (ie noncolorectal ornonrectal cancers tumor types other than adenocarci-noma or adenocarcinoma with neuroendocrinedifferentiation)

Sepulveda et al

2 Patients with noninvasive tumors (ie intraepithelialdysplasia in situ polyps without carcinoma)

3 Studies of colorectal cancers without biomarker testingnovel biomarkersdfor example VEG-F (vascularendothelial growth factor) XRCC1 (X-ray repair com-plementing defective repair in Chinese hamster cells 1)IGF (insulin-like growth factor) ERCC (excision repaircross-complementing rodent repair deficiency comple-mentation group 1) micro-RNA TYMS (thymidylatesynthetase) GCC (guanylyl cyclase C) LINE (longinterspersed nucleotide element) methylation CIMP(CpG island methylator phenotype) HER2 (V-erb-b2erythroblastic leukemia viral oncogene homolog 2) CIN(chromosomal instability) status LOH (loss of heterozy-gosity) and germline (genetics only) testing

4 Non-English-language articles5 Animal studies6 Studies published prior to 20027 Noncomparative studies letters commentaries or

editorials8 Studies that did not address at least one of the defined

inclusion criteria9 Studies with fewer than 50 patients per comparison arm

Outcomes of Interest

The primary outcomes of interest included survival outcomesand performance characteristics of laboratory testing assaysSurvival outcomes included overall survival (OS) disease-free survival (DFS) progression-free survival (PFS)recurrence-free survival time to recurrence response totherapy (eg complete and partial response) Laboratory dataand test performance characteristics included percent muta-tion concordance of testing methods sensitivity of testingmethods specificity of testing methods concordance ofdetected mutations between primary and metastatic muta-tions [number () of cases with mutations vs number of caseswith no mutations in the gene of interest] and concordance ofmutations (synchronous primary vs metastatic metachro-nous primary vs metastatic between synchronous metasta-ses between metachronous metastases)

Quality Assessment

An assessment of the quality of the evidence was performedfor all retained studies following application of the inclusion

Table 1 Levels of Evidence

Level Description

Level I Evidence derived from systematic reviewsLevel II Evidence derived from randomized controLevel III Evidence derived from comparative studieLevel IV Evidence without a comparator (eg case

Data derived from National Health and Medical Research Council10

and exclusion criteria by the methodologist Using thismethod studies deemed to be of low quality would not beexcluded from the systematic review but would be retainedand their methodologic strengths and weaknesses discussedwhere relevant Studies would be assessed by confirmingthe presence of items related to both internal and externalvalidity which are all associated with methodologic rigorand a decrease in the risk of bias The quality assessment ofthe studies was performed by determining the risk of bias byassessing key indicators based on study design againstknown criteria (Refer to the SDC for detailed discussion ofthe quality assessment)For strength of the evidence the panel considered the level

of evidence as well as its quantity and quality of includedstudies The level of evidence was based on the study designas described in Table 110 In general level I and II evidence isconsidered most appropriate to answer clinical questions butin the absence of such high-quality evidence the panelconsidered data from lower quality studies The quantity ofevidence refers to the number of studies and number of casesincluded for each outcome in the recommendation Thequality of studies reflects how well the studies were designedto eliminate bias and threats to validityThe appropriateness of the study design and data

collected relevance and clarity of findings and adequacy ofconclusions were evaluated Each study was assessed indi-vidually (refer to the SDC for individual assessments andresults) and then summarized by study type Componentssuch as generalizability and applicability were also consid-ered when determining the strength of evidence A summaryof the overall quality of the evidence was given consideringthe evidence in totality Ultimately the designation (ierating or grade) of the strength of evidence is a judgment bythe expert panel of its level of confidence that the evidencefrom the studies informing the recommendations reflectstrue effect Table 2 describes the grades for strength ofevidence11

Assessing the Strength of Recommendations

Development of recommendations requires that the panelreview the identified evidence and make a series of keyjudgments (using procedures described in the SDC) Gradesfor strength of recommendations were developed by theCAP Pathology and Laboratory Quality Center and aredescribed in Table 311

of appropriate level II studies andor clinical practice guidelineslled trialss (eg prospective cohort studies retrospective cohort studies)reports case series narrative reviews)

Table 2 Grades for Strength of Evidence

Designation Description Quality of Evidence

Convincing High confidence that available evidence reflects true effectFurther research is very unlikely to change the confidencein the estimate of effect

Highintermediate quality of evidence

Adequate Moderate confidence that available evidence reflects trueeffect Further research is likely to have an importantimpact on the confidence in estimate of effect and maychange the estimate

Intermediatelow quality of evidence

Inadequate Little confidence that available evidence reflects trueeffect Further research is very likely to have animportant impact on the confidence in the estimate ofeffect and is likely to change the estimate

Lowinsufficient quality of evidence and expert panel usesformal consensus process to reach recommendation

Insufficient Evidence is insufficient to discern net effect Any estimateof effect is very uncertain

Insufficient evidence and expert panel uses formalconsensus process to reach recommendation

Adapted from Guyatt et al11 by permission of BMJ Publishing Group Limited

Guideline Revision

This guideline will be reviewed every 4 years or earlier inthe event of publication of substantive and high-qualityevidence that could potentially alter the original guidelinerecommendations If necessary the entire panel willreconvene to discuss potential changes When appropriatethe panel will recommend revision of the guideline to theASCP CAP AMP and ASCO for review and approval

Disclaimer

Practice guidelines and consensus statements reflect the bestavailable evidence and expert consensus supported inpractice They are intended to assist physicians and patientsin clinical decision making and to identify questions and

Table 3 Grades for Strength of Recommendation

Designation Recommendation

Strong recommendation Recommend for or against a particular motesting practice for colorectal cancer (camust or should )

Recommendation Recommend for or against a particular motesting practice for colorectal cancer (cashould or may)

Expert consensus opinion Recommend for or against a particular motesting practice for colorectal cancer (cashould or may)

No recommendation No recommendation for or against a particmolecular testing practice for colorectal

Data derived from Guyatt et al11

settings for further research With the rapid flow of scientificinformation new evidence may emerge between the time apractice guideline or consensus statement is developed andwhen it is published or read Guidelines and statements arenot continually updated and may not reflect the most recentevidence Guidelines and statements address only the topicsspecifically identified therein and are not applicable to otherinterventions diseases or stages of diseases Furthermoreguidelines and consensus statements cannot account forindividual variation among patients and cannot be consid-ered inclusive of all proper methods of care or exclusive ofother treatments It is the responsibility of the treatingphysician or other health care provider relying on inde-pendent experience and knowledge to determine the bestcourse of treatment for the patient Accordingly adherenceto any practice guideline or consensus statement is volun-tary with the ultimate determination regarding its

Rationale

lecularn include

Supported by convincing or adequate strength ofevidence high or intermediate quality ofevidence and clear benefit that outweighs anyharms

lecularn include

Some limitations in strength of evidence (adequateor inadequate) and quality of evidence(intermediate or low) balance of benefits andharms values or costs but panel concludes thatthere is sufficient evidence andor benefit toinform a recommendation

lecularn include

Serious limitations in strength of evidence(inadequate of insufficient) quality of evidence(intermediate or low) balance of benefits andharms values or costs but panel consensus isthat a statement is necessary

ularcancer

Insufficient evidence or agreement of the balance ofbenefits and harms values or costs to provide arecommendation

Table 4 Guideline Statements and Strength of Recommendations

Guideline Statement Strength of Recommendation

1 Patients with colorectal carcinoma being considered for anti-EGFR therapy must receive RAS mutationaltesting Mutational analysis should include KRAS and NRAS codons 12 and 13 of exon 2 59 and 61 ofexon 3 and 117 and 146 of exon 4 (ldquoexpandedrdquo or ldquoextendedrdquo RAS )

Recommendation

2a BRAF pV600 [BRAF c1799 (pV600)] mutational analysis should be performed in colorectal cancertissue in patients with colorectal carcinoma for prognostic stratification

Recommendation

2b BRAF pV600 mutational analysis should be performed in deficient MMR tumors with loss of MLH1 toevaluate for Lynch syndrome risk Presence of a BRAF mutation strongly favors a sporadic pathogenesisThe absence of a BRAF mutation does not exclude risk of Lynch syndrome

Recommendation

3 Clinicians should order mismatch repair status testing in patients with colorectal cancers for theidentification of patients at high risk for Lynch syndrome andor prognostic stratification

Recommendation

4 There is insufficient evidence to recommend BRAF c1799 pV600 mutational status as a predictivemolecular biomarker for response to anti-EGFR inhibitors

No recommendation

5 There is insufficient evidence to recommend PIK3CA mutational analysis of colorectal carcinoma tissuefor therapy selection outside of a clinical trialNote Retrospective studies have suggested improved survival with postoperative aspirin use inpatients whose colorectal carcinoma harbors a PIK3CA mutation

No recommendation

6 There is insufficient evidence to recommend PTEN analysis (expression by immunohistochemistry ordeletion by fluorescence in situ hybridization) in colorectal carcinoma tissue for patients who are beingconsidered for therapy selection outside of a clinical trial

No recommendation

7 Metastatic or recurrent colorectal carcinoma tissues are the preferred specimens for treatment predictivebiomarker testing and should be used if such specimens are available and adequate In their absenceprimary tumor tissue is an acceptable alternative and should be used

Expert consensus opinion

8 Formalin-fixed paraffin-embedded tissue is an acceptable specimen for molecular biomarker mutationaltesting in colorectal carcinoma Use of other specimens (eg cytology specimens) will require additionaladequate validation as would any changes in tissue-processing protocols

9 Laboratories must use validated colorectal carcinoma molecular biomarker testing methods withsufficient performance characteristics for the intended clinical use Colorectal carcinoma molecularbiomarker testing validation should follow accepted standards for clinical molecular diagnostics tests

Strong recommendation

10 Performance of molecular biomarker testing for colorectal carcinoma must be validated in accordancewith best laboratory practices

11 Laboratories must validate the performance of IHC testing for colorectal carcinoma molecularbiomarkers (currently IHC testing for MLH1 MSH2 MSH6 and PMS2) in accordance with best laboratorypractices

12 Laboratories must provide clinically appropriate turnaround times and optimal utilization of tissuespecimens by using appropriate techniques (eg multiplexed assays) for clinically relevant molecular andimmunohistochemical biomarkers of colorectal cancer

13 Molecular and IHC biomarker testing in colorectal carcinoma should be initiated in a timely fashionbased on the clinical scenario and in accordance with institutionally accepted practicesNote Test ordering can occur on a case-by-case basis or by policies established by the medical staff

14 Laboratories should establish policies to ensure efficient allocation and utilization of tissue formolecular testing particularly in small specimens

15 Members of the patientrsquos medical team including pathologists may initiate colorectal carcinomamolecular biomarker test orders in accordance with institutionally accepted practices

16 Laboratories that require send-out of tests for treatment predictive biomarkers should process and sendcolorectal carcinoma specimens to reference molecular laboratories in a timely mannerNote It is suggested that a benchmark of 90 of specimens should be sent out within 3 working days

17 Pathologists must evaluate candidate specimens for biomarker testing to ensure specimen adequacytaking into account tissue quality quantity and malignant tumor cell fraction Specimen adequacyfindings should be documented in the patient report

18 Laboratories should use colorectal carcinoma molecular biomarker testing methods that are able todetect mutations in specimens with at least 5 mutant allele frequency taking into account theanalytical sensitivity of the assay (limit of detection or LOD) and tumor enrichment (egmicrodissection)Note It is recommended that the operational minimal neoplastic carcinoma cell content tested shouldbe set at least two times the assayrsquos LOD

(table continues)

Sepulveda et al

192 jmdamjpatholorg - The Journal of Molecular Diagnostics

Table 4 (continued )

19 Colorectal carcinoma molecular biomarker results should be made available as promptly as feasible toinform therapeutic decision making both prognostic and predictiveNote It is suggested that a benchmark of 90 of reports be available within 10 working days fromdate of receipt in the molecular diagnostics laboratory

20 Colorectal carcinoma molecular biomarker testing reports should include a results and interpretationsection readily understandable by oncologists and pathologists Appropriate Human Genome VariationSociety and Human Genome Organisation nomenclature must be used in conjunction with any historicalgenetic designations

21 Laboratories must incorporate colorectal carcinoma molecular biomarker testing methods into theiroverall laboratory quality improvement program establishing appropriate quality improvement monitorsas needed to ensure consistent performance in all steps of the testing and reporting process Inparticular laboratories performing colorectal carcinoma molecular biomarker testing must participate informal proficiency testing programs if available or an alternative proficiency assurance activity

EGFR epidermal growth factor receptor IHC immunohistochemistry MMR mismatch repair PTEN phosphatase and tensin homolog

application to be made by the physician in light of eachpatientrsquos individual circumstances and preferences TheASCP CAP AMP and ASCO make no warranty expressor implied regarding guidelines and statements and spe-cifically exclude any warranties of merchantability andfitness for a particular use or purpose The ASCP CAPAMP and ASCO assume no responsibility for any injury ordamage to persons or property arising out of or related toany use of this statement or for any errors or omissions

Results

A total of 4197 studies met the search term requirements Atotal of 123 articles were included for data extractionExcluded articles were available as discussion or back-ground references The panel convened 14 times (11 tele-conference webinars and three face-to-face meetings) fromJuly 27 2013 through September 24 2015 to develop thescope draft recommendations review and respond to soli-cited feedback and assess the quality of evidence thatsupports the final recommendations Additional work wascompleted via electronic mail An open comment periodwas held from March 30 2015 through April 22 2015during which draft recommendations were posted on theAMP website Twenty-one guideline statements had anagreement ranging from 60 to 94 for each statementfrom the open-comment period participants (refer to Out-comes in the SDC for full details) The website received atotal of 248 comments Teams of three to four expert panelmembers were assigned three to five draft recommendationsto review all comments received and provide an overallsummary to the rest of the panel Following panel discus-sion and the final quality of evidence assessment the panelmembers determined whether to maintain the original draftrecommendation as is revise it with minor language changeor consider it as a major recommendation change Theexpert panel modified eight draft statements based on the

feedback during the open-comment period and the consid-ered judgment process Resolution of all changes was ob-tained by majority consensus of the panel using nominalgroup technique (rounds of email discussion and multipleedited recommendations) among the panel members Thefinal recommendations were approved by the expert panelwith a formal vote The panel considered the risks andbenefits throughout the whole process in their consideredjudgment process Formal cost analysis or cost-effectivenesswas not performed

Each organization instituted a review process to approvethe guideline The ASCP assigned the review of theguideline to a Special Review Panel For the CAP an in-dependent review panel (IRP) representing the Council onScientific Affairs was assembled to review and approve theguideline The IRP was masked to the expert panel andvetted through the COI process The AMP approval processrequired the internal review of an independent panel led bythe Publications and Communications Committee Chair andExecutive Committee approval The ASCO approval pro-cess required the review and approval of the ClinicalPractice Guidelines Committee

Guideline Statements

1 Recommendation Patients with CRC being consideredfor anti-EGFR therapy must receive RAS mutational testingMutational analysis should include KRAS and NRAS codons12 and 13 of exon 2 59 and 61 of exon 3 and 117 and 146of exon 4 (ldquoexpandedrdquo or ldquoextendedrdquo RAS ) (Table 4)

Aberrant activation of EGFR signaling pathways in CRCis primarily associated with activating mutations of genes inthe mitogen-activated protein kinase and phosphatidylino-sitol-3-kinase (PI3K) pathways Together KRAS NRASand BRAF mutations have been reported to occur in morethan half of all CRC cases and KRAS or NRAS and BRAFmutations are inversely associated with a small proportion

Table 5 KRAS Clinical Practice Guidelines Systematic Reviews Meta-Analyses Prospective Cohort Studies and Retrospective CohortStudies

Author YearNo of Studies(No of Patients) Comparison Tests Used Codons Studied OS PFS ORR

CPGs systematic reviews and meta-analyses on KRAS mutationthorn vs mutatione (n Z 30)Petrelli et al35

2013SR 12 studies

including

2226 patients

with mCRC treated

with bevacizumab

Mutthorn vs Mute NR NR Median HR 065

95 CI 046-

092 P lt 05in favor of Mute

Median PFS HR

085 95 CI

074-098 P lt

05 in favor ofMute

KRAS Mutthorn483 vs KRASMute 548(OR 142 95CI 105-192P lt 05)

Mao et al32

2013SR 10 studies

including 1487

patients with mCRC

treated with

cetuximab

pG13D vs codon 12

MutthornNR G13D 12 HR 052 95 CI

033-080

P lt 05 in favorof G13D

PFS HR 054 95

CI 036-081

KRAS pG13D 22KRAS 12 16KRAS Mute 44(pG13D vs 12 RR

164 95 CI113-238P lt 05)

pG13D vs MuteRR 054 CI038-077P lt 05)

Jiang et al27

2013SR 13 studies

including 1174

patients with mCRC

treated with

cetuximab or

panitumumab

Increased vs not

increased EGRF

FISH CISH SISH

NR Increased GCN

associated with

improved OS

among patients

treated with

anti-EGFR mAbs

(HR 062 95

CI 050-077

P lt 05)

GCN associated

with improved

PFS (HR 065

95 CI 047-

089 P lt 05)

Hoyle et al25

2013SR-HTA 2 studies

including EGFR-

expressing mCRC

patients total with

cetuximab

bevacizumab or

panitumumab in

the second-line

and greater

Mutthorn vs Mute NR NR Median 95

months vs 48

months HR

055 95 CI

041-075 P lt

05 in favor ofcetuximab overBSC in Mute

Panitumumab thornBSC comparedwith BSC alonein MuteP Z ns

Median PFS HR

040 95 CI

030-054 P lt

05 third-linecetuximab thornBSC comparedwith BSC alonein Mute

Median PFS HR045 95 CI034-059P lt 05panitumumab thornBSC comparedwith BSC alonein Mute

KRAS Mute128

KRAS Mutthorn12 P lt 05

cetuximab thornBSC comparedwith BSC alonein Mute

KRAS Mute 10KRAS mut 0

P lt 05panitumumabthorn BSCcompared withBSC alone inMute

Chen et al21

2013SR 7 studies

including 2802

patients with mCRC

Codon 13 Mutthorn vs

other mutations

PCR direct

sequencing

13 other Mutthorn

Median OS

146 monthscodon 13

118 months(othermutation)

173 monthsMute

Median PFS

64 months codon13

41 months (othermutation)

66 months Mute

Codon 13 Mutthornvs other

mutations RR152 (95 CI110-209P lt 05)

Codon 13 Mutthornvs Mute RR061 (95 CI045-083P lt 05)

Zhou et al16

2012SR 4 RCTs including

1270 first-line

patients with mCRC

(all Mute)

Oxaliplatin CT anti-EGFR mAbs

Anti-EGFR thorn CT vs CT

Mute only HR 100 95 CI

088-113

P Z ns

HR 086 95 CI

071-104

P Z ns

RR 108 95 CI

086-136

P Z ns

Zhang et al41

2011SR 4 studies

including 2912

patients with mCRC

Mutthorn vs Mute NR NR Cetuximab thorn CT vs

CT alone Mute

HR 084 95 CI064-111P Z ns

Cetuximab thorn CT vs

CT alone Mute

HR 064 95

CI 050-084 Plt 05 favorsthorncetuximab

Cetuximab thorn CT

vs CT alone

RR 193 95

CI 114-326

P lt 05 favorsthorncetuximab

(table continues)

Sepulveda et al

CT alonemutation HR103 95 CI074-144P Z ns

Cetuximab thorn CT vsCT alonemutation HR137 95 CI081-231P Z ns

CetuximabthornCT vsCT alone Mute RR 14495 CI 120-173 P lt 05favorsthorncetuximab

Yang et al40

2012SR 19 studies

including 1077

patients with mCRC

Mutthorn vs Mute

GCNthorn vs GCNeFISH qPCRCISH Exon 20 No pooling due to

statistical

heterogeneity

No pooling due to

statistical

heterogeneity

No pooling due to

statistical

heterogeneity

Vale et al39

5996 patients

with advanced CRC

Mutthorn vs Mute NR NR NR

Third line HR076 95 CI062-092P lt 05

Firstsecond line

PFS HR 083

95 CI 076-

090 P lt 05Third line PFS HR

043 95 CI035-052 P lt

05 in favor ofanti-EGFR MAbsfor Mute only

Tsoukalaset al38

SR 13 studies

including 1394

patients with CRC

Mutthorn vs Mute

Response tocetuximab vs noresponse

NR NR NR NR NR

Ross et al42

2012SR Six studies

including 2526

patients with mCRC

Mutthorn vs Mute

Antibody vs controlSanger

pyrosequencing

PCR ARMS

Scorpion

NR NR NR NR

Ren et al37

2012SR 23 studies

including 1362

patients with

mutations

(w100 at

codons 12 and 13

n Z 1 at codon

Mutthorn vs Mute d 12 13 61 HR 161 95 CI

119-218

P lt 05 in favorof treatment inMute vs Mutthornpatients

Petrelli et al34

484 Mute patients

with mCRC

Mutthorn vs Mute

Cetuximab andorpanitumumab thorn CTvs CT alone

NR NR P Z ns PFS HR 068

P lt 05 in favorof addingcetuximab andor panitumumabto CT in Mutepatients

RR 167 P lt

05 in favorof addingcetuximabandorpanitumumabto CT in Mutepatients

Modest et al14

2012M-A 3 trials

including 119

patients with mCRC

with codon 12

mutations vs other

mutations

Cetuximab CT NR 12 P Z ns NR NR

Loupakiset al31

SR 8 trials including

6609 patients

with mCRC

Mutthorn vs Mute NR NR NR PFS HR 091 95

CI 084-099 Plt 05 in favorof adding anti-EGFR MAbs toCT in Mutepatients(irinotecanfavoring CTP lt 05)

RR 117 95 CI

104-133 P lt

05 in favor ofKRAS Mute

Ku et al28

261 patients with

Cetuximab thorn 5FU

with oxaliplatin vs

irinotecan

Mutthorn vs Mute

NR NR No pooling

performed in

this comparison

AIO trial P Z ns

No pooling

performed in

this comparison

AIO trial P Z ns

(table continues)

The Journal of Molecular Diagnostics - jmdamjpatholorg 195

CECOG trial P lt

05 in favor ofcetuximab thornFOLFOX in Mutepatients

CECOG trial P Zns in favor ofcetuximab thornFOLFOX in Mutepatients

Petrelli et al15

2011SR 7 trials including

5212 patients

with advanced

CRC KRAS Muteonly

Cetuximab or

panitumumab thornCT vs BSC

NR NR HR 084 95 CI

073-098

P lt 05 in favorof anti-EGFRmAbs vs nomAbs in Mutepatients

PFS HR 065 95

CI 051-083 Plt 05 in favorof anti-EGFRmAbs vs nomAbs in Mutepatients

RR 169 95 CI

120-238

P lt 05 infavor of anti-EGFR

Mao et al33

2012SR 13 studies

including 576

patients with

mCRC all KRAS Mutetreated withanti-EGFR MAbs

Mutthorn vs Mute Direct sequencing

survey analysis

alleic

discrimination

Sanger

PIK3CA exon 920

HR 329 95 CI

160-674

P lt 05

PFS HR 252 95

CI 133-478

P lt 05 PIK3CAexon 20mutationsassociated withsignificantlyshorter PFSduration

RR 025 95 CI

005-119

P lt 05PIK3CA exon20 mutationsassociatedwith lower ORR

Lin et al29

2011SR 8 studies

including 5325

patients with

advanced CRC

Mutthorn vs Mute NR NR P Z ns PFS HR 066 95

CI 053-082

P lt 05 in favorof adding anti-EGFR to CT inMute patients

Ibrahimet al13

SR 4 studies

including 2115

patients with mCRC

with Mute KRAS

Panitumumab-based

treatment vs

control

NR NR P Z ns PFS HR 058 95

CI 036-093 Plt 05 in favorof addingpanitumumab toCT in Mutepatients

OR 108 95 CI

075-158

P Z ns

Dahabrehet al22

SR 29 poolable

studies including

5032 patients

with mCRC treated

with anti-EGFR

Mutthorn vs Mute

Cetuximab orpanitumumab thorn CTvs CT alone

NR NR HR 130 95

CI 095-178

P Z ns in Mutepatients

PFS HR 222 95

CI 174-284

P lt 05 in favorof anti-EGFR thornCT in Mutepatients only

Positive

likelihood

ratio 735

(95 CI 372-

Negativelikelihoodratio 055(95 CI 049-061)

KRAS mutationsassociatedwith higherlikelihood ofresponsefailure

Baas et al20

2011SR 21 studies

including w1213

patients with mCRC

(one study N ZNR)

Concordance between

KRAS MutthornMutebetween primaryand metastases

Sequencing

pyrosequencing

PCR-RFLP SSCP

AS-PCR ASO

KRAS PIK3CABRAF or of lossof PTEN

NR NR NR

Adelsteinet al18

SR 11 studies

including 8924

patients with mCRC

treated with anti-

EGFR mAbs

Mutthorn vs Mute

NR 12 13 61 NR PFS HR 080 95

CI 064-099

P lt 05 in favorof anti-EGFRmAbs in Mutepatients

RD 15 95 CI

P lt 05 infavor of KRASMute thorn anti-EGFR treatment

(table continues)

Sepulveda et al

Qiu et al36

2010SR 22 studies

including 2188

patients with mCRC

Mutthorn vs Mute

Cetuximab thorn CT vs CTalone

DS surveyor

analysis qPCR

AD melting

curve analysis

Exon 1 2 Median OS 69 vs

135 months

HR 217 95

CI 172-274

P lt 05 longermedian survivalshown in Mutepatients whoreceived anti-EGFR mAbs thorn CT

Median PFS 30 vs

58 months HR

194 95 CI

162-233 P lt

05 longermedian PFSshown in Mutepatients whoreceived anti-EGFR mAbs thorn CT

KRAS Mute 39KRAS Mutthorn 14RR 024 95 CI

016-038P lt 05

HealthQuality Ontario24

SR 14 observational

studies in patients

with advanced CRC

Mutthorn vs Mute

NR NR Mean OS MD

e411 95 CI

e560 toe262

P lt 05 longersurvival detectedin Mute patientstreated withcetuximabthornirinotecan

Mean PFS MDZe332 95 CI

e486 toe178

P lt 05 longerdurationdetected in Mute patientstreated withcetuximabthornirinotecan

Ibrahimet al26

SR 10 studies

including 2703

patients with mCRC

Mutthorn vs Mute

NR NR P lt 05 in favor oftreatment withcetuximab thorn CTin Mutepatients

PFS P lt 05 infavor oftreatment withcetuximab thorn CTin Mutepatients

OR 210 95 CI

142-310

P lt 05

De Roocket al23

MA 7 studies

including 774

patients with mCRC

who received

cetuximab-based

treatment CT

pG13D vs other

mutation

NR pG13D 13 Median (95 CI)

pG13D 76 months(57-205)

Other mutations57 months(49-68)

Mute 101months(94-113)

P lt 05 pG13Dsuperior toother mutations

Median (95 CI)

pG13D 40 months(19-62)

Mute 42 months(39-54)

Allegra et al19

627 patients with

mCRC and 5 single-

arm studies

including 247

patients

Mutthorn vs Mute PCR direct

sequencing

12 13 No pooling was

performed

No pooling was

performed

No pooling was

performed

Linardouet al30

SR 8 studies

including 817

patients with mCRC

(306 with KRASmutations)

Mutthorn vs Mute NR 12 13 61 NR NR

Sorich et al12

5948 patients

with mCRC

Mutthorn vs Mute

Anti-EGFR mAbtreatment effectsize between RASsubgroupsincluding Mutthorn vsMute

Bidirectional

Sanger

sequencing

pyrosequencing

MALDI-TOF

analysis and

WAVE-based

Surveyor

analysis

KRASNRAS 1213 59 61117 146

RAS Mute vs RASMutthorn

HR 072 (95 CI056-092 P lt

01) RAS Mutesuperior

KRAS exon 2mutant vs newRAS mutantP Z ns

RAS Mute anti-EGFR vs no anti-EGFR HR 087

HR 060 (95 CI048-076 P lt

(table continues)

(95 CI 077-099 P lt 04)

KRAS exon 2 Muteanti-EGFR vs noanti-EGFR HR090 (95 CI083-098P Z ns)

Any RAS mutantanti-EGFR vs noanti-EGFR HR108 (95 CI097-121P Z ns)

KRAS exon 2mutant anti-EGFR vs no anti-EGFR HR 105(95 CI 095-117 P Z ns)

(95 CI 050-076 P lt 001)

KRAS exon 2 Muteanti-EGFR vs noanti-EGFR HR068 (95 CI058-080P lt 001)

Randomized controlled trials (n Z 1)Douillard

et al44

RCT reanalysis of

PRIME trial (NCT

00364013) data

including 1060

patients

RAS Mut andFOLFOX4 anti-EGFR mAb

PCR Sanger

Surveyor

KRASNRAS 1213 61 117146

Mut and anti-

EGFR mAb

26 months vs 202months

HR 078 (95 CI062-099 P lt

05) in favor ofMute and thornanti-EGFR mAb

Mut and anti-

EGFR mAb

HR 072 (95 CI058-090 P lt

Prospective cohort studies (n Z 1)Etienne-

Grimaldiet al45

251 patients KRAS Mutthorn vs KRASMute

NR KRAS 12 13 NR RR 240 (95 CI

127-455 P lt

05) RFSshorter in KRASMutthorn patientswith stage IIItumors

Retrospective cohort studies (n Z 1)Bando et al43

201382 samples from 376

patients

All Mute vs KRAS 1213 vs KRAS 61 146

Luminex xMAP vs

(concordance

rate 100)

KRAS 121361146

All Mute 138

months (92-

184) vs KRASMutthorn 82months (57-107 P lt 05)

All Mute 61

months (31-

All Mute 388

vs KRAS Mutthorn48 P lt 05

AD allelic discrimination-PCR AIO German AIO colorectal study group ARMS amplification refractory mutation system AS-PCR allele-specific polymerasechain reaction ASO allele-specific oligonucleotide BRAF proto-oncogene B-Rafv-Raf murine sarcoma viral oncogene homolog B BSC best supportive careCECOG Central European Cooperative Oncology Group CI confidence interval CISH chromogenic in situ hybridization CPG clinical practice guideline CRCcolorectal cancer CT chemotherapy DS direct sequencing EGFR epidermal growth factor receptor FOLFOX4 folacin 4-fluorouracil oxaliplatin FISHfluorescence in situ hybridization FOLFOX folinic acid (leucovorin calcium) 5-fluorouracil and oxaliplatin 5FU fluorouracil GCN gene copy number HRhazard ratio HTA health technology assessment KRAS Kirsten rat sarcoma viral oncogene homolog M-A meta-analysis MALDI-TOF matrix-assisted laserdesorptionionization-time of flight mCRC metastatic colorectal cancer MD mean difference mAbs monoclonal antibodies Mute mutation negative or wildtype Mutthorn mutation positive NR not reported NRAS neuroblastoma RAS viral (v-ras) oncogene homolog ns nonsignificant OR odds ratio ORR objectiveresponse rate OS overall survival PCR polymerase chain reaction PCR-RFLP polymerase chain reactionerestriction fragment length polymorphism PFSprogression-free survival PIK3CA phosphatidylinositol-45-bisphosphate 3-kinase catalytic subunit alpha PRIME Panitumumab Randomized Control Trial inCombination with Chemotherapy for Metastatic Colorectal Cancer to Determine Efficacy PTEN phosphatase and tensin homolog qPCR quantitative polymerasechain reaction RCT randomized controlled trial RAS rat sarcoma viral oncogene homolog RD risk difference RFS recurrence-free survival RR response rateSISH silver in situ hybridization SR systematic review SSCP single-strand conformation polymorphism xMAP multiplex assayTests used by Ren et al37 hybridization PCR direct sequencing topographic genotyping AS-PCR tissue transglutaminase enzyme high-performance liquid

chromatography pyrosequencing capillary sequencing

Sepulveda et al

of individual CRCs showing co-occurrence of RAS and RAFmutations312

Cetuximab and panitumumab are antibodies that bind tothe extracellular domain of EGFR blocking the binding ofEGF and other EGFR endogenous ligands thereby blockingEGFR signaling Earlier studies reported the effects of anti-EGFR antibody treatment independent of KRAS status13-16

However it was later reported that targeted EGFR therapieswith cetuximab or panitumumab improve PFS and OS inpatients with metastatic CRCwith wild-typeKRAS but not forpatients with mutatedKRAS2317 In these earlier studies onlymutations of KRAS exon 2 were considered Based on theavailable clinical trial data in 2009 the ASCO recommendedthat patients with metastatic CRCwho are candidates for anti-EGFR antibody therapy should have their tumor tested forKRAS mutations in a Clinical Laboratory ImprovementsAmendments rsquo88 (CLIA)eaccredited laboratory2

A large body of evidence was available to guide therecommendation in the current guideline for RAS testing incolorectal cancers (Table 5 and Supplemental Table 14 allsupplemental materials can be found at httpdxdoiorg101016jjmoldx201611001) From 2008 to 2015 there were311 primary studies that included 74546 patients and re-ported treatment outcomes for patients with RAS mutationscompared with nonmutatedwild type12-1618-45 The mostcommon comparison of anti-EGFR antibody treatmentoutcomes was between KRAS mutation vs KRAS non-mutatedwild type18-202224-2628-3133-42 Some studies alsocompared the effects of adding an anti-EGFR inhibitor toKRAS nonmutatedwild-type patients vs chemotherapyalone182224262836-38 A few studies reported anti-EGFRantibody treatment outcomes for the following compari-sons KRAS G13D vs codon 12 mutations32 KRAS codon13 mutations vs other mutations21 and G13D vs other exon2 mutations23

The reported anti-EGFR therapy outcomes in thesestudies were pooled survival13-1621-272932-373941 pooledPFS1315161821-272931-363941 and pooled objectiveresponse rate (ORR)131516182122252630-3641 Thirteenstudies reported significant differences between compara-tors152123-27323335-3739 The systematic review literatureof data on anti-EGFR therapy outcomes is presented inSupplemental Table 14 Five of these studies detected asignificant pooled survival advantage of antieEGFR-treatedpatients for KRAS nonmutatedwild type compared withKRAS mutation2133353739 Three studies detected anadvantage for patients with nonmutated tumors given anti-EGFR treatment compared with KRAS mutation-positivepatients given chemotherapy alone242636 Twenty of theincluded studies pooled PFS1315161821-272931-363941 with19 reporting significant differences between compara-tors13151821-272931-363941 Fourteen papers detected asignificant PFS advantage for adding an anti-EGFR inhibi-tor to chemotherapy for KRAS nonmutatedwild-type pa-tients compared with chemotherapy alone1315182224-2629313334363941 Sixteen of the included papers pooled

ORR131516182122252630-3641 with 14 reporting significantdifferences between comparators15182122252630-3641

Eight studies detected ORR advantages for adding an anti-EGFR inhibitor to chemotherapy for patients withnonmutatedwild-type tumors compared with chemotherapyalone1825263033343641 and four detected an ORR advan-tage for KRAS nonmutatedwild-type patients over mutationpatients22313235 Survival advantages (OS and PFS ORR)for G13D mutations over codon 12 and G13D over othermutations were reported in two studies2332 and codon 13over other KRAS mutations21

Recent studies showed conclusive evidence that in addi-tion to mutations in KRAS exon 2 other RAS mutations inKRAS exons 3 and 4 and NRAS exons 2 3 and 4 were alsoassociated with nonresponse of metastatic CRC to anti-EGFR monoclonal antibody therapy124446 Douillardet al44 published a reanalysis of the Panitumumab Ran-domized Control Trial in Combination with Chemotherapyfor Metastatic Colorectal Cancer to Determine Efficacy(PRIME) trial reporting that patients with any RAS muta-tions were associated with inferior PFS and OS withpanitumumab-FOLFOX4 treatment which was consistentwith the findings previously reported for patients with KRASmutations in exon 2 Subsequently a meta-analysis of ninerandomized clinical trials provided further evidence that notall KRAS exon 2 nonmutatedwild-type tumors benefit fromanti-EGFR monoclonal antibody treatment in metastaticCRC12 Patients with colorectal cancers that are KRAS exon2 nonmutatedwild type but harbor RAS mutations in KRASexons 3 and 4 or NRAS exons 2 3 and 4 also havesignificantly inferior anti-EGFR treatment outcomes benefitcompared with those without any RAS mutations (Table 5and Table 6) RAS mutations occur mostly at exon 2 fol-lowed by mutations in exons 3 and 4 (Table 7) The resultssuggest that ldquoextendedrdquo or ldquoexpandedrdquo RAS mutation testing(KRAS exons 2 3 and 4 and NRAS exons 2 3 and 4) mustbe performed before the administration of an anti-EGFRmonoclonal antibody therapy12 In summary current evi-dence indicates that both cetuximab and panitumumabshould only be prescribed for patients with metastatic CRCsthat are nonmutatedwild type for all known RAS-activatingmutations12

This recommendation is supported by 34 studies12-1618-4547 comprising 29 systematic studies1213151618-2224-4247 two meta-analyses1423 one randomized controlledtrial44 one prospective cohort study45 and one retrospectivecohort study43

Of the 29 systematic reviews1213151618-2224-4247 onlythree reported using a multidisciplinary panel192530 andonly one reported taking patient preferences into account37

although 13 examined important patient sub-types12151618212224273033373940 All but one had well-described and reported methods sections42 Seven did notreport on conflict of interest13151634384142 Only ninerated the quality of the included evidence and these samenine were the only ones that reported on the strength of the

Table 6 Outcomes of RAS Mutations and Anti-EGFR Therapy12

Characteristic

Overall Survival Progression-Free Survival

HR (95 CI) P Value HR (95 CI) P Value

RAS nm vs RAS mutation RAS nm superior 072 (056-092) lt01 060 (048-076) lt001KRAS exon 2 mutant vs new RAS mutant ns nsKRAS nm exon 2 anti-EGFR vs no anti-EGFR 090 (083-098) ns 068 (058-080) lt001KRAS exon 2 mutant anti-EGFR vs no anti-EGFR 105 (095-117) ns 114 (095-136) nsRAS nm anti-EGFR vs no anti-EGFR 087 (077-099) lt04 062 (050-076) lt001Any RAS mutant anti-EGFR vs no anti-EGFR 108 (097-121) ns 112 (094-134) ns

CI confidence interval EGFR EGFR epidermal growth factor receptor HR hazard ratio KRAS Kirsten rat sarcoma viral oncogene homolog nm nonmutatedns nonsignificant RAS rat sarcoma viral oncogene homolog

Sepulveda et al

included evidence161821222425323739 None of the studiesincluded a plan for updating None of the systematic re-views reported industry funding two reported no fund-ing1631 and 11 did not report on the source of funding ifany1315263234-3638414247 Two of these systematic re-views were deemed to have a low risk of bias2437 14 weredeemed to have a low to moderate risk ofbias1216181921222527293032353947 12 were deemed tohave a moderate risk of bias131520262831333436384041

and one was deemed to have a high risk of bias42

Of the two meta-analyses obtained1423 both had well-reported and reproducible methods sections bothdescribed the planned pooling a priori and both discussedthe limitations of their analyses Neither was based on asystematic review of the literature and neither did aquality assessment of the included studies One reportednonindustry funding23 and the other reported industry

Table 7 Prevalence of New RAS Mutations Across Studies

Study New RAS Totaly KRAS Exon 3y KRAS Exo

Codons 59 61 Codons 1OPUS 263 59 93PICCOLO 98 NRz 37x

20020408 176 48z 5020050181 205 46 79PRIME 174 37z 56FIRE-3 160 43z 49x

PEAK 201 41 77COIN 84 21c NECRYSTAL 147 33 56Summary (95 CI) 199 (167-234) 43 (33-55) 67 (57

CI confidence interval COIN Combination Chemotherapy With or Without CetuCancer Trial CRYSTAL Cetuximab Combined with Irinotecan in First-Line Therapy(FOLFIRI) Plus Cetuximab vs FOLFIRI Plus Bevacizumab in First-Line Treatmentevaluated but not reported OPUS Effect of Roflumilast on Exacerbation Rate in PPEAK Panitumumab Plus mFOLFOX6 vs Bevacizumab Plus mFOLFOX6 for First-LineKirsten Rat Sarcoma-2 Virus (KRAS) Tumors Trial PICCOLO Panitumumab and IrinResistant Advanced Colorectal Cancer Trial PRIME Panitumumab Randomized TrDetermine Efficacy TrialModified from Sorich et al12 by permission of Oxford University Press on behayNew RAS mutations are reported as a proportion of the KRAS exon 2 nonmutazKRAS and NRAS codon 59 mutation was not evaluatedxKRAS codon 117 mutation was not evaluatedExon 3 codon 61 mutations in addition to the exon 2 mutationskOnly NRAS mutation G12C evaluatedRandom-effects meta-analysis summary estimates

funding14 One was deemed to have a low to moderate riskof bias23 and the other was deemed to have a moderaterisk of bias14

The single randomized controlled trial did not report on anydetails of the randomization including blinding the expectedeffect size and power calculation and the length of follow-up44 It did report on differences in baseline patient charac-teristics This trial did report at least partial industry fundingand was deemed to have a low to moderate risk of bias44

The single prospective cohort study reported a balancebetween treatment and assessment groups reported onbaseline characteristics and made adjustments in the anal-ysis when differences were found45 It reported nonindustryfunding and was deemed to have a low risk of bias45

The single retrospective cohort study reported that thetreatment and assessment groups were in balance and alsoreported on baseline patient characteristics43 It did not

n 4y NRAS Exon 2y NRAS Exon 3y NRAS Exon 4y

17 146 Codons 12 13 Codons 59 61 Codons 117 14668 51 0863 NRz NE42 30z 1123 58 0034 41z 0038 20z 0054 59 0009k 30z NE35 28 09

-79) 38 (30-48) 48 (34-68) 05 (02-12)

ximab as First-Line Therapy in Treating Patients With Metastatic Colorectalfor Metastatic Colorectal Cancer Trial FIRE-3 Folinic Acid and IrinotecanColorectal Cancer (CRC) Trial NA not applicable NE not evaluated NRatients With Chronic Obstructive Pulmonary Disease (BY217M2-111) TrialTreatment of Metastatic Colorectal Cancer (mCRC) Patients With Wild-Typeotecan vs Irinotecan Alone for Patients With KRAS Wild-Type Fluorouracil-ial in Combination With Chemotherapy for Metastatic Colorectal Cancer to

lf of the European Society for Medical Oncologytedwild-type group

report that adjustments were made in the analysis to accountfor differences where differences were found This studyreported nonindustry funding and was deemed to have a lowrisk of bias43

All of the evidence that supported this recommendationwas assessed and none was found to have methodologicflaws that would raise concerns about their findings

2a Recommendation BRAF pV600 [BRAF c1799(pV600)] position mutational analysis should be performedin CRC tissue in selected patients with colorectal carcinomafor prognostic stratification

BRAF activating mutations occur in about 8 ofadvanced disease patients with CRC4748 and in approxi-mately 14 of patients with localized stage II and IIICRC849 As such mutations in BRAF constitute a sub-stantial subset of patients with CRC The key questionsrelated to BRAF mutations are whether patients whosecancers carry a BRAF mutation have a poorer outcomecompared with BRAF mutation-negative tumors andwhether the presence of a mutation predicts benefit from orlack thereof to anti-EGFR therapy

Four systematic reviews2050-52 and three systematic re-views that included meta-analyses474853 pertaining to theprognostic and predictive value of BRAF mutations inpatients with CRC were identified through our systematicreview process (Table 8 and Supplemental Table 14) Thesestudies revealed that patients with advanced CRC whopossess a BRAF mutation have significantly poorer out-comes as measured by PFS and OS and have a decreasedresponse rate to anti-EGFR therapy relative to those withnonmutated BRAF Poorer OS was also demonstrated forthose patients with earlier stage II and III CRC having aBRAF mutation854 however the poorer outcome appears tobe primarily the result of decreased OS after relapse in thesepatients rather than a harbinger of an increased rate ofrelapse Finally while outcomes in advanced disease pa-tients with BRAF mutations were poorer relative to non-mutation patients the data were consistent with a modestbeneficial impact from the use of anti-EGFR agents relativeto those patients whose tumors contained a RAS mutation55

In summary patients with CRC that contains a BRAFmutation have a worse outcome relative to nonmutationpatients Selected patients for BRAF mutation testinginclude patients with metastatic disease since these patientshave particularly poor outcomes It is important to know theBRAF c1799 (pV600) mutation status of a patientrsquos CRCsince standard therapy is not adequate for patients withmetastatic disease and BRAF mutation For these patientssome studies suggest the use of FOLFIRINOX [folinic acid(leucovorin calcium) 5-fluorouracil irinotecan hydrochlo-ride and oxaliplatin] as first-line therapy followed byenrollment in a clinical trial56 Furthermore early clinicaltrials data suggest that the combination of a BRAFplus EGFR inhibitor appears to be effective in this popu-lation57-59 Data in support of molecular testing for BRAF

c1799 (pV600) mutations in CRC continue to emerge fromclinical trials A recent publication of the PETACC-8(oxaliplatin fluorouracil and leucovorin with or withoutcetuximab in patients with resected stage III colon cancerrandomised phase 3) trial reported that trials in the adjuvantsetting should consider mismatch repair BRAF and KRASstatus for stratification since BRAF pV600 and KRASmutations were associated with shorter DFS and OS in pa-tients with microsatellite-stable colon cancer but not in thosewith tumors with MSI6061

This recommendation is supported by seven systematicreviews20474850-53 three of which included meta-anal-ysis474853 None of the systematic reviews reported thecomposition of their panel so multidisciplinary panel rep-resentation could not be confirmed and none reported pa-tient representation on the panel All but the systematicreview reported by Baas et al20 reported examining impor-tant patient subgroups All of the systematic reviews re-ported well-described and reproducible methods Three didnot report how conflicts of interest were managed and re-ported on475153 Only two reported on a quality assessmentof the included literature4850 and only one rated thestrength of the evidence50 None reported a plan forupdating While none of the systematic reviews reportedindustry funding one study did not report any fundingsupport47 Overall the risk of bias assessment for this bodyof evidence ranged from low4850 to moderate205153 andnone were found to have methodologic flaws that wouldraise concerns about their findings

2b Recommendation BRAF pV600 mutational anal-ysis should be performed in dMMR tumors with loss ofMLH1 to evaluate for Lynch syndrome risk Presence of aBRAF mutation strongly favors a sporadic pathogenesisThe absence of BRAF mutation does not exclude risk ofLynch syndrome

dMMR occurs via several mechanisms In sporadic CRCdMMR is most frequently caused by epigenetic silencingthrough CpG methylation primarily of MLH1 with fewcases resulting from somatic mutation of one of the MMRgenes In Lynch syndrome CRC the underlying mechanismis usually a germline mutation of one of the four (MLH1MSH2 MSH6 and PMS2) mismatch repair genes and inrare patients a deletion involving EPCAM (epithelial celladhesion molecule) a gene adjacent to MSH2 that leads toepigenetic inactivation of the MSH2 gene dMMR occurs in15 to 20 of all colorectal cancers and of these aboutthree-fourths are due to MLH1 epigenetic silencing562

dMMR underlies widespread mutations in the genome andMSI BRAF pV600 mutations rarely occur in patients withgermline-based dMMR but have been reported in up tothree-fourths of those with epigenetic MMR gene silencing(Table 8 and Table 9) Thus testing for BRAF mutationsserves as a means for distinguishing germline from epige-netic dMMR particularly in those cases where the dMMR isthe result of epigenetic silencing of MLH1 For tumors with

Table 8 BRAF Clinical Practice Guidelines Systematic Reviews Meta-Analyses Prospective Cohort Studies and Retrospective CohortStudies

Author Year Study Type and Evidence Comparison Tests Used

CPGs systematic reviews and meta-analyses (n Z 8)Parsons et al52 2012 SR 36 studies including 4562 CRC

tumors (BRAF ) 43 studies including2975 CRC tumors (MLH1)

Correlation study NR

Mao et al51 2011 SR 11 studies including 1046patients with mCRC

Mutthorn vs Mute NR

Lin et al50 2011 SR 1 study of 649 patients withmCRC all KRAS Mute 65 wereBRAF Mutthorn

Mutthorn vs Mute NR

Baas et al20 2011 SR 7 studies including 538 patientswith mCRC

Mutthorn vs Mute Sequencing pyrosequencing

Cui et al53 2014 SR 4 studies including 1245 patients Mutthorn vs MuteCT anti-EGFR mAbs

Yang et al71 2013 SR 17 studies (patients n Z NR) Mutthorn vs Mute d

Yuan et al48 2013 SR 21 studies including 5229patients

Mutthorn vs Mute NR

Xu et al47 2013 SR 19 studies including 2875patients

Mutthorn vs Mute NR

Prospective cohort studies (n Z 1)Etienne-Grimaldi et al45 2014 251 patients Mutthorn vs Mute NR

Retrospective cohort studies (n Z 1)Bando et al43 2013 82 samples from 376 patients All Mut- vs BRAF Mutthorn and

PIK3CA MuteLuminex xMAP vs DS(concordance rate 100)

(table continues)

BRAF proto-oncogene B-Rafv-Raf murine sarcoma viral oncogene homolog B CI confidence interval CPG clinical practice guideline CRC colorectalcancer CT chemotherapy DS direct sequencing EGFR epidermal growth factor receptor HR hazard ratio KRAS Kirsten rat sarcoma viral oncogenehomolog mAbs monoclonal antibodies mCRC metastatic colorectal cancer MLH1 mutL homolog 1 Mute mutation negative or wild type Mutthornmutation positive NR not reported ns nonsignificant ORR objective response rate OS overall survival PCR polymerase chain reaction PFSprogression-free survival PIK3CA phosphatidylinositol-45-bisphosphate 3-kinase catalytic subunit alpha RR response rate RFS recurrence-freesurvival SR systematic review xMAP multiplex assayYang et al71 adenovirus-PCR pyrosequencing allele-specific PCR DS PCR amplification quantitative PCR Sanger real-time PCR genotypingthornDS PCR

clamping melting curve analysis DNA sequencing and Taqman single-nucleotide polymorphism assay

Sepulveda et al

Table 8 (continued)

Codons Studied OS PFS ORR

BRAF pV600E MLH1 NR NR NR

V600E NR NR BRAF Mutthorn 0 BRAF Mute 363P lt 05 RR 014 95 CI 004-053

V600E Shorter duration in BRAF Mutthornpatients difference 28 weeksP lt 05

PFS shorter duration in BRAFMutthorn patients difference 18weeks P lt 05

V600E NR NR NR

V600E NR NR Mutthorn vs Mute (all KRAS Mute) RR043 (95 CI 016-075 P lt 05)in favor of Mute

Mut vs CT anti-EGFR mAbs (allKRAS Mute) RR 038 (95 CI020-073 P lt 05) in favor ofMute

Mutthorn and CT anti-EGFR mAbsP Z ns

Mute and KRAS Mute and CT anti-EGFR mAbs RR 148 (95CI 128-171 P lt 05) in favor ofBRAF Mute with CT thorn anti-EGFRmAbs

V600E 599 466 469 (7 studies)BRAF Mut HR 274 (95 CI179-419 P lt 05) in favor ofBRAF Mute

(8 studies)BRAF Mut HR 259 (167 403 P lt 05) infavor of BRAF Mute

BRAF Mute 464BRAF Mut 185P lt 05 in favor of BRAF Mute

V600E HR 035 (95 CI 029-042P lt 05) in favor of BRAF Mute

HR 038 (95 CI 029-051P lt 05) in favor of BRAF Mute

RR 031 (95 CI 018-053P lt 05) in favor of BRAF andKRAS Mute

V600E K601E (1 study)D549C (1 study)

ORR 058 (95 CI 035-094P lt 05) in favor of BRAF Mute

BRAF pV600E NR Shorter RFS in KRAS Mute andBRAF Mute patients with stageIII tumors (P lt 05)

600 All Mute 138 months (95 CI92-184) vs BRAFPIK3CA Mut63 months (95 CI 13-113P lt 05)

All Mute 61 months (95 CI31-92) vs BRAFPIK3CA Mutthorn16 months (95 CI 15-17P lt 05)

All Mute 388 vs BRAFPIK3CAMutthorn 0 P lt 05

Table 9 Summary of Frequencies of Tumor V600E Mutation Status

Sample GroupNo ofStudies

PositivepV600E No

Negative BRAFpV600E No BRAF pV600E (95 CI)

Known negative MMR mutation statusMSI-H known mutation status 11 115 216 3610 (2095-5284)MLH1 methylation or MLH1 loss of expression(known or assumes MSI-H status)

9 191 141 6350 (4698-7853)

MSS 11 85 1538 500 (355-668)Known positive MMR mutation status

All mutation carriers 26 4 546 140 (006-225)

BRAF proto-oncogene B-Rafv-Raf murine sarcoma viral oncogene homolog B CI confidence interval MMR mismatch repair MSI-H microsatelliteinstability high MLH1 mutL homolog 1 MSS microsatellite stableAdapted from Parsons et al52 by permission from BMJ Publishing Group Limited

Sepulveda et al

a mutation in BRAF and dMMR it may be concluded thatthe basis for their dMMR is less likely to be germline55262

In contrast tumors with dMMR in the absence of a BRAFmutation may have either germline or an epigenetic (MLH1gene promoter hypermethylation) basis for the dMMR andspecific testing for MLH1 promoter hypermethylation maybe used to further refine the risk of Lynch syndrome beforeinitiating definitive genetic testing Identification of thosepatients with germline-based dMMR has clear implicationsfor the patientrsquos family members

3 Recommendation Clinicians should order mismatchrepair status testing in patients with colorectal cancers forthe identification of patients at high risk for Lynch syn-drome andor prognostic stratification

The molecular pathology underlying most MSI tumors issomatically acquired CpG methylation of the promoter ofthe gene MLH1 About three-fourths of colorectal cancerswith MSI due to MLH1 promoter hypermethylation willhave an acquired BRAF mutation as well The reason for thisis not understood Less than one-third of individuals with

Table 10 Mismatch RepairMicrosatellite Instability Systematic Review

Guastadisegniet al7 2010

SR 31 studies including12782 patients withCRC

MSI vs MSS MSI by PCR iIHC in 6 st

Des Guetzet al6 2009

SR 7 studies including3690 patients with CRCon effect of adjuvantchemotherapy

1444 treated with 5-FUebased therapy and1518 not treated

MSI vs MSS PCR in all an2 studies

CI confidence interval CRC colorectal cancer DFS disease-free survival 5microsatellite instability MSI-H microsatellite instability high MSS microsatepolymerase chain reaction PFS progression-free survival RFS relapse-free surviv

dMMRMSI colorectal tumors do not have underlyingMLH1 promoter hypermethylation but rather have a germ-line mutation affecting any one of the four DNA MMRgenes noted above Individuals with germline mutations inthe MMR genes are said to have Lynch syndrome anautosomal dominant disorder that confers dramaticallyincreased risks for colorectal and endometrial cancers andmoderately increases risks for a variety of other tumors63

Diagnosis of Lynch syndrome is important as active man-agement of cancer risks has been demonstrated to benefitgene mutation carriers56465 and establishing a diagnosiscreates opportunities for prevention among all at-risk rela-tives Testing for dMMR can be performed by immuno-histochemistry for the four MMR proteins (MLH1 MSH2PMS2 and MSH6) or by MSI DNA-based testing as dis-cussed in detail in a report by Funkhouser et al66 (recom-mendation 11)A systematic review of 31 studies7 reporting survival on

12782 patients whose tumors were characterized for MSIshowed a favorable prognosis as determined by both OS

OS PFS

n all andudies

OR 06 95 CI 053-069 P lt 0001 MSI isassociated with longersurvival

DFS OR 058 95 CI047-072 P lt 0001MSI is associated with alonger PFS duration

d IHC in MSI-HHR 070 95 CI 044-109 P Z ns nosignificant benefit ofchemotherapy in MSI-Hpatients

MSI-HRFS HR 096 95 CI062-149 P Z ns nosignificant difference iftreated or not treated

MSI-H vs MSSRFS HR 077 95 CI067-087 P lt 05 MSIpatients had no effectof treatment comparedwith beneficial effect inMSS patients

-FU 5-fluorouracil HR hazard ratio IHC immunohistochemistry MSIllite stable ns nonsignificant OR odds ratio OS overall survival PCRal SR systematic review

and DFS (Table 10) but this is dependent on stage Inaddition the presence of MSI in CRC was reported to bepredictive for nonresponse to 5-fluorouracilebased adjuvantchemotherapy of early stage disease6 although this has notbeen corroborated (Table 10)67 Emerging data indicate thatMMR status may have predictive value in some settingsspecifically in patients with advanced disease beingconsidered for anti-programmed cell death protein-1 (PD-1)programmed cell death ligand protein-1 (PD-L1) immunecheckpoint inhibitor therapy68-70

This recommendation is supported by two systematicreviews that included 38 studies and 16472 patients67

Both of these systematic reviews included a well-described and reproducible methods section and both re-ported on potential conflicts of interest Only one the sys-tematic review reported by Guastadisegni et al7 reported thesource of funding which was nonindustry Due to deficits inthe reporting one of these systematic reviews was deemedto have a moderate risk of bias6 and the other was deemedto have a low to moderate risk of bias7 however neither ofthese were found to have any major methodologic flaws thatwould cause us to question their findings

4 No Recommendation There is insufficient evidenceto recommend BRAF c1799 (pV600) mutational status as apredictive molecular biomarker for response to anti-EGFRinhibitors

As noted in recommendation 2a mutations in positionpV600 in BRAF are associated with poor prognosisespecially in patients with metastatic disease Responserates to chemotherapy regimens including regimens withcetuximab and panitumumab are lower in patientsharboring BRAF pV600 mutations515371 (Table 8)Similarly the PFS and OS after treatment with EGFRmonoclonal antibodies in combination with chemotherapyare lower in patients with BRAF pV600 mutations4748

Many of these analyses used nonrandomized cohortsthereby making evaluation of the potential predictive valueof the BRAF pV600 mutation impossible to discern(Table 8) In addition the poor prognosis and low muta-tion prevalence make evaluation of the relative benefit ofEGFR inhibitors difficult to evaluate in individual ran-domized clinical trials

Meta-analyses of randomized studies of EGFR mono-clonal antibodies have been completed to address thequestion of the predictive role of BRAF pV600 mutationsA meta-analysis of 463 patients with KRAS wild-type andBRAF pV600 mutated tumors did not provide sufficientevidence to exclude a magnitude of benefits seen in KRASBRAF wild-type tumors Nor was there sufficient evidenceto identify a statistically significant benefit to this treat-ment55 A second meta-analysis showed that EGFR mono-clonal antibody treatment in patients whose tumors containa BRAF pV600 mutation was not associated with signifi-cant OS (P Z 43) although there was a trend for betterPFS (P Z 07)72 This suggests insufficient evidence to

recommend the use of BRAF pV600 as a predictive markerfor benefit of anti-EGFR monoclonal antibodies More dataare required to definitively determine the predictive value ofBRAF mutations relative to anti-EGFR therapy

This recommendation was supported by five systematicreviews4748515371 (Table 8) None of these systematic re-views reported forming a multidisciplinary panel and nonereported including patient representatives in developing theirresearch questions or interpreting their outcomes All of thesystematic reviews examined important patient subtypes andall used well-described and reproducible methods Only thesystematic review by Yuan et al48 reported on any potentialconflicts of interest the article by Mao et al51 stated conflictswere not examined and the other three did not report any-thing regarding conflicts475371 Only two the systematicreviews reported by Yang et al71 and Yuan et al48 rated thequality of the included evidence although none of the studiesreported on the strength of the evidence None of the studiesdiscussed any plans for future updating Four reportednonindustry funding for their systematic reviews48515371

and one did not report the source of funding if any47 Twoof the systematic reviews were deemed to have a low riskof bias4871 one was deemed to have a low to moderate risk ofbias47 and two were deemed to have a moderate risk ofbias5153 Overall none of the systematic reviews were foundto have methodologic flaws that would raise concerns abouttheir findings

5 No Recommendation There is insufficient evidenceto recommend PIK3CA mutational analysis of colorectalcarcinoma tissue for therapy selection outside of a clinicaltrial

Note Retrospective studies have suggested improvedsurvival with postoperative aspirin use in patients whosecolorectal carcinoma harbors a PIK3CA mutation

Despite comprehensive RAS testing (recommendation 1)many patients still fail to respond to EGFR monoclonalantibody therapy Additional biomarkers to guide patientselection for such therapy are desired

PIK3CA mutations are observed in 10 to 18 of pa-tients with CRC primarily in exons 9 and 20 and lead to aconstitutive activation of p100a enzymatic activity leadingto an increased PI3K activity and high oncogenic trans-formation ability However mutations of KRAS or NRASand PIK3CA mutations can be detected alternatively and insome cases concurrently in a single CRC38 PIK3CA mu-tations are positively correlated with KRAS exon 12 and 13mutations3 Several meta-analyses and one individual pa-tient data large pooled analysis have examined the prog-nostic role of PIK3CA in patients with stage IV CRC bothoverall and in the KRAS nonmutatedwild-type populationThese studies have generally indicated poorer response rateand PFS in patients with the PIK3CA mutation a findingthat appears to be driven primarily by patients with exon 20mutation3335071 (Table 11) These meta-analyses haveincluded many of the same studies as well as observed and

Table 11 PIK3CA Clinical Practice Guidelines Systematic Reviews Meta-Analyses Prospective Cohort Studies and Retrospective CohortStudies

CPGs systematic reviews and meta-analyses on PIK3CA Mutthorn vs Mute (n Z 5)Wu et al73 2013 SR 8

839 patients with mCRC who all receivedanti-EGFR mAbs

Mutthorn vs Mute Sanger allelic discrimination directsequencing pyrosequencing

Mao et al33 2012 SR 13 studies including patients all KRASMutetreated with anti-EGFR mAbs

E20 Mutthorn vs E20 Mute NR

Lin et al50 2011 SR 4 studies 1030 patients with mCRCall KRAS Mute subgroup analysisexons 9 and 20

Mutthorn vs Mute NR

Baas et al20 2011 SR 3 studies including 195 patients withmCRC

Yang et al71 2013 SR 10 studies (patient number Z NR) Mutthorn vs Mute DS PCR amplification AS-PCRgenotyping RT-PCR Sanger DNAsequencing pyrosequencing

Retrospective cohort studies (n Z 1)Bando et al43 2013 82 samples from 376 patients All Mutevs BRAF Mutthorn

and PIK3CA MutthornLuminex xMAP vs DS (concordancerate 100)

(table continues)

AS-PCR allele-specific polymerase chain reaction CI confidence interval CPG clinical practice guideline DS direct sequencing EGFR epidermal growthfactor receptor HR hazard ratio mAbs monoclonal antibodies mCRC metastatic colorectal cancer Mute mutation negative or wild type Mutthorn mutationpositive NR not reported ns nonsignificant ORR objective response rate OS overall survival PCR polymerase chain reaction PFS progression-freesurvival PIK3CA phosphatidylinositol-45-bisphosphate 3-kinase catalytic subunit alpha RD risk difference RR response rate RT-PCR reverse transcriptionpolymerase chain reaction SR systematic review xMAP multiplex assay

Sepulveda et al

acknowledged between-study heterogeneity and all haveconcluded further prospective data are necessary Contra-dictory recent studies have also been recently reported74

None of the studies considered the independent role ofPIK3CA in the context of comprehensive RAS testing DeRoock et al3 estimated that comprehensive PIK3CA testingwould increase response rate in the first-line setting by only1 The prognostic impact of PIK3CA in stage I to IIIdisease has been inconsistent75-77

Multiple prospective observational studies have demon-strated an association between aspirin use and decreasedCRC mortality78-80 Data on aspirin as a treatment for CRC(postdiagnosis usage) are more limited and drawn only fromobservational studies Domingo et al81 and Liao et al82

found a survival advantage for posttreatment aspirin usersonly in patients whose tumors exhibit PIK3CA mutationshowever a recent cohort study did not validate these ob-servations83 Multiple prospective studies are under way toaddress the potential benefit of adding aspirin or othernonsteroidal anti-inflammatory drugs to adjuvant therapy

This recommendation is supported by two systematicreviews3340 obtained from our systematic review Nonereported the composition of a multidisciplinary panel re-ported patient representation or study quality rated strengthof the evidence reviewed or disclosed a plan for updating

However both systematic reviews did include relevant pa-tient subgroups and included methods that were welldescribed and reproducible In both systematic reviewsinformation about the potential conflicts of the panelists wasreported and funding was provided by nonindustry sourcesBoth were found to have a moderate risk of bias but neitherof the studies providing the evidence base for recommen-dation 5 were found to have methodologic flaws that wouldraise concerns about their findingsAt the present time the retrospective data for the use of

PIK3CA mutation to deny anti-EGFR antibody therapy inpatients with stage IV CRC or as a selection factor for use ofaspirin in stage I to III tumors are insufficient for clinical useoutside of a clinical trial

6 No Recommendation There is insufficient evidenceto recommend PTEN analysis [expression by immunohis-tochemistry (IHC) or deletion by fluorescence in situ hy-bridization (FISH)] in colorectal carcinoma tissue forpatients who are being considered for therapy selectionoutside of a clinical trialPTEN functions as a tumor suppressor gene and loss of

PTEN results in upregulation of the PI3KAKT pathwayPTEN mutations occur in approximately 5 to 14 ofcolorectal cancers484 and loss of PTEN expression can be

Table 11 (continued)

Exons 9 20 HR 128 95 CI 105-156 P lt 05patients with PIK3CA Mutthorn hadshorter PFS

PFS HR 153 95 CI 128-184P lt 05 patients with PIK3CA Mutthornhad shorter PFS

Exon 20 HR 329 95 CI 160-674 P lt 05 PFS HR 252 95 CI 133-478P lt 05 PIK3CA exon 20 mutationsassociated with shorter PFS

ORRExon 20 Mutthorn 0Exon 20 Mute 37RR 025 95 CI 005-119 P Z ns(subset 377 patients)

Exons 9 20 P Z ns no difference between Mutthornand Mute patients

Exon 20 Mutthorn predicts poorer survival

P Z ns no difference between Mutthornand Mut- patients

Exons 9 20 NR NR NR

Exons 7 8 918 19 20

(6 studies)HR 143 (95 CI 102-20 P lt 05)in favor of Mute

(6 studies)HR 191 (95 CI 078-468 P Z ns)P lt 05 in favor of exon 9 comparedwith exon 20 mutations

(6 studies)RD e23 (-35 -10 P lt 05) infavor of exon 9 compared exon 20mutations

Exon 9 All Mute 138 months (95 CI 92-184) vs BRAFPIK3CA Mutthorn 63months (95 CI 13-113 P lt 05)

All Mute 61 months (95 CI 31-92) vs BRAFPIK3CA Mutthorn 16months (95 CI 15-17 P lt 05)

All Mut- 388 vs BRAFPIK3CAMutthorn 0 P lt 05

observed in tumors with KRAS BRAF and PIK3CAmutations

Although there is evidence suggesting that PTEN is acritical factor in cancer development the association be-tween PTEN expression and predictiveprognostic valueremains controversial with several studies suggesting anassociation with poorer prognosis and others finding noassociation at all Four systematic reviews were obtainedthat reported on loss of PTEN expression compared withnormal PTEN expression and 31 primary studies includinga total of 2545 patients20508586 (Supplemental Table 14)Tests used included IHC and FISH Of the four studies thatreported overall survival rates20508586 three studies re-ported on pooled outcomes508586 One study reported asignificant difference in favor of normal PTEN expression86

and the others reported no significant differences205085 ForPFS three studies pooled outcomes508586 two detected asignificant difference in favor of normal PTEN expres-sion8586 and one showed no significant difference50 ForORR two studies pooled outcomes and both found loss ofPTEN expression associated with a poorer response8586

Several studies have shown an association between PTENloss and local recurrence advanced TNM stage lymph nodemetastasis and a lower 5-year survival rate87-90 Howeverseveral other studies have found no correlation betweenPTEN status and patient survival tumor grade TNM stagelymphatic invasion and liver metastasis91-93 Regardingresponse to EGFR-targeted therapies several studies have

shown an association with PTEN loss and lack of responseto cetuximab and panitumumab94-97 However other pub-lished studies failed to demonstrate a clear correlationbetween loss of PTEN expression and response to anti-EGFR therapy98-100 Given the significant discordance inresults the role of PTEN as a prognostic or predictivebiomarker in CRC is still largely unknown and researchinto the prognostic and predictive significance of PTEN isongoing

This recommendation is supported by 20 studies4205084-100 four20508586 of which met the inclusion criteria forinclusion in our systematic review All four of these weresystematic reviews and included 42 studies and 3412 pa-tients None of these systematic reviews reported using amultidisciplinary panel or reported including the patientperspective or a plan for future updating Three508586 re-ported on important patient subgroups All four had well-described and reproducible methods sections Three205086

reported that potential conflicts of interest were examinedOnly two5086 rated the quality of the included evidence andthese same two were also the only two that rated the strengthof the evidence Only three205086 reported on the source ofany funding but all three reported nonindustry funding Onewas deemed to have a low risk of bias50 one was deemed tohave a low to moderate risk of bias86 and two were deemedto have a moderate risk of bias2085 None of the studieswere found to have any methodologic flaws that wouldbring doubt to their findings

Sepulveda et al

7 Expert Consensus Opinion Metastatic or recurrentcolorectal carcinoma tissues are the preferred specimens fortreatment predictive biomarker testing and should be used ifsuch specimens are available and adequate In their absenceprimary tumor tissue is an acceptable alternative and shouldbe used

In clinical practice one or more specimens of CRC from anindividual patient may become available for molecular testingduring the course of the disease These specimens may includeinitial diagnostic biopsy or surgical resection specimens of theprimary tumor and resection biopsy or cytologic specimensfrom metastatic and recurrent tumor Discordance betweenprimary andmetastatic lesionsmay be attributed to a number ofmechanisms including tumor heterogeneity already present inthe primary tumor tumor evolutionwhere novelmutations areacquired and in some cases the presence of separate pri-maries The systematic literature review for the CRC guidelinewas done to identify studies that compared the mutationalstatus of primary vs metastatic CRC

An earlier systematic literature search that was conductedto include studies testing concordance of KRAS BRAFPIK3CA and loss of PTEN expression in CRC20 reportedthe results of 21 studies with an overall concordance rate of93 (range 76-100) for KRAS 93 for BRAF status arange of 89 to 94 for PIK3CA and 68 for loss ofPTEN Table 12 shows the summary of two subsequentstudies where KRAS NRAS BRAF and PIK3CA mutationand PTEN expression were compared in paired primary vsmetastatic tumor lesions101102 Overall concordance ratesbetween primary and metastatic lesions were high withmorethan 90 concordance (Table 12)101102 In the study by Leeet al analysis of KRAS mutation in primary and recurrenttumors after radical resection showed 203 discordance103

This recommendation was supported by two retrospectivecohort studies101102 that were obtained in the systematicreview Both of these studies compared results within a

Table 12 Concordance Rates Between Primary and MetastaticLesions

Genes Tested (n) Concordance Rate

KRAS (117)101 910KRAS NRAS BRAF (84)102 988PIK3CA (117)101 940PIK3CA (84)102 928PTEN IHC (117)101 660

Summary of two randomized clinical trials where comparison of muta-tion in KRAS NRAS BRAF and PIK3CA was performed for paired primarytumor and metastatic lesions Immunohistochemistry for PTEN was done inCejas et al101 In the study by Cejas et al101 metastases were synchronousor metachronous DNA was extracted from formalin-fixed paraffin-embedded tissue and mutational analysis was performed with a polymerasechain reactionedirect sequencing assay KRAS mutations were detected in42 of metastatic lesions and 39 of primary tumors In the study byVakiani et al102 DNA was extracted from frozen tissue and the iPLEX(Agena Bioscience San Diego CA) assay was used to examine the followingmutations KRAS 12 13 22 61 117 and 146 NRAS 12 13 and 61 BRAF600 and PIK3CA 345 420 542 545 546 1043 and 1047

single cohort The study reported by Cejas et al101 reportedat least partial industry funding and the study reported byVakiani et al102 did not report the source of funding if anyThe study by Cejas et al101 was deemed to have a low tomoderate risk of bias and the study by Vakiani et al102 wasdeemed to be low Overall neither of these studies had anymethodologic flaws that would raise concerns about thereported findingsIn summary given that discordance of mutational status

between primary and metastatic or recurrent CRC lesionsmay occur in a number of cases metastatic or recurrentCRC tissues are the preferred specimens for treatment pre-dictive biomarker testing However if these specimens arenot available primary tumor tissue is an acceptable alter-native given the overall high rates of concordance for themutation status of EGFR pathway genes

8 Expert Consensus Opinion Formalin-fixed paraffin-embedded (FFPE) tissue is an acceptable specimen formolecular biomarker mutational testing in colorectal carci-noma Use of other specimens (eg cytology specimens) willrequire additional adequate validation as would anychanges in tissue-processing protocolsThe systematic review identified a number of studies

summarized in Table 13 where CRC KRAS mutationaltesting was performed using FFPE specimens as well asfresh or frozen specimens Recommendation 17 highlightsthe importance of review of stained sections of tumorselected for testing by a pathologist to verify the tumor cellcontent population of the sample and demarcate regions forpotential macrodissection or microdissection to enrich forcancer cells Biopsy and resection specimens are similarlyacceptable as long as sufficient tumor cells are present(Table 13) Cytology specimens may be adequate for testingbut will require proper validation The use of FFPE cellblocks allows for the evaluation of tumor cell content andviability104 Laboratories will need to establish the mini-mum tumor cell content for specimens based on the per-formance characteristics of their validated assay105126

Liquid biopsy tests use serum or plasma and may be usedfor monitoring tumor recurrence and emergence of treat-ment resistance The noninvasive nature of this approach(monitoring through blood testing) offers great potential forclinical use106 However at the present time the clinicalapplication of liquid biopsy assays awaits robust validationand further studies to determine their clinical utility

9 Strong Recommendation Laboratories must usevalidated colorectal carcinoma molecular biomarker testingmethods with sufficient performance characteristics for theintended clinical use Colorectal carcinoma molecularbiomarker testing validation should follow accepted stan-dards for clinical molecular diagnostics testsClinical validation assesses the molecular biomarker

testing method in light of clinical characteristics of thedisease or marker being tested to ensure the test is ldquofit forpurposerdquo Elements of clinical validation include analytical

sensitivity analytical specificity clinical sensitivity andclinical specificity Data for clinical validation can be ob-tained from studies performed by the laboratory studiesreported in peer-reviewed literature or other reliable sour-ces CLIA requires clinical laboratories to have a qualifiedlaboratory director who is responsible for ensuring that thelaboratory provides quality laboratory services for all as-pects of test performance107 Rigorous validation should beperformed to ensure all molecular marker testing methodssuch as those used for colorectal carcinoma are ready forimplementation in the clinical laboratory To reach that goaleach step of the testing process must be carefully evaluatedand documented Excellent and comprehensive documentshave been published on this topic and a detailed review isprovided under recommendation 10 Our systematic reviewof the available literature provided information regardingthe performance characteristics of molecular marker testingmethods of colorectal carcinoma in clinical use for RASmutational testing (Table 13) Most studies reported theperforming characteristic of assays that detected KRAS exon2 mutations as detailed in Table 13 Direct sequencing ofgenomic DNA even after polymerase chain reaction (PCR)amplification of the fragment of interest has low analyticalsensitivity requiring a mutant allele frequency of about 20for mutation detection A number of more sensitive assayshave been developed for RAS testing including those listedin Table 13

Sanger sequencing was used as the most commonbaseline assay for comparison against other moleculardetection methods for KRAS mutations Testing methodsvary widely including direct Sanger sequencing amplifi-cation refractory mutation system real-time PCRehigh-resolution melting (HRM) assays allele-specific PCRLuminex (Austin TX) bead microarray PCR restrictionfragment length polymorphism strip assays pyrosequenc-ing and more recently NGS Population or clinicalsensitivity of the testing methods for KRAS mutations asshown in Table 13 ranged between 36 and 59 Assaysensitivity ranged from 844 to 100 with Sangersequencing on the lower end of the range Analyticalsensitivity defined as the lowest detectable mutant allelefraction was between 05 and 20 across all testingmethods with most methods performing between 1 and5 mutant allele fraction Specificity was between 98and 100 for most assays with two studies demonstratinglower specificity Positive predictive value percentagesvaried between 66 and 100 with most studies reportingbetween 99 and 100 Negative predictive value per-centages were between 97 and 100 Minimal tumorpercentages reported varied widely between studiesConcordance between assays was between 93 and 100with some variability noted in two retrospective cohortstudies The available evidence from assays to detect KRASmutations supports the use of a number of alternativeassays as long as their performing characteristics adjustedfor sample type and percent tumor purity meet the clinical

sensitivity with acceptable specificity Recently NGS hasbeen used in a number of studies and in laboratory practicefor solid tumor mutational analysis108 NGS has shown tomeet the sensitivity of detection used in CRC clinical trials(detecting at least 5 mutant alleles) permitting simulta-neous testing of hundreds of mutations and is becomingwidely used Testing for mutations in multiple genes orgene loci with multiplex assays such as NGS and othermethods should be done on patients at the time of me-tastases to obtain comprehensive genomic information andidentify mutations beyond RASBRAF status that might beable to be targeted if conventional therapies becomeineffective

10 Strong Recommendation Performance of molecularbiomarker testing for colorectal carcinoma must be vali-dated in accordance with best laboratory practices

Proper validation of CRC biomarker testing is importantto ensure appropriate patient care If validation is inade-quate this can lead to erroneous results and improperdiagnosis prognosis andor therapeutic intervention Forexample with regard to RAS testing a false-positive resultwould lead to an improper withholding of therapy whereasa false-negative result would lead to distribution of anineffective therapy resulting in increased costs and unnec-essary side effects As molecular oncology testing growsmore complex with NGS thorough and proper validation ofpreanalytical (specimen type and processing) analytical(assay performance) and postanalytical (bioinformaticsannotation and reporting) steps is imperative109110

The design of a validation study somewhat depends onthe analyte (gene) mutations or molecular alterationsassessed and chosen platform and technology Howeverassay validation should be done using best laboratorypractices in accordance with CLIA (42 CFR4931253(b)(2) also known as Title 42 Chapter IV Sub-chapter G Part 493 Subpart Kx4931253)111 as applicable tothe assay type Laboratories should comply with CLIA andtheir individual accrediting agency (eg CAP New YorkState) to fulfill requirements for validation111112 Additionalresources for establishing clinical molecular testing areavailable to assist laboratories113 For the US Food andDrug Administration (FDA)eclearedapproved assays(without any modification) verification of test specifica-tions including accuracy precision reportable range andreference range only needs to be done114 For nonwaivednoneFDA-approved assays (laboratory-developed proced-ures or LDPs) validation must be performed Validationdesign must include the required elements of analyticalaccuracy (specificity and sensitivity) precision andanalytical sensitivity (limit of detection) and interferingsubstances and reportable range as applicable Clinicalsensitivity and specificity as well as positive and negativepredictive value should be considered additions

Additional considerations should include specimenprocessing (including microdissection or macrodissection

Table 13 Comparison of Test Performing Characteristic of Assays for KRAS Mutation Detection

Author Year No Comparison Testing Method Codons Tissue Site ProcedureSampleType

Ma et al 2009130 100 Sequencing HRM 12 13 Primary NR FFPEPinto et al 2011131 372 Consensusz Sequencing 12 13 NR NR FFPE

184 DxS182 HRM372 Snapshot

Tol et al 2010132 511 Sequencing DxS 12 13 Primary Resection FrozenBuxhofer-Ausch et al 2013133 60 Sequencing SA 12 13 Primary NR BiopsyChang et al 2010136 60 Sequencing MPCR PE 12 13 61 Primary NR FrozenChen et al 2009137 90 Sequencing SSCP 12 13 Primary NR FreshChow et al 2012138 204 Sequencing ASP 12 13 NR NR FFPESundstrom et al 2010142 100 DxS Pyro 12 13 61 Primary or met BiopsyFranklin et al 2010128 59 Sequencing HRM 12 13 Primary Resection FFPE

59 Sequencing ARMS 12 13 NRLaosinchai-Wolf et al 2011129 86 Sequencing BMA 12 13 Primary NR FFPECarotenuto et al 2010134 540 Sequencing DxS 12 13 Primary NR FFPE

540 Sequencing SangerCavallini et al 2010135 112 DxS SA 12 13 NR NR FFPE

112 DxS PCR-RFLPKristensen et al 2010139 61 COLD-PCR DxS 12 13 Primary Resection FFPE

61 PCR MCAKristensen et al 2012140 100 CADMA DxS 12 13 Primary Resection FFPE

100 DxS CADMALang et al 2011141 125 Sequencing ASP 12 13 Primary Resection FFPE

(table continues)

ARMS amplification refractory mutation system ASP allele specific (nonquantitative) BMA Luminex bead microarray CADMA competitive ampli-fication of differentially melting amplicons COLD-PCR coamplification at lower denaturation temperature-PCR DxS QIAGEN method FFPE formalinfixed paraffin embedded HRM high-resolution melting M missing MCA melting curve analysis met metastatic MPCR PE multiplex polymerase chainreaction (KRAS NRAS HRAS ) and primer extension NPV negative predictive value NR not reported PPV positive predictive value PCR-RFLPpolymerase chain reactionerestriction fragment length polymorphism PCS prospective cohort study Pyro pyrosequencing RCS retrospective cohort studySA KRAS-BRAF strip assay SSCP single-strand conformation polymorphismPopulation or clinical sensitivity of testing method () of cases positive for KRAS mutation testedyFour (95) of 42 samples negative for KRAS mutation by direct sequencing were positive for KRAS mutations by HRM analysiszTotal of 844 of consensus mutation resultxDetected one mutation in 23 Mut- allelesVariable concordance for different tumor percentage in the samplekThe sensitivity was increased by 5- to 100-fold for melting temperature decreasing mutations when using coamplification at lower

denaturation temperature-PCR (COLD-PCR) compared with standard PCR Mutations undetectable by the TheraScreen (QIAGEN Valencia CA)kit in clinical samples were detected by coamplification at lower denaturation temperature-PCR followed by HRM and verified by sequencingSequencing (PCR of fragment of interest followed by sequence analysis) described as direct sequencing

Sepulveda et al

histologic processing and fixation times) and reagentstability and storage Proper controls should be introducedand used to assess as many of the potential mutationsdetected by the assay and to verify the limit of detectionidentified in the validation With high-throughput (NGS)sequencing assessing all possible mutations throughcontrol material and specimens is impossible andcontinuing validation may need to occur If NGS is usedbioinformatics pipelines should be properly validatedusing multiple types of mutations (single-nucleotidevariants and insertionsdeletions) Finally reportingshould be carefully considered during the validation

process Resources to assist laboratories with solid tumormolecular testing have also been made available throughthe CLSI115

Preanalytical Variables

Histologic or preanalytical processing should be consideredand representative processes should be included in thevalidation set Specific specimen types should also beproperly validated Most tissue used in CRC biomarkertesting is derived from FFPE tissue Formalin fixation re-sults in fragmentation of DNA as a result of histone proteinfixation to the DNA Therefore most assays for FFPE tissue

PopulationSensitivityof TestingMethod

Sensitivityof Assay

AnalyticalSensitivity (Mutant AlleleFraction) Specificity PPV NPV

MinimalTumor

ConcordanceBetweenAssays Study

59 Increasedy (gt100) 5-10 98 NR NR 30 95 PCS364 844z 15-20 NR NR NR gt50 NR PCS431 96 1 NR NR NR NR427 98 3-10 NR NR NR NR433 99 5 NR NR NR NR394 965 1 997 995 972 3-90 9530 PCS470 100 1 100 NR NR At least 50 100 PCS340 100 NRx 100 100 100 NR 100 PCS360 100 NR 100 100 100 NR 100 PCS407 100 125-25 100 100 100 NR NR PCS390 91 125-25 125 NR NR NR NR NR PCS540 100 1 87 81 100 1-90 NR RCS43 100 5 71 66 100 1-90 93 RCS450 100 1 100 100 100 NR NR or M RCS386 958 1 100 100 973 lt30 vs gt70 Variable RCS

986 NR 100 100 991 NR NR RCS925-100 NR 100 NR NR 70 NR RCS925-100 NR 100 NR NR NR NR RCS

NR 93 01-5 100 NR NR NR k RCS97 5-10 100 NR NR NR RCS

444 98 050 98 NR NR NR 959 RCS99 NR 100 NR NR NR NR RCS

368 957x 1 NR NR NR gt50 NR RCS

are designed to amplify products less than 200 base pairsLength of formalin fixation and age of blocks may also befactors to consider in validation of FFPE tissues Othertissue sources should also be separately validated if offeredas clinical tests especially cytology-based specimensVarious cytology fixative preparations should be validatedas used by the laboratory If cell-free assays are consideredthese should be validated as a separate source Finallytesting should be limited to invasive carcinoma withexclusion of adenomatous tissue and benign backgroundtissue cellular components (eg normal mucosa muscularisinflammation) as much as possible

Analytical Variables

Careful specimen selection should be undertaken to cover asmany of the potential detected mutations and expectedspecimen types as possible to ensure analytical accuracy Agold-standard method (dideoxy sequencing or other vali-dated test method) andor interlaboratory comparison shouldbe used to verify accuracy of the assay For example theCAP Laboratory Accreditation Program COM40350 in-dicates that at least 20 specimens (including positive low-positive and negative specimens) should be included forqualitative and quantitative assays112 More specimens may

be required If it is a single-gene assay the design shouldinclude as many of the mutations covered by the assay aspossible If it is a real-timeebased allele-specific assay allmutations for which a primer probe reaction is built shouldbe analyzed as reasonably as possible If it is apyrosequencing-based assay similarly all of the possiblecommon mutations for which targeted therapies are indi-cated should be tested Multigene assays based on NGS orother technology [such as SNaPshot (ThermoFisher Scien-tific Waltham MA)] require an increased number ofspecimens to test as many of the hotspot regions as possiblein all genes included in the assay With such assays not allpossible mutations can be validated It is recommended thatan ongoing validation occur after initial validation withverification of novel mutations by either dideoxysequencing or real-time PCR depending on the laboratorycapability and limit of detection Depending on the tech-nology employed important parameters (eg variant allelefrequency cyclic threshold values allele coverage) shouldbe monitored for interrun and intrarun precision

CRC specimens can vary from large primary resectionblocks with plenty of tumor cells to small primary tumor ormetastatic CRC liver biopsy specimens to rectal specimensafter neoadjuvant therapy with minimal tumor percentage

Sepulveda et al

Many of these tests are ordered for metastatic disease forwhich only a small needle core biopsy specimen or cyto-logic sampling is available Presently tissue volume andaccessibility are decreasing while ancillary testing (IHCand molecular studies) is increasing The ability of an assayto be highly analytically sensitive is important if a labo-ratory is to test specimens with low tumor burden It isrecommended that an assay be able to identify a mutation ina specimen that has at minimum 20 tumor cells (mutantallele frequency of 10 assuming heterozygosity) WithNGS and highly sensitive PCR technologies mutationsshould be identifiable in specimens with as little as 10tumor (mutant allele frequency of 5 assuming heterozy-gosity and diploidy) Lower analytically sensitive assayssuch as dideoxy sequencing can be used but it is recom-mended that PCR enrichment strategies (eg coamplificationat lower denaturation temperature-PCR) be used to increasethe analytical sensitivity of the test and require less tumorpercentage A proper validation study should use cell lineDNA (preferably FFPE treated) or reference control materialmanufactured by good manufacturing processes to assesslimit of detection for as many mutations as possibleImportantly the limit of detection may differ for mutationsof varying types (small indels vs point mutations)

Postanalytical Variables

Postanalysis is as important to consider in validation aspreanalytical and analytical variables For single-geneassays the software used in analysis should be validatedwith verification of updates If NGS is used the bioinfor-matics pipeline should be thoroughly and rigorouslyvalidated include potential problematic mutations(eg large indels) and be verified or revalidated for newupgrades as applicable to the change Any analysis shouldbe performed on validation specimens as it would be forclinical specimens

Reporting format should also be considered and decidedduring validation Interpretation comments for inclusion inthe patient report to ensure that the reports are correctlyunderstood should be developed during the validation pro-cess112 Human Genome Organisation (HUGO)ebasednomenclature should be used for reports and a designatedNational Center for Biotechnology Information (NCBI)transcript number (NM_) should be used within thevalidation and report116 For multigene panels based onNGS reporting protocols and any used software should beincluded in the validation procedure Databases and anno-tation guidelines should be discussed and included in thevalidation as one prepares to report variants based on NGSdata In addition decisions should be made during thevalidation process as to whether normal tissue will be testedto assist in variant interpretation with NGS

In conclusion validation of assays used in CRC molec-ular testing is extremely important for accuracy of reportingand proper patient care There are several documents(eg CLIA CAP and CLSI)111-113115 available to assist in

proper validation which should be consulted to validateaccording to best laboratory practices

11 Strong Recommendation Laboratories must vali-date the performance of IHC testing for colorectal carci-noma molecular biomarkers (currently IHC testing forMLH1 MSH2 MSH6 and PMS2) in accordance with bestlaboratory practicesFour proteins (MLH1 MSH2 MSH6 and PMS2) are

currently considered important in the normal biochemistryof DNA MMR117-119 As detailed in recommendation 2baltered DNA mismatch repair proteins due to mutation orepigenetic silencing result in interference with normalMMR protein heterodimerization and loss of normal repairof mispaired bases and short insertionsdeletions resultingin MSI119120 overall categorized as dMMR Loss ofMMR function usually correlates with loss of proteinexpression such that immunohistochemical testing forMMR proteins is optimized to detect loss of MMR proteinexpression in tumor cell nuclei Each of these four proteinscan be detected in paraffin sections using commerciallyavailable primary and secondary antibodies standardizedantigen retrieval and 33rdquo-diaminobenzidine chromogendetection Development of anti-MMR protein antibodystaining protocols follows a standard approach that in-volves (1) demonstration of absent background noise withsecondary antibody alone and (2) empirical optimizationof the signal-to-noise ratio by testing different antibodyconcentrations antigen retrieval buffers and reactionconditions taking advantage of internal control cellsincluding lymphocytes stromal cells and other nonneo-plastic nucleiValidation of the final staining protocol is required prior to

implementation for clinical use Peer-reviewed literature-based guidelines for validation and revalidation of immuno-histochemical tests have been defined as 14 recommendationsand expert consensus opinions121 Concordance with internalor external known comparator tests is required to exceed 90Proficiency testing is a good approach to confirm interlabor-atory test reproducibility Test result concordance acrosslaboratories implies accuracy of participant laboratorydiagnosisOnce the protocol is defined and validated for a given

primary antibody clone and antigen retrieval conditions aknown positive external control (eg tonsil) is routinely runin parallel with each unknown This demonstrates that theMMR protein was detectable on that staining run and allowstrust in a loss of expression result in the unknown specimenEach of the four MMR proteins is expressed in nonneo-plastic tissue in most lymphocytes and overexpressed ingerminal centers such that most colon block sections willalso have positive internal control stainingOverall validated immunohistochemical detection of

MMR proteins is a trustworthy method for identification ofloss of expression of individual MMR proteins in paraffinsections of CRC In most CRCs with high-level

microsatellite instability (MSI-H) the loss of DNA MMRprotein expression in tumor cell nuclei by immunohisto-chemical detection is uniform throughout the tumor122123

Rare cases of MSI tumors have been reported to showheterogeneous staining124 Loss of MMR protein expressionusually correlates with MSI particularly for MSI-H tumorsand is indicative of dMMR If MSH2 or MLH1 shows lossof expression due to loss of function then their heterodimerpartners (MSH6 and PMS2 respectively) will also not beexpressed In contrast inactivation of MSH6 or PMS2 re-sults in loss of expression of the individual MMR proteinMSH6 or PMS2 respectively

Although loss of MMR protein immunoreactivity isgenerally detected in dMMR CRC normal immunoreac-tivity can be seen in up to 10 of dMMR cases125 there-fore MSI DNA testing may be performed either stepwise oras a concurrent test

12 Expert Consensus Opinion Laboratories mustprovide clinically appropriate turnaround times and optimalutilization of tissue specimens by using appropriate tech-niques (eg multiplexed assays) for clinically relevantmolecular and immunohistochemical biomarkers of CRC

Expediency in reporting of biomarker results for colo-rectal tumors is dictated primarily by two factors need forpatient management decisions and more generally patientanxiety Consequently results of such evaluations should beavailable within a timeframe for the involved clinician torelay this information to the patient This need is com-pounded by the patientrsquos need to receive a complete un-derstanding of his or her diagnosis and treatment plansgoing forward A reasonable benchmark is that nonacutebiomarker results be available to the treating physicianwithin 10 working days of receipt in the molecular di-agnostics laboratory This turnaround time has been rec-ommended in other guidelines for molecular tumortesting105126127 Ideally the transitional time between testordering tissue block selection block retrieval and ship-ment to the performing laboratory should be included in the10-day timeframe Consequently laboratories should makeevery effort to minimize delays in securing appropriate tis-sue blocks for testing Testing laboratories should makeevery effort to minimize processing time and return ofresults

The availability of tumor tissue for biomarker evaluationis generally not limiting in most cases of resected CRCOccasionally following neoadjuvant therapy the amount ofresidual tumor in resection specimens can be very small andfocal Similarly the amount of tumor tissue obtained bybiopsy or fine-needle aspiration procedures from primary ormetastatic foci can be very small and challenging to test forthe desired biomarkers In such circumstances availabletissue blocks should be sectioned judiciously reservingsufficient sections for testing by molecular methods orimmunohistochemical techniques as deemed appropriate tosecure as accurate and informative an evaluation as possible

Test turnaround times for RAS testing in instances ofadvanced stage tumors are dictated by the need to select andinitiate appropriate chemotherapy options Ideally such in-formation should be available either at the time of post-operative oncology evaluation where decisions regardingtherapeutic options are entertained or at the tumor boardswhere patient treatment options are discussed In some in-stitutions these discussions may occur in the weekfollowing surgery or biopsy and probably no later than inthe second week following tissue diagnosis and stagingHere too a timeframe of no more than 10 days would seeman appropriate benchmark for biomarker result availability

In exceptional circumstances even shorter test turn-around times may be called for Occasional patients havehistories sufficiently suggestive of Lynch syndrome thatprompt consideration and discussion regarding extent ofsurgery (ie complete colectomy or prophylactic hysterec-tomy in select affected patients) Efforts should be made insuch circumstances to obtain appropriate test results asrapidly as possible to allow for informed decision makingMMR immunohistochemistry can be performed and re-ported with a turnaround time of 48 hours or less and in theappropriate clinical context a result of preserved expressionof MMR proteins would argue against Lynch syndromeConversely any loss of MMR protein expression will needto be integrated with additional clinical information familyhistory and further testing such as BRAF mutation MLH1methylation testing and potential germline genetic testingFurthermore DNA MMR status performed by MMRimmunohistochemistry or by MSI DNA tests as a goodprognostic biomarker for CRC overall should be availablewithin the recommended 10 working day turnaround timefor test results

13 Expert Consensus Opinion Molecular and IHCbiomarker testing in colorectal carcinoma should be initiatedin a timely fashion based on the clinical scenario and inaccordance with institutionally accepted practices

Note Test ordering can occur on a case-by-case basis orby policies established by the medical staff

Molecular and IHC biomarker testing is increasinglybeing used in patient management Prognostic biomarkersare being used for early stage disease to guide decisions onthe use of adjuvant chemotherapy Such discussions requirethe availability of tests in a timely manner and delays ininitiation of therapy have been associated with worse out-comes127 Predictive biomarkers such as those for EGFRmonoclonal antibody therapy should be initiated in a timelyfashion to guide chemotherapy options and long-termtreatment planning Institutional policies and practices thatencourage the rapid initiation of appropriate molecular andIHC marker testing should be encouraged Such policiesmay include reflexive ordering of molecular and IHCmarkers as guided by the clinical scenario and incorporationof testing initiation by multiple members of the multidisci-plinary team as noted in recommendation 15

Sepulveda et al

14 Expert Consensus Opinion Laboratories shouldestablish policies to ensure efficient allocation and utiliza-tion of tissue for molecular testing particularly in smallspecimens

The number of molecular and immunohistochemical testsbecoming available that have a direct benefit to patient carewill continue to increase Most of these tests are performedon FFPE specimens the most common preservation tech-nique including pretreatment and posttreatment biopsiesand resections (Table 13) Tissues from patients with cancershould be processed according to established laboratoryprotocols which include quality controls of preservationmaterials tissue dissection time to fixation fixation timeand processing

Laboratory protocols need to include procedures forhandling small samples such as endoscopic or core biopsyspecimens and fine-needle aspirate samples of metastatic le-sions (eg from liver or lung) Limiting the number of tissuefragments per individual cassette is encouraged Establishedprotocols may allow upfront ordering of required tissue sec-tions (eg extra unstained slides) which limit tissue wastingand improve turnaround time of final results Immunohisto-chemistry studies if needed to diagnose metastatic CRCshould be limited in scope and standardized to preserve tissues

It is imperative to identify suspected metastatic CRCspecimens at specimen accessioning to limit unneededancillary tests such as liver biopsy special stains Recog-nition of previous CRC diagnoses from the patient clinicalhistory should limit the need for immunohistochemistryprofiles in many cases Established laboratory procedures toidentify patients undergoing cancer biopsy or fine-needleaspiration specifically for predictive molecular biomarkerassessments need to be in place

Laboratories must maintain appropriate cataloguing andstorage of tissue specimens and diagnostic slides to allowfor retrospective timely testing of cancer samples

This recommendation is supported by 15 studies128-142

comprising eight prospective cohort studies130-133136-138142

and seven retrospective cohort studies128129134135139-141

For the eight prospective cohort studies130-133136-138142

all reported balance between the treatment and assessmentgroups as all but one132 used a single cohort designallowing for within-group comparisons Only this singlestudy reported by Tol et al132 would have required makingadjustments for imbalances between the treatment andassessment groups but none were needed Fivestudies130133136-138 reported nonindustry funding one132

reported at least partial industry funding one142 reportedindustry funding and one131 did not disclose the source offunding if any Seven130131133136-138142 were deemed tohave a low risk of bias and one132 was deemed to have alow to moderate risk of bias

For the seven retrospective cohortstudies128129134135139-141 all used a single cohort designallowing for within-group comparisons Four reportednonindustry funding134135139140 one reported industry

funding129 and two did not disclose the source of fundingif any128141 Six were deemed to have a low risk ofbias128134135139-141 and one was deemed to have a mod-erate risk of bias129

All of the evidence that supported this recommendationwas assessed and none had methodologic flaws that wouldraise concerns about their findings

15 Expert Consensus Opinion Members of the pa-tientrsquos medical team including pathologists may initiatecolorectal carcinoma molecular biomarker test orders inaccordance with institutionally accepted practicesFor patients with CRC timely diagnosis or therapeutic

initiation is critical and molecular testing that is to beconsidered should be ordered as efficiently as possible inaccordance with institutional practices and guidelines MSItesting is often ordered at the time of diagnosis to identifypatients with Lynch syndrome direct adjuvant chemo-therapy or determine prognosis Many institutions employalgorithms to ensure that all colorectal cancers are evaluatedfor MMR deficiency and these are often initiated by pa-thologists when the diagnosis occurs after joint generalprocess approval by pathologists oncologists and othermembers of the patient medical team Molecular testing thatis performed to direct targeted therapy (eg RAS ) may beordered at a later date than the primary diagnosis at meta-static presentation for example and so institutions maydiffer as to whether one should order such testing upfront onthe primary diagnostic biopsy or resection specimen or waituntil metastatic disease arises requiring targeted therapyOften oncologists order predictive molecular assays sincethey are used to direct therapy but this should not neces-sarily be limited to oncologists as pathologists serve asimportant stewards of the tissue and make the tumor diag-nosis There are also issues to consider including logisticalissues cost-effectiveness patient access to molecular testingin rural or underserved areas and even heterogeneity con-siderations between primary and metastatic tumor Sinceeach institution differs in patient population facilitiesdepartmental organization regulatory and reimbursementclimates and practitioner preference whether to submittesting at initial diagnosis of a primary lesion or when ametastatic lesion arises should be discussed collaborativelybetween oncologists pathologists and medical executive orhospital committees as applicableldquoReflexrdquo testing a testing policy that does not require a

separate clinician order for each case is appropriate ifagreed upon by the CRC care team as an institutionallyapproved standing order and may help to ensure expeditedand consistent routing of specimens for molecular testingHowever some patients may not be candidates for targetedtherapy for clinical reasons and good communication be-tween the clinical care team and the testing laboratory isneeded to ensure testing is performed for patients whosemanagement will be affected by the test result Specificallytesting is not necessary for patients with stage IV disease

who are being considered for palliative or hospice careonly Similarly in settings in which reflex testing is thepractice a mechanism should be provided for the clinicalcare team to communicate to the pathologist examining asmall biopsy or cytology sample when a more suitablediagnostic specimen (eg a resection) is expected to be ob-tained and the molecular testing should be deferred to thesubsequent more generous sample All reflex testing shouldbe approved institutionally by the hospital or institutionrsquosmedical executive committee as local policies dictate

16 Expert Consensus Opinion Laboratories thatrequire send out of tests for treatment predictive biomarkersshould process and send colorectal carcinoma specimens toreference molecular laboratories in a timely manner

Note It is suggested that a benchmark of 90 of speci-mens should be sent out within 3 working days

It is critical to provide the results of molecular tests in atimely fashion to start the most appropriate cancer treatmentoption for each patient Delays in initiation of therapy havebeen associated with worse outcomes127 To date labora-tories have had limited guidance on the recommendedtiming or turnaround time of molecular test results andstudies addressing the impact of specific turnaround timeshave not been conducted Therefore the panel reached anexpert consensus opinion based on each panel memberrsquospractical experience in the laboratory and clinical setting

For laboratories that do not perform molecular testingandor biomarker immunohistochemistry for CRC therapyselection the consensus opinion was that send out ofspecimens should occur within 3 working days startingfrom the day the test order was received in the laboratoryprovided the specimens (eg biopsy or resection specimens)are received at the same time of the test order or specimensare already in the laboratory (eg archived paraffin blocks)The underlying rationale stems from the usual workflow fortissue processing In practice the longest process would bethe processing of large surgical specimens such as colec-tomies A possible approach is to obtain a designated mo-lecular tissue block at the time of specimen grossing andmolecular protocols for obtaining tissue sections may beused to have the necessary sections for test send-out in atimely fashion by the third working day for most casesAnother scenario may be the retrieval of archived tissueparaffin blocks that may be stored outside of the laboratorylocation In this case a protocol for block retrieval formolecular testing may be operationalized to streamline theprocess and reach the desired turnaround time for send-outThis turnaround time of 3 working days was also recom-mended for RAS testing of colorectal carcinoma in theguidance document from the Association of Clinical Pa-thologists Molecular Pathology and Diagnostics Group inthe United Kingdom126

Laboratories should develop written policies as part oftheir quality assurance program to monitor turnaround timesfor all cancer therapeutic and prognostic biomarkers

17 Expert Consensus Opinion Pathologists must eval-uate candidate specimens for biomarker testing to ensurespecimen adequacy taking into account tissue qualityquantity and malignant tumor cell fraction Specimen ade-quacy findings should be documented in the patient report

It is critical that pathologists selecting blocks forbiomarker testing understand the specimen requirements ofthe method being employed in terms of total tissue amount(a reflection of the total amount of DNA required for theassays) and the fraction of malignant tumor cells in thespecimen focus to be evaluated The total amount of tissueselected for evaluation is significant in two respects First theamount of tissue sampled should be of sufficient quantity toproduce a result that is reliably representative of the entiretumor While recent evidence indicates that some genescontinue to evolve during tumor progression leading tosubstantial tumor genetic heterogeneity those driver muta-tions of importance to CRC are usually but not always ho-mogeneous throughout the tumor The amount of tumornecessary however for a particular analytical method canvary and demands knowledge and due attention to the indi-cated tissue requirements for the specific assay employedThe minimal required proportion of tumor DNA in a samplefrom cancer is dictated by the analytical sensitivity of theparticular validated assay As shown in Table 13 the amountof tumor used in the analyses of KRAS mutations in severalstudies comparing the test-performing characteristics ofvarious assays varied widely ranging from 1 to 90

The proportion of malignant tumor cells (as opposed totumor-associated nonmalignant cells eg stromal fibro-blasts endothelial cells infiltrating inflammatory cells)should be evaluated as accurately as possible and docu-mented This evaluation is most readily performed by esti-mating the proportion of malignant cell nuclei tononmalignant cell nuclei within the focus selected forevaluation143 Understanding that the number of mutatedalleles for a particular gene may represent as few as half ofthe alleles in diploid tumor cells a tumor cell focus with anominal proportion of 50 tumor cells would have a mutantallele fraction of 25 a value approaching the analyticalsensitivity of some molecular assays So while variety ofmolecular methods can be used to evaluate tissue speci-mens it is critical that these be carefully matched to theirspecific tissue and tumor cell proportion requirementsWhen adhered to all these of these methods can produceaccurate and reliable results

Pathologists evaluating tissue section for biomarkerevaluation should also be aware that necrosis and tissuedegeneration can lead to erroneous results and focidemonstrating significant necrosis should be avoided formolecular testing Any amount of necrosis in the sampleselected for biomarker testing should be estimated anddocumented

18 Expert Consensus Opinion Laboratories should usecolorectal carcinoma molecular biomarker testing methods

Sepulveda et al

that are able to detect mutations in specimens with at least 5mutant allele frequency taking into account the analyticalsensitivity of the assay (limit of detection or LOD) and tumorenrichment (eg microdissection)

Note It is recommended that the operational minimalneoplastic carcinoma cell content tested should be set atleast two times the assayrsquos LOD

Since the accuracy and results of testing for molecularmarkers are dependent on both tumor cell content and theassay-specific sensitivity in the identification of a mutantallele against a background of wild-typenonmutated allelesit is suggested that laboratories should establish minimumacceptable tumor cell content as a component of theirspecimen requirements It is recommended that a patholo-gist reviews all cases for tumor cell content and quality Dueto the stochastic nature of mutant allele identification at thelower LOD it is recommended that the minimal tumor cellcontent be at least two times the lower LOD of a validatedmolecular method or assay This LOD was also recom-mended for RAS testing of colorectal carcinoma in theguidance document from the United Kingdom126 Hence ifa particular assay has a lower limit of mutant allele detectionof 5 then the minimum tumor cell content in samplesanalyzed by this assay should be at least 10 to reliablydetect heterozygous mutations in those neoplasms Due tointratumoral heterogeneity subclones and the nature oftissue sampling clinical trials have used 5 as the lowerLOD and for clinical purposes it is recommended that thelower LOD for a mutant allele be at least 512 Thereforethe utilization of methods such as PCR HRM single-strandconformation polymorphism pyrosequencing or commer-cially available kits that achieve this level of sensitivity isrecommended130137138142 (Table 13)

This recommendation is supported by four prospectivecohort studies130137138142 and two retrospective cohortstudies102144 The four prospective cohort studies all studieda single cohort allowing for within-group comparisons Forthis reason all were balanced between comparison groupsand no adjustments were needed to account for baselinedifferences All four reported nonindustry funding and allwere deemed to have a low risk of bias

The two retrospective cohort studies102144 also usedsingle cohorts allowing for within-group comparisons onlyOne102 did not report the source of funding while theother144 reported nonindustry funding Both were deemed tohave a low risk of bias

None of the studies had methodologic flaws that wouldraise concerns about their findings

19 Expert Consensus Opinion Colorectal carcinomamolecular biomarker results should be made available aspromptly as feasible to inform therapeutic decision makingboth prognostic and predictive

Note It is suggested that a benchmark of 90 of reportsbe available within 10 working days from date of receipt inthe molecular diagnostics laboratory

Combined chemotherapy including anti-EGFR therapyin patients with CRC in the absence of mutations in theEGFR signaling pathway is associated with significantsurvival advantage No significant therapeutic benefit isderived from anti-EGFR therapy in the presence of muta-tions in KRAS and NRAS44 The presence of deficient MMRin stage II CRC indicates a good prognosis and identifiespatients for whom adjuvant 5-fluorouracil mono-basedtherapies have no significant benefit145146 The presenceof deficient MMR or BRAF pV600E mutation in proficientMMR CRCs has important prognostic significance54

In the absence of published data establishing an evidence-based recommendation it is our expert consensus opinionthat the above results regardless of testing methods beavailable from test ordering in the initial diagnostic pathologylaboratory to the clinical team within 2 weeks (10 workingdays) The 10 working days does not include the time beforethe tissue specimen is available for testing (ie from diag-nostic procedure to receipt in laboratory) or time to retrievetissue samples from an outside laboratory Laboratories un-able to maintain this standard either through in-house testingor use of a reference laboratory need to implement measuresto improve test result turnaround time A turnaround time of7 working days was recommended for RAS testing of colo-rectal carcinoma in the guidance document from the Asso-ciation of Clinical Pathologists Molecular Pathology andDiagnostics Group in the United Kingdom126

This recommendation is supported by evidence from onerandomized controlled trial reported by Douillard et al44

This report used prospective patient data collected withinthe PRIME trial While it did not report details on therandomization blinding statistical power calculation sam-ple size or length of follow-up it did report on baselinecharacteristics and was otherwise well reported Fundingwas reported to be partially from industry sources Overallthis trial was found to have a low to moderate risk of biasand did not have methodologic flaws that would raiseconcerns about its findingsEach laboratory should develop a quality assurance pro-

gram to monitor turnaround times for all cancer therapeuticand prognostic biomarkers

20 Expert Consensus Opinion Colorectal carcinomamolecular biomarker testing reports should include a resultsand interpretation section readily understandable by oncol-ogists and pathologists Appropriate Human Genome Vari-ation Society (HGVS) and HUGO nomenclature must beused in conjunction with any historical genetic designationsReporting of molecular results is becoming more com-

plex as new information and clinical utility are discoveredfor somatic variants Single-gene assays are still beingwidely used but multiplexing has allowed for multiplepossible results With the introduction of NGS into theclinical setting multiple somatic mutations with clinicalsignificance may be identified However panel assays byNGS can also reveal variants with unknown clinical

significance As pathogenic genes and somatic mutationshave been discovered over the past 30 years there has beendivergent nomenclature employed making clinical report-ing and clinical analysis difficult Presently and in thefuture as national databases are constructed annotatingclinical somatic variants it is imperative that standardizednomenclature be employed to identify the clinical signifi-cance of rare variants

Clinicians want a report that is easily readable and un-derstandable but that gives pertinent clinical informationconcisely accurately and thoroughly Reported variantsshould be identified using both DNA and protein nomen-clature Citing codon positivity only is not encouraged (egpositive for a KRAS codon 12 mutation) The specific mu-tation should be explained using standardized nomenclaturepreferably HUGO gene nomenclature112147 Historical des-ignations (eg historical HER-2neu for HUGO ERBB2)should also be included as appropriate in the report to avoidconfusion among oncologists Importantly the messengerRNA transcript number (NM_) from the NCBI used todesignate the specific codon numbering should be named inthe report since numbering can differ between the differentalternative transcript designations for the same gene If usingNGS variants should at least be classified as pathogeniclikely pathogenic variant of unknown significance likelybenign or benign but classification of somatic mutations isstill awaiting specifically approved guidelines148 How-ever a numerical classification scheme for somatic variantshas been proposed taking into consideration actionabilityof the variant in the patientrsquos tumor type vs other tumortypes predicted pathogenicity (using programs such asSIFT and PolyPhen 2) in the patientrsquos tumor type vs othertumor types variant recurrence in a certain cancer type orunknown significance149 Such a classification schememay be better suited to somatic variants considering theindications for which most of these assays are beingordered

Reports should contain the analytical result the methodused and information about the genes and loci tested orincluded in the assay the assay limit of detection and anydisclaimers (eg ASR) that are required to meet regulationsWhen reasonable and applicable an interpretive commentshould be given to ensure that results are correctly under-stood112 Such an interpretive comment may include infor-mation regarding therapeutic implications prognosticimplications andor pathogenic significance of the mutationand when appropriate or desired potential applicable clin-ical trials

In summary molecular reports should be easily under-standable by clinical oncologists and use standardizednomenclature outlined by HGVSHUGO All reports shouldcontain the elements of result interpretation variant clas-sification and information as applicable limit of detectionof the assay and methods to assist the oncologist in under-standing the test result and limitations as they consider theresult in a clinical context

21 Strong Recommendation Laboratories must incor-porate colorectal carcinoma molecular biomarker testingmethods into their overall laboratory quality improvementprogram establishing appropriate quality improvementmonitors as needed to ensure consistent performance in allsteps of the testing and reporting process In particularlaboratories performing colorectal carcinoma molecularbiomarker testing must participate in formal proficiencytesting programs if available or an alternative proficiencyassurance activity

Proficiency testing (PT) is an important component ofquality assurance for laboratory tests in general and appliesto the molecular tests discussed in the current CRC mo-lecular testing guidelines These include mutational as wellas immunohistochemical testing Participation in PT allowsthe assessment and comparison of test performance amongdifferent clinical laboratories and technologies and allowsverification of accuracy and reliability of laboratorytests150

From a regulatory standpoint PT in the United States is arequirement for accreditation by the Centers for Medicare ampMedicaid Services Participation in PT may be done throughCAP PT programs or through other providers acceptedby CLIA151 Other countriesdnamely the UnitedKingdomdfollow similar guidelines recommending thatlaboratories providing RAS testing of CRC should demon-strate successful participation in a relevant external qualityassurance scheme and be appropriately accredited126

Formal external proficiency testing programs for analytesother than KRAS MSI MMR and BRAF may not beavailable at the time of this publication Alternative profi-ciency testing activities should be used Appropriate alter-native performance assessment procedures may includesplit sample analysis with other laboratories or if that is notavailable assessment of split samples with an establishedin-house method and previously assayed material which arerun and interpreted by laboratory personnel who do nothave access to the prior results151 If exchanging specimenswith other laboratories is the laboratory proficiencyapproach this should be done with one or more other lab-oratories at least twice per year105 Methods-based profi-ciency testing (MBPT) refers to a testing approach that isbased on method rather than based on each individual an-alyte tested MBPT is well established for several pathologysubspecialty areas and the concept of MBPT complies withfederal laboratory regulations151

Discussion on Emerging Biomarkers

Numerous studies have reported potential molecular bio-markers for CRC prognosis while fewer studies evaluatedmarkers that could be predictive of response to specifictreatments Many published studies are limited due to earlyexploratory and retrospective analyses and those bio-markers while of potential interest have not made it to

Sepulveda et al

clinical practice Our systematic review identified severalCRC molecular biomarkers that showed either prognostic ortreatment predictive characteristics in single studies(Supplemental Table 15) Most of the molecular biomarkersreported in the studies listed in the Supplemental Table 15were tested for expression by immunohistochemistryImmunohistochemistry is notable for its widespread avail-ability in pathology laboratories but has limited quantitativecapabilities due to difficult standardization of quantitative orsemiquantitative scoring and is fraught by significantinterobserver variability A problem of quantitative assayssuch as gene expression microRNA expression andmethylation levels tested in solid tumors results from theintrinsic mixed nature of the tissue with significant vari-ability of tumor and nontumor tissue content Anotherlimitation of molecular biomarker discovery approaches thatrely on expression levels is that these biomarkers have notbeen evaluated in the context of complex molecular regu-lation of individual cancer subtypes Their fruitful use in theclinic may require further studies that take into accountcomputational predictions of biological behavior and vali-dation in prospective cohorts

A great deal of interest has been raised recently fornoninvasive prognostic andor therapy-predictive molecularbiomarkers such as those tested in circulating tumor cells orcirculating nucleic acids either as free nucleic acid in serumor associated with extracellular vesicles or exosomes Thishas been referred to as ldquoliquid biopsyrdquo152 Liquid biopsiesmay be particularly useful in the management of patients withCRC to identify recurrence RAS mutation testing for emer-gence of treatment resistance associated with anti-EGFRtherapy and potential early cancer detection in defined sub-populations such as those at high risk of CRC Overallmolecular biomarkers for colorectal cancer tested in liquidbiopsy samples are promising but await further validation

Emerging data indicate that MMR status may have pre-dictive value in some settings specifically in patients withadvanced disease being considered for anti-PD-1PD-L1therapy6869

Conclusions

Evidence supports mutational testing of specific genes in theEGFR signaling pathway since they provide clinicallyactionable information for targeted therapy of CRC withanti-EGFR monoclonal antibodies Mutations in some ofthe biomarkers have clear prognostic value (BRAF MMR)and at least two (KRAS and NRAS ) have relatively strongevidence as negative predictors of benefit to anti-EGFRtherapies and should be used to guide the use of theseagents BRAF mutations are consistently associated withpoor outcomes in patients with metastatic CRC includingthose who relapse after adjuvant therapy Patients withlocalized colon cancer and dMMR have improved out-comes Emerging data suggest that MMR status has

predictive value in some settings specifically in patientswith advanced disease being considered for anti-PD-1PD-L1 therapyLaboratory approaches to operationalize molecular

testing for predictive and prognostic molecular biomarkersinvolve selection of assays type of specimens to be testedtiming of ordering of tests and turnaround time for testingresults A number of alternative technical approaches caneffectively be used as long as test specificity and sensitivitymeet the clinical needs While earlier testing approacheswere focused on one or a few testing targets (eg BRAFpV600 mutations) currently new approaches are usinggene panels such as targeted NGS cancer panels which canrange from a few to hundreds of genes and amplicons withknown mutational hotspots in cancerThese guidelines will be subjected to regular updates

such that new advances in the field can be captured andintegrated in the guidelines in a timely mannerThis guideline was developed through collaboration

between the American Society for Clinical PathologyCollege of American Pathologists Association for Molec-ular Pathology and American Society of Clinical Oncologyand has been jointly published by invitation and consent inthe American Journal of Clinical Pathology Archives ofPathology amp Laboratory Medicine Journal of MolecularDiagnostics and Journal of Clinical Oncology It has beenedited in accordance with standards established at theAmerican Journal of Clinical PathologyThe authors gratefully acknowledge the following expert

and advisory panel members Kevin Halling MD CharlesBlanke MD Pr Jean-Franccedilois Fleacutejou Heather HampelMS CGC J Randolph Hecht MD Kazunori KanehiraMD Faye Kastrinos MD MPH Carla B MacLeod MDPeter J Odwyer MD Weijing Sun MD Josep TaberneroMD PhD Laura H Tang MD PhD Mary Kay Wash-ington MD PhD Shuji Ogino MD PhD MS LorenJoseph MD Kim Ryan and Pamela McAllister PhD fortheir review of the key questions recommendations anddraft manuscript as well as Kaitlin Einhaus Lisa AFatheree SCT(ASCP) Jeff Jacobs MA Sandra LarsenMBA MT(ASCP) Melvin Limson PhD John Olsen MDMrudula Pullambhatla MS Shiwen Song MD MeganWick MT(ASCP) and Mary Steele Williams MNAMT(ASCP)SM CAE for their support throughout theguideline development process

Supplemental Data

Supplemental material for this article can be found athttpdxdoiorg101016jjmoldx201611001

References

1 Febbo PG Ladanyi M Aldape KD et al NCCN task force reportevaluating the clinical utility of tumor markers in oncology J NatlCompr Cancer Netw 2011 9 Suppl 5S1eS33

2 Grothey A EGFR antibodies in colorectal cancer where do theybelong J Clin Oncol 2010 284668e4670 doi101200JCO2010293407

3 De Roock W Claes B Bernasconi D et al Effects of KRAS BRAFNRAS and PIK3CA mutations on the efficacy of cetuximab pluschemotherapy in chemotherapy-refractory metastatic colorectal can-cer a retrospective consortium analysis Lancet Oncol 2010 11753e762 doi101016S1470-2045(10)70130-3

4 De Roock W De Vriendt V Normanno N et al KRAS BRAFPIK3CA and PTEN mutations implications for targeted therapies inmetastatic colorectal cancer Lancet Oncol 2011 12594e603 doi101016S1470-2045(10)70209-6

5 Rubenstein JH Enns R Heidelbaugh J et al American Gastroen-terological Association Institute guideline on the diagnosis andmanagement of Lynch syndrome Gastroenterology 2015 149777e782 doi101053jgastro201507036

6 Des Guetz G Schischmanoff O Nicolas P et al Does microsatelliteinstability predict the efficacy of adjuvant chemotherapy in colorectalcancer a systematic review with meta-analysis Eur J Cancer 2009451890e1896 doi101016jejca200904018

7 Guastadisegni C Colafranceschi M Ottini L et al Microsatelliteinstability as a marker of prognosis and response to therapy a meta-analysis of colorectal cancer survival data Eur J Cancer 2010 462788e2798 doi101016jejca201005009

8 Gavin PG Colangelo LH Fumagalli D et al Mutation profiling andmicrosatellite instability in stage II and III colon cancer an assess-ment of their prognostic and oxaliplatin predictive value Clin CancerRes 2012 186531e6541 doi1011581078-0432CCR-12-0605

9 Institute of Medicine Clinical Practice Guidelines We Can TrustWashington DC National Academies Press 2011

10 National Health and Medical Research Council A guide to thedevelopment implementation and evaluation of clinicl practiceguidelines 1999 Available at httpswwwnhmrcgovau_files_nhmrcpublicationsattachmentscp30pdf Accessed March 23 2016

11 Guyatt GH Oxman AD Vist GE GRADE Working Group et alGRADE an emerging consensus on rating quality of evidence andstrength of recommendations BMJ 2008 336924e926 doi101136bmj39489470347AD

12 Sorich MJ Wiese MD Rowland A et al Extended RAS mutationsand anti-EGFR monoclonal antibody survival benefit in metastaticcolorectal cancer a meta-analysis of randomized controlled trialsAnn Oncol 2015 2613e21 doi101093annoncmdu378

13 Ibrahim EM Abouelkhair KM Clinical outcome of panitumumab formetastatic colorectal cancer with wild-type KRAS status a meta-analysis of randomized clinical trials Med Oncol 2011 28 Suppl 1S310eS317 doi101007s12032-010-9760-4

14 Modest DP Brodowicz T Stintzing S et al Impact of the specificmutation in KRAS codon 12 mutated tumors on treatment efficacy inpatients with metastatic colorectal cancer receiving cetuximab-basedfirst-line therapy a pooled analysis of three trials Oncology 2012 83241e247 doi101159000339534

15 Petrelli F Borgonovo K Cabiddu M et al Cetuximab and pan-itumumab in KRAS wild-type colorectal cancer a meta-analysisInt J Colorectal Dis 2011 26823e833 doi101007s00384-011-1149-0

16 Zhou SW Huang YY Wei Y et al No survival benefit from addingcetuximab or panitumumab to oxaliplatin-based chemotherapy in thefirst-line treatment of metastatic colorectal cancer in KRAS wild typepatients a meta-analysis PLoS One 2012 7e50925 doi101371journalpone0050925

17 Karapetis CS Khambata-Ford S Jonker DJ et al K-ras mutationsand benefit from cetuximab in advanced colorectal cancer N Engl JMed 2008 3591757e1765 doi101056NEJMoa0804385

18 Adelstein BA Dobbins TA Harris CA et al A systematic reviewand meta-analysis of KRAS status as the determinant of response toanti-EGFR antibodies and the impact of partner chemotherapy in

metastatic colorectal cancer Eur J Cancer 2011 471343e1354 doi101016jejca201103031

19 Allegra CJ Jessup JM Somerfield MR et al American Society ofClinical Oncology provisional clinical opinion testing for KRASgene mutations in patients with metastatic colorectal carcinoma topredict response to anti-epidermal growth factor receptor monoclonalantibody therapy J Clin Oncol 2009 272091e2096 doi101200JCO2009219170

20 Baas JM Krens LL Guchelaar HJ et al Concordance of predictivemarkers for EGFR inhibitors in primary tumors and metastases incolorectal cancer a review Oncologist 2011 161239e1249 doi101634theoncologist2011-0024

21 Chen J Ye Y Sun H et al Association between KRAS codon 13mutations and clinical response to anti-EGFR treatment in patientswith metastatic colorectal cancer results from a meta-analysis Can-cer Chemother Pharmacol 2013 71265e272 doi101007s00280-012-2005-9

22 Dahabreh IJ Terasawa T Castaldi PJ et al Systematic review anti-epidermal growth factor receptor treatment effect modification byKRAS mutations in advanced colorectal cancer Ann Intern Med2011 15437e49 doi1073260003-4819-154-1-201101040-00006

23 De Roock W Jonker DJ Di Nicolantonio F et al Association ofKRAS pG13D mutation with outcome in patients with chemo-therapy-refractory metastatic colorectal cancer treated with cetux-imab JAMA 2010 3041812e1820 doi101001jama20101535

24 Health Quality Ontario KRAS testing for Anti-EGFR therapy inadvanced colorectal cancer an evidence-based and economic anal-ysis Ont Health Technol Assess Ser 2010 101e49

25 Hoyle M Crathorne L Peters J et al The clinical effectiveness andcost-effectiveness of cetuximab (mono- or combination chemo-therapy) bevacizumab (combination with non-oxaliplatin chemo-therapy) and panitumumab (monotherapy) for the treatment ofmetastatic colorectal cancer after first-line chemotherapy (review oftechnology appraisal No150 and part review of technology appraisalNo 118) a systematic review and economic model Health TechnolAssess 2013 171e237

26 Ibrahim EM Zekri JM Bin Sadiq BM Cetuximab-based therapy formetastatic colorectal cancer a meta-analysis of the effect of K-rasmutations Int J Colorectal Dis 2010 25713e721 doi101007s00384-010-0927-4

27 Jiang Z Li C Li F et al EGFR gene copy number as a prognosticmarker in colorectal cancer patients treated with cetuximab or pan-itumumab a systematic review and meta analysis PLoS One 2013 8e56205 doi101371journalpone0056205

28 Ku GY Haaland BA de Lima Lopes G Jr Cetuximab in the first-linetreatment of K-ras wild-type metastatic colorectal cancer the choiceand schedule of fluoropyrimidine matters Cancer Chemother Phar-macol 2012 70231e238 doi101007s00280-012-1898-7

29 Lin AY Buckley NS Lu AT et al Effect of KRAS mutational statusin advanced colorectal cancer on the outcomes of anti-epidermalgrowth factor receptor monoclonal antibody therapy a systematicreview and meta-analysis Clin Colorectal Cancer 2011 1063e69doi103816CCC2011n009

30 Linardou H Dahabreh IJ Kanaloupiti D et al Assessment of somatick-RAS mutations as a mechanism associated with resistance toEGFR-targeted agents a systematic review and meta-analysis ofstudies in advanced non-small-cell lung cancer and metastatic colo-rectal cancer Lancet Oncol 2008 9962e972 doi101016S1470-2045(08)70206-7

31 Loupakis F Cremolini C Salvatore L et al Clinical impact of anti-epidermal growth factor receptor monoclonal antibodies in first-linetreatment of metastatic colorectal cancer meta-analytical estimationand implications for therapeutic strategies Cancer 2012 1181523e1532 doi101002cncr26460

32 Mao C Huang YF Yang ZY et al KRAS pG13Dmutation and codon12 mutations are not created equal in predicting clinical outcomes of

Sepulveda et al

cetuximab in metastatic colorectal cancer a systematic review andmeta-analysis Cancer 2013 119714e721 doi101002cncr27804

33 Mao C Yang ZY Hu XF et al PIK3CA exon 20 mutations as apotential biomarker for resistance to anti-EGFR monoclonal anti-bodies in KRAS wild-type metastatic colorectal cancer a systematicreview and meta-analysis Ann Oncol 2012 231518e1525 doi101093annoncmdr464

34 Petrelli F Barni S Resectability and outcome with anti-EGFR agentsin patients with KRAS wild-type colorectal liver-limited metastases ameta-analysis Int J Colorectal Dis 2012 27997e1004 doi101007s00384-012-1438-2

35 Petrelli F Coinu A Cabiddu M et al KRAS as prognostic biomarkerin metastatic colorectal cancer patients treated with bevacizumab apooled analysis of 12 published trials Med Oncol 2013 30650 doi101007s12032-013-0650-4

36 Qiu LX Mao C Zhang J et al Predictive and prognostic value ofKRAS mutations in metastatic colorectal cancer patients treated withcetuximab a meta-analysis of 22 studies Eur J Cancer 2010 462781e2787 doi101016jejca201005022

37 Ren J Li G Ge J et al Is K-ras gene mutation a prognostic factor forcolorectal cancer a systematic review and meta-analysis Dis ColonRectum 2012 55913e923 doi101097DCR0b013e318251d8d9

38 Tsoukalas N Tzovaras AA Tolia M et al Meta-analysis of thepredictive value of KRAS mutations in treatment response usingcetuximab in colorectal cancer J BUON 2012 1773e78

39 Vale CL Tierney JF Fisher D et al Does anti-EGFR therapyimprove outcome in advanced colorectal cancer A systematic reviewand meta-analysis Cancer Treat Rev 2012 38618e625 doi101016jctrv201111002

40 Yang ZY Shen WX Hu XF et al EGFR gene copy number as apredictive biomarker for the treatment of metastatic colorectal cancerwith anti-EGFR monoclonal antibodies a meta-analysis J HematolOncol 2012 552 doi1011861756-8722-5-52

41 Zhang L Ma L Zhou Q Overall and KRAS-specific results ofcombined cetuximab treatment and chemotherapy for metastaticcolorectal cancer a meta-analysis Int J Colorectal Dis 2011 261025e1033 doi101007s00384-011-1197-5

42 Ross JS Clinical implementation of KRAS testing in metastaticcolorectal carcinoma the pathologistrsquos perspective Arch Pathol LabMed 2012 1361298e1307 doi105858arpa2011-0478-RA

43 Bando H Yoshino T Shinozaki E et al Simultaneous identificationof 36 mutations in KRAS codons 61 and 146 BRAF NRAS andPIK3CA in a single reaction by multiplex assay kit BMC Cancer2013 13405 doi1011861471-2407-13-405

44 Douillard JY Oliner KS Siena S et al Panitumumab-FOLFOX4treatment and RAS mutations in colorectal cancer N Engl J Med2013 3691023e1034 doi101056NEJMoa1305275

45 Etienne-Grimaldi MC Mahamat A Chazal M et al Molecular pat-terns in deficient mismatch repair colorectal tumours results from aFrench prospective multicentric biological and genetic study Br JCancer 2014 1102728e2737 doi101038bjc2014213

46 Van Cutsem E Lenz HJ Kohne CH et al Fluorouracil leucovorinand irinotecan plus cetuximab treatment and RAS mutations incolorectal cancer J Clin Oncol 2015 33692e700 doi101200JCO2014594812

47 Xu Q Xu AT Zhu MM et al Predictive and prognostic roles ofBRAF mutation in patients with metastatic colorectal cancer treatedwith anti-epidermal growth factor receptor monoclonal antibodies ameta-analysis J Dig Dis 2013 14409e416 doi1011111751-298012063

48 Yuan ZX Wang XY Qin QY et al The prognostic role of BRAFmutation in metastatic colorectal cancer receiving anti-EGFR mono-clonal antibodies a meta-analysis PLoS One 2013 8e65995 doi101371journalpone0065995

49 Forbes SA Bhamra G Bamford S et al The catalogue of somaticmutations in cancer (COSMIC) Curr Protoc Hum Genet 2008chap10unit 1011

50 Lin JS Webber EM Senger CA et al Systematic review of phar-macogenetic testing for predicting clinical benefit to anti-EGFRtherapy in metastatic colorectal cancer Am J Cancer Res 2011 1650e662

51 Mao C Liao RY Qiu LX et al BRAF V600E mutation and resis-tance to anti-EGFR monoclonal antibodies in patients with metastaticcolorectal cancer a meta-analysis Mol Biol Rep 2011 382219e2223 doi101007s11033-010-0351-4

52 Parsons MT Buchanan DD Thompson B et al Correlation oftumour BRAF mutations and MLH1 methylation with germlinemismatch repair (MMR) gene mutation status a literature reviewassessing utility of tumour features for MMR variant classifica-tion J Med Genet 2012 49151e157 doi101136jmedgenet-2011-100714

53 Cui D Cao D Yang Y et al Effect of BRAF V600E mutation ontumor response of anti-EGFR monoclonal antibodies for first-linemetastatic colorectal cancer treatment a meta-analysis of randomizedstudies Mol Biol Rep 2014 411291e1298 doi101007s11033-013-2974-8

54 Lochhead P Kuchiba A Imamura Y et al Microsatellite instabilityand BRAF mutation testing in colorectal cancer prognostication JNatl Cancer Inst 2013 1051151e1156 doi101093jncidjt173

55 Rowland A Dias MM Wiese MD et al Meta-analysis of BRAFmutation as a predictive biomarker of benefit from anti-EGFRmonoclonal antibody therapy for RAS wild-type metastatic colorectalcancer Br J Cancer 2015 1121888e1894 doi101038bjc2015173

56 Cremolini C Loupakis F Antoniotti C et al FOLFOXIRI plusbevacizumab versus FOLFIRI plus bevacizumab as first-line treat-ment of patients with metastatic colorectal cancer updated overallsurvival and molecular subgroup analyses of the open-label phase 3TRIBE study Lancet Oncol 2015 161306e1315 doi101016S1470-2045(15)00122-9

57 Corcoran RB Atreya CE Falchook GS et al Phase 1-2 trial of theBRAF inhibitor dabrafenib (D) plus MEK inhibitor trametinib (T) inBRAF V600 mutant colorectal cancer (CRC) updated efficacy andbiomarker analysis [ASCOmeeting abstract 3517] J Clin Oncol 2014325s httpmeetinglibraryascoorgcontent131743-144 AccessedNovember 18 2016

58 Bendell JCAtreyaCEAndreacuteT et alEfficacy and tolerability in anopen-label phase III study of MEK inhibitor trametinib (T) BRAF inhibitordabrafenib (D) and anti-EGFR antibody panitumumab (P) in combina-tion in patients (pts) with BRAFV600Emutated colorectal cancer (CRC)[ASCO meeting abstract 3515] J Clin Oncol 2014325s httpmeetinglibraryascoorgcontent131642-144 Accessed November 182016

59 Temraz S Mukherji D Shamseddine A Dual inhibition of MEK andPI3K pathway in KRAS and BRAF mutated colorectal cancers Int JMol Sci 2015 1622976e22988 doi103390ijms160922976

60 Taieb J Zaanan A Le Malicot K et al Prognostic effect of BRAFand KRAS mutations in patients with stage III colon cancer treatedwith leucovorin fluorouracil and oxaliplatin with or without cetux-imab a post hoc analysis of the PETACC-8 trial [published onlineJanuary 14 2016] JAMA Oncol doi101001jamaoncol20155225

61 Barras D BRAF mutation in colorectal cancer an update BiomarkCancer 2015 79e12 doi104137BICS25248

62 Geiersbach KB Samowitz WS Microsatellite instability and colo-rectal cancer Arch Pathol Lab Med 2011 1351269e1277 doi105858arpa2011-0035-RA

63 Barrow E Hill J Evans DG Cancer risk in lynch syndrome FamCancer 2013 12229e240 doi101007s10689-013-9615-1

64 Schmeler KM Lynch HT Chen LM et al Prophylactic surgery toreduce the risk of gynecologic cancers in the Lynch syndrome NEngl J Med 2006 354261e269 doi101056NEJMoa052627

65 Vasen HF Blanco I Aktan-Collan K et al Revised guidelines for theclinical management of Lynch syndrome (HNPCC) recommenda-tions by a group of European experts Gut 2013 62812e823 doi101136gutjnl-2012-304356

66 Funkhouser WK Jr Lubin IM Monzon FA et al Relevance path-ogenesis and testing algorithm for mismatch repair-defective colo-rectal carcinomas a report of the Association for MolecularPathology J Mol Diagn 2012 1491e103 doi101016jjmoldx201111001

67 Kawakami H Zaanan A Sinicrope FA Implications of mismatchrepair-deficient status on management of early stage colorectal can-cer J Gastrointest Oncol 2015 6676e684

68 Diaz LA Jr Le DT PD-1 blockade in tumors with mismatch-repairdeficiency N Engl J Med 2015 3731979 doi101056NEJMc1510353

69 Le DT Uram JN Wang H et al PD-1 blockade in tumors withmismatch-repair deficiency N Engl J Med 2015 3722509e2520doi101056NEJMoa1500596

70 Le DT Uram JN Wang H et al Programmed death-1 blockade inmismatch repair deficient colorectal cancer [ASCO meeting abstract103] J Clin Oncol 201634 httpmeetinglibraryascoorgcontent167415-176 Accessed November 18 2016

71 Yang ZY Wu XY Huang YF et al Promising biomarkers for pre-dicting the outcomes of patients with KRAS wild-type metastaticcolorectal cancer treated with anti-epidermal growth factor receptormonoclonal antibodies a systematic review with meta-analysis Int JCancer 2013 1331914e1925 doi101002ijc28153

72 Pietrantonio F Petrelli F Coinu A et al Predictive role of BRAFmutations in patients with advanced colorectal cancer receivingcetuximab and panitumumab a meta-analysis Eur J Cancer 2015 51587e594 doi101016jejca201501054

73 Wu S Gan Y Wang X et al PIK3CA mutation is associated withpoor survival among patients with metastatic colorectal cancerfollowing anti-EGFR monoclonal antibody therapy a meta-analysisJ Cancer Res Clin Oncol 2013 139891e900 doi101007s00432-013-1400-x

74 Karapetis CS Jonker D Daneshmand M et al PIK3CA BRAFand PTEN status and benefit from cetuximab in the treatment ofadvanced colorectal cancerdresults from NCIC CTGAGITGCO17 Clin Cancer Res 2014 20744e753 doi1011581078-0432CCR-13-0606

75 Liao X Morikawa T Lochhead P et al Prognostic role of PIK3CAmutation in colorectal cancer cohort study and literature review ClinCancer Res 2012 182257e2268 doi1011581078-0432CCR-11-2410

76 Ogino S Liao X Imamura Y et al Predictive and prognostic analysisof PIK3CA mutation in stage III colon cancer intergroup trial J NatlCancer Inst 2013 1051789e1798 doi101093jncidjt298

77 Ogino S Nosho K Kirkner GJ et al PIK3CA mutation is associatedwith poor prognosis among patients with curatively resected coloncancer J Clin Oncol 2009 271477e1484 doi101200JCO2008186544

78 Flossmann E Rothwell PM Effect of aspirin on long-term risk ofcolorectal cancer consistent evidence from randomised and obser-vational studies Lancet 2007 3691603e1613 doi101016S0140-6736(07)60747-8

79 Jacobs EJ Thun MJ Bain EB et al A large cohort study of long-termdaily use of adult-strength aspirin and cancer incidence J Natl CancerInst 2007 99608e615 doi101093jncidjk132

80 Chan AT Giovannucci EL Meyerhardt JA et al Aspirin dose andduration of use and risk of colorectal cancer in men Gastroenterology2008 13421e28 doi101053jgastro200709035

81 Domingo E Church DN Sieber O et al Evaluation of PIK3CAmutation as a predictor of benefit from nonsteroidal anti-inflammatorydrug therapy in colorectal cancer J Clin Oncol 2013 314297e4305doi101200JCO2013500322

82 Liao X Lochhead P Nishihara R et al Aspirin use tumor PIK3CAmutation and colorectal-cancer survival N Engl J Med 2012 3671596e1606 doi101056NEJMoa1207756

83 Kothari N Kim R Jorissen RN et al Impact of regular aspirin use onoverall and cancer-specific survival in patients with colorectal cancer

harboring a PIK3CA mutation Acta Oncol 2015 54487e492 doi1031090284186X2014990158

84 Berg M Danielsen SA Ahlquist T et al DNA sequence profiles ofthe colorectal cancer critical gene set KRAS-BRAF-PIK3CA-PTEN-TP53 related to age at disease onset PLoS One 2010 5e13978 doi101371journalpone0013978

85 Shen Y Yang J Xu Z et al Phosphatase and tensin homologexpression related to cetuximab effects in colorectal cancer patients ameta-analysis World J Gastroenterol 2012 182712e2718 doi103748wjgv18i212712

86 Wang ZH Gao QY Fang JY Loss of PTEN expression as a predictorof resistance to anti-EGFR monoclonal therapy in metastatic colo-rectal cancer evidence from retrospective studies Cancer ChemotherPharmacol 2012 691647e1655 doi101007s00280-012-1886-y

87 Nassif NT Lobo GP Wu X et al PTEN mutations are common insporadic microsatellite stable colorectal cancer Oncogene 2004 23617e628 doi101038sjonc1207059

88 Sawai H Yasuda A Ochi N et al Loss of PTEN expression isassociated with colorectal cancer liver metastasis and poor patientsurvival BMC Gastroenterol 2008 856 doi1011861471-230X-8-56

89 Lin MS Huang JX Chen WC et al Expression of PPARgamma andPTEN in human colorectal cancer an immunohistochemical studyusing tissue microarray methodology Oncol Lett 2011 21219e1224

90 Li XH Zheng HC Takahashi H et al PTEN expression and mutationin colorectal carcinomas Oncol Rep 2009 22757e764

91 Colakoglu T Yildirim S Kayaselcuk F et al Clinicopathological sig-nificance of PTENloss and the phosphoinositide 3-kinaseAkt pathway insporadic colorectal neoplasms is PTENloss predictor of local recurrenceAm J Surg 2008 195719e725 doi101016jamjsurg200705061

92 Eklof V Wikberg ML Edin S et al The prognostic role of KRASBRAF PIK3CA and PTEN in colorectal cancer Br J Cancer 20131082153e2163 doi101038bjc2013212

93 Price TJ Hardingham JE Lee CK et al Prognostic impact and therelevance of PTEN copy number alterations in patients with advancedcolorectal cancer (CRC) receiving bevacizumab Cancer Med 20132277e285 doi101002cam475

94 Frattini M Saletti P Romagnani E et al PTEN loss of expressionpredicts cetuximab efficacy in metastatic colorectal cancer patientsBr J Cancer 2007 971139e1145 doi101038sjbjc6604009

95 Perrone F Lampis A Orsenigo M et al PI3KCAPTEN deregulationcontributes to impaired responses to cetuximab in metastatic colo-rectal cancer patients Ann Oncol 2009 2084e90 doi101093annoncmdn541

96 Sartore-Bianchi A Martini M Molinari F et al PIK3CA mutationsin colorectal cancer are associated with clinical resistance to EGFR-targeted monoclonal antibodies Cancer Res 2009 691851e1857doi1011580008-5472CAN-08-2466

97 Negri FV Bozzetti C Lagrasta CA et al PTEN status in advancedcolorectal cancer treated with cetuximab Br J Cancer 2010 102162e164 doi101038sjbjc6605471

98 Laurent-Puig P Cayre A Manceau G et al Analysis of PTENBRAF and EGFR status in determining benefit from cetuximabtherapy in wild-type KRAS metastatic colon cancer J Clin Oncol2009 275924e5930 doi101200JCO2008216796

99 Tol J Dijkstra JR Klomp M et al Markers for EGFR pathwayactivation as predictor of outcome in metastatic colorectal cancerpatients treated with or without cetuximab Eur J Cancer 2010 461997e2009 doi101016jejca201003036

100 Ulivi P Capelli L Valgiusti M et al Predictive role of multiple gene al-terations in response to cetuximab in metastatic colorectal cancer a singlecenter study J Transl Med 2012 1087 doi1011861479-5876-10-87

101 Cejas P Lopez-Gomez M Aguayo C et al Analysis of theconcordance in the EGFR pathway status between primary tumorsand related metastases of colorectal cancer patients implications forcancer therapy Curr Cancer Drug Targets 2012 12124e131 doi102174156800912799095162

Sepulveda et al

102 Vakiani E Janakiraman M Shen R et al Comparative genomicanalysis of primary versus metastatic colorectal carcinomas J ClinOncol 2012 302956e2962 doi101200JCO2011382994

103 Lee KH Kim JS Lee CS et al KRAS discordance between primaryand recurrent tumors after radical resection of colorectal cancers JSurg Oncol 2015 1111059e1064 doi101002jso23936

104 Aisner DL Deshpande C Baloch Z et al Evaluation of EGFRmutation status in cytology specimens an institutional experienceDiagn Cytopathol 2013 41316e323 doi101002dc21851

105 Lindeman NI Cagle PT Beasley MB et al Molecular testingguideline for selection of lung cancer patients for EGFR and ALKtyrosine kinase inhibitors guideline from the College of AmericanPathologists International Association for the Study of Lung Cancerand Association for Molecular Pathology Arch Pathol Lab Med2013 137828e860 doi105858arpa2012-0720-OA

106 Bettegowda C Sausen M Leary RJ et al Detection of circulatingtumor DNA in early- and late-stage human malignancies Sci TranslMed 2014 6224ra224

107 Clinical Laboratory Improvement Amendments of 1988 42 CFRx4931351e1495 (1992)

108 Sakai K Tsurutani J Yamanaka T et al Extended RAS and BRAFmutation analysis using next-generation sequencing PLoS One 201510e0121891 doi101371journalpone0121891

109 Salto-Tellez M Gonzalez de Castro D Next-generation sequencing achange of paradigm in molecular diagnostic validation J Pathol 20142345e10 doi101002path4365

110 Cottrell CE Al-Kateb H Bredemeyer AJ et al Validation of a next-generation sequencing assay for clinical molecular oncology J MolDiagn 2014 1689e105 doi101016jjmoldx201310002

111 Clinical Laboratory Improvement Amendments of 1988 42 CFRx4931253 (2003)

112 College of American Pathologists CAP laboratory accreditationchecklists Available at httpwwwcaporgappscapportal AccessedNovember 18 2016

113 Clinical and Laboratory Standards Institute Establishing MolecularTesting in Clinical Laboratory Environments Approved Guideline(MM19-A) Wayne PA Clinical Laboratory Standards Institute2011

114 Halling KC Schrijver I Persons DL Test verification and validationfor molecular diagnostic assays Arch Pathol Lab Med 2012 13611e13 doi105858arpa2011-0212-ED

115 Clinical and Laboratory Standards Institute Molecular DiagnosticMethods for Solid Tumors (Nonhematological Neoplasms)(MM23-ED 1) Wayne PA Clinical Laboratory StandardsInstitute 2015

116 Ogino S Gulley ML den Dunnen JT et al Standard mutation nomen-clature in molecular diagnostics practical and educational challenges JMol Diagn 2007 91e6 doi102353jmoldx2007060081

117 Fishel R Lescoe MK Rao MR et al The human mutator genehomolog MSH2 and its association with hereditary nonpolyposiscolon cancer Cell 1993 751027e1038

118 Bronner CE Baker SM Morrison PT et al Mutation in the DNAmismatch repair gene homologue hMLH1 is associated with heredi-tary non-polyposis colon cancer Nature 1994 368258e261 doi101038368258a0

119 Liu B Nicolaides NC Markowitz S et al Mismatch repair genedefects in sporadic colorectal cancers with microsatellite instabilityNat Genet 1995 948e55 doi101038ng0195-48

120 Gologan A Sepulveda AR Microsatellite instability and DNAmismatch repair deficiency testing in hereditary and sporadicgastrointestinal cancers Clin Lab Med 2005 25179e196 doi101016jcll200412001

121 Fitzgibbons PL Bradley LA Fatheree LA et al Principles of analyticvalidation of immunohistochemical assays guideline from the Col-lege of American Pathologists Pathology and Laboratory QualityCenter Arch Pathol Lab Med 2014 1381432e1443 doi105858arpa2013-0610-CP

122 Sepulveda AR The importance of microsatellite instability in colonicneoplasms Medscape 2008571610 wwwmedscapeorgviewarticle571610 Accessed November 18 2016

123 Hatch SB Lightfoot HM Jr Garwacki CP et al Microsatelliteinstability testing in colorectal carcinoma choice of markers affectssensitivity of detection of mismatch repair-deficient tumors ClinCancer Res 2005 112180e2187 doi1011581078-0432CCR-04-0234

124 Watson N Grieu F Morris M et al Heterogeneous staining formismatch repair proteins during population-based prescreening forhereditary nonpolyposis colorectal cancer J Mol Diagn 2007 9472e478 doi102353jmoldx2007060162

125 Cicek MS Lindor NM Gallinger S et al Quality assessment andcorrelation of microsatellite instability and immunohistochemicalmarkers among population- and clinic-based colorectal tumors resultsfrom the Colon Cancer Family Registry J Mol Diagn 2011 13271e281 doi101016jjmoldx201012004

126 Wong NA Gonzalez D Salto-Tellez M et al RAS testing of colo-rectal carcinoma-a guidance document from the Association ofClinical Pathologists Molecular Pathology and Diagnostics GroupJ Clin Pathol 2014 67751e757 doi101136jclinpath-2014-202467

127 Biagi JJ Raphael MJ Mackillop WJ et al Association between timeto initiation of adjuvant chemotherapy and survival in colorectalcancer a systematic review and meta-analysis JAMA 2011 3052335e2342 doi101001jama2011749

128 Franklin WA Haney J Sugita M et al KRAS mutation compari-son of testing methods and tissue sampling techniques in coloncancer J Mol Diagn 2010 1243e50 doi102353jmoldx2010080131

129 Laosinchai-Wolf W Ye F Tran V et al Sensitive multiplex detec-tion of KRAS codons 12 and 13 mutations in paraffin-embeddedtissue specimens J Clin Pathol 2011 6430e36 doi101136jcp2010081539

130 Ma ES Wong CL Law FB et al Detection of KRAS mutations incolorectal cancer by high-resolution melting analysis J Clin Pathol2009 62886e891 doi101136jcp2008063677

131 Pinto P Rocha P Veiga I et al Comparison of methodologies forKRAS mutation detection in metastatic colorectal cancer CancerGenet 2011 204439e446 doi101016jcancergen201107003

132 Tol J Dijkstra JR Vink-Borger ME et al High sensitivity of bothsequencing and real-time PCR analysis of KRAS mutations in colo-rectal cancer tissue J Cell Mol Med 2010 142122e2131 doi101111j1582-4934200900788x

133 Buxhofer-Ausch V Ausch C Zeillinger R et al Duplex reverse-hy-bridization assay for the simultaneous detection of KRASBRAF mu-tations in FFPE-extracted genomic DNA from colorectal cancerspecimens Dis Markers 2013 34171e177 doi1011552013740659

134 Carotenuto P Roma C Rachiglio AM et al Detection of KRASmutations in colorectal carcinoma patients with an integratedPCRsequencing and real-time PCR approach Pharmacogenomics2010 111169e1179 doi102217pgs1086

135 Cavallini A Valentini AM Lippolis C et al KRAS genotyping asbiomarker in colorectal cancer a comparison of three commercial kitson histologic material Anticancer Res 2010 305251e5256

136 Chang YS Yeh KT Hsu NC et al Detection of N- H- and KRAScodons 12 13 and 61 mutations with universal RAS primermultiplex PCR and N- H- and KRAS-specific primer extensionClin Biochem 2010 43296e301 doi101016jclinbiochem200910007

137 Chen YL Chang YS Chang JG et al Genotyping of K-ras codons12 and 13 mutations in colorectal cancer by capillary electrophoresisJ Chromatogr A 2009 12165147e5154 doi101016jchroma200904083

138 Chow L Lin PC Chang JS et al Differences in the frequencies ofK-ras c12-13 genotypes by gender and pathologic phenotypes incolorectal tumors measured using the allele discrimination methodEnviron Mol Mutagen 2012 5322e31 doi101002em20673

139 Kristensen LS Daugaard IL Christensen M et al Increased sensi-tivity of KRAS mutation detection by high-resolution melting anal-ysis of COLD-PCR products Hum Mutat 2010 311366e1373 doi101002humu21358

140 Kristensen LS Kjeldsen TE Hager H et al Competitive amplifica-tion of differentially melting amplicons (CADMA) improves KRAShotspot mutation testing in colorectal cancer BMC Cancer 2012 12548 doi1011861471-2407-12-548

141 Lang AH Drexel H Geller-Rhomberg S et al Optimized allele-specific real-time PCR assays for the detection of common mutationsin KRAS and BRAFJ Mol Diagn 2011 1323e28 doi101016jjmoldx201011007

142 Sundstrom M Edlund K Lindell M et al KRAS analysis in colo-rectal carcinoma analytical aspects of pyrosequencing and allele-specific PCR in clinical practice BMC Cancer 2010 10660 doi1011861471-2407-10-660

143 Viray H Li K Long TA et al A prospective multi-institutionaldiagnostic trial to determine pathologist accuracy in estimation ofpercentage of malignant cells Arch Pathol Lab Med 2013 1371545e1549 doi105858arpa2012-0561-CP

144 Nardon E Glavac D Benhattar J et al A multicenter study tovalidate the reproducibility of MSI testing with a panel of 5quasimonomorphic mononucleotide repeats Diagn Mol Pathol 201019236e242 doi101097PDM0b013e3181db67af

145 Ribic CM Sargent DJ Moore MJ et al Tumor microsatellite-instability status as a predictor of benefit from fluorouracil-based

adjuvant chemotherapy for colon cancer N Engl J Med 2003 349247e257 doi101056NEJMoa022289

146 Jover R Zapater P Castells A et al The efficacy of adjuvantchemotherapy with 5-fluorouracil in colorectal cancer depends on themismatch repair status Eur J Cancer 2009 45365e373 doi101016jejca200807016

147 Human Genome Variation Society Human Genome Variation Soci-ety website wwwhgvsorg Updated September 25 2016 AccessedNovember 18 2016

148 Richards S Aziz N Bale S et al Standards and guidelines for theinterpretation of sequence variants a joint consensus recommenda-tion of the American College of Medical Genetics and Genomics andthe Association for Molecular Pathology Genet Med 2015 17405e424 doi101038gim201530

149 Sukhai MA Craddock KJ Thomas M et al A classification system forclinical relevance of somatic variants identified inmolecular profiling ofcancer Genet Med 2016 18128e136 doi101038gim201547

150 Kalman LV Lubin IM Barker S et al Current landscape and newparadigms of proficiency testing and external quality assessment formolecular genetics Arch Pathol Lab Med 2013 137983e988 doi105858arpa2012-0311-RA

151 Schrijver I Aziz N Jennings LJ et al Methods-based proficiencytesting in molecular genetic pathology J Mol Diagn 2014 16283e287 doi101016jjmoldx201402002

152 Diaz LA Jr Bardelli A Liquid biopsies genotyping circulating tumorDNA JClinOncol 2014 32579e586 doi101200JCO2012452011

Appendix 2 Disclosed Positions of Interest from 2010 e June 2016

Carmen J Allegra MD NRG Oncology Deputy Director and GI Committee co-Chair 2014 e PresentNSABP Foundation Senior vice-President 2014 e PresentJournal National Cancer Institute Editor-in-Chief 2012 e Present

Allison M Cushman-Vokoun MD PhD CAP Molecular Diagnostic Genomic and Precision Medicine Resource Guide CAP PHCCommittee

Molecular Pathology Representative ASCP Curriculum CommitteeGenetics Representative Association for Molecular Pathology (AMP) Training and EducationCommittee

CAP Personalized Health Care (PHC) CommitteeWilliam K Funkhouser MD PhD Molecular Oncology Committee College of American Pathologists 2012-2016

Association of Directors of Anatomic and Surgical Pathology (ADASP)Wayne Grody MD PhD President American College of Medical Genetics and Genomics 2011-2013Stanley R Hamilton MD PhD NCI Colon Task Force member 2015-present

Journal of Pathology Archives of Pathology and Laboratory Medicine 2015-presentNCI Program for Assessment of Clinical Cancer Tests (PACCT) Working Group 2015-presentFDA Immunology Devices Panel Member 2015-presentActionable Genome Consortium Member 2015-presentMolDX Advisory Panel member Palmetto Medicare Administrative Contractor (MAC)2015-present

College of American Pathologists Co-Chair of Colorectal Molecular Markers Expert PanelPhysician Quality Reporting Committee and Economic Affairs Committee member2015-present

College of American Pathologists Cancer Biomarker Reporting Committee member2015-present

Institute of Medicine Committee on Policy Issues in the Clinical Development and Use ofBiomarkers for Molecularly Targeted Therapies Sponsor and testimony on January 29 20152015-present

Medical Evidence Development Consortium (MED-C) member 2015-presentCenter for Medicare and Medicaid Services (CMS) Medicare Evidence Development andCoverage Advisory Committee (MEDCAC) member 2015-present

CMS Advisory Panel on Clinical Diagnostic Laboratory Tests (CDLT) member 2015-presentScott Kopetz MD PhD NCI Colon Task Force Translational Sciences Representative 2012-2016

NCI Colon Task Force Genomics Subcommittee 2012-2016Southwest Oncology Group 2012-2016Colon Cancer National Surgical Adjuvant Breast and Bowel Project (NSABP) the RadiationTherapy Oncology Group (RTOG) and the Gynecologic Oncology Group (GOG) (NRG)Cooperative Group 2013-2016

Gastrointestinal Steering Committee National Cancer Institute 2014-2016Southwest Oncology Group (SWOG) Translational Sciences GI Sub-Committee 2014-2016GI Program NIH Cancer Center Support Grant (CCSG) 2015-2016

(appendix continues)

Appendix 1 Disclosed Interests and Activities 2010 e June 2016

Name InterestActivity Type Entity

Federico A Monzon MD Employment Stock Optionsbonds InVitae 2011-2015Castle Biosciences 2015-present

Antonia Sepulveda MD PhD Consultant American Gastroenterological Society onendoscopy and tissue sampling 2015

Veena M Singh MD Employment Stock Optionsbonds BioceptConsultant bioTheranostics Inc

Lab PMM

Sepulveda et al

Appendix 2 (continued )

Christopher Lieu MD Southwest Oncology Group (SWOG) Translational Sciences GI Sub-Committee 2015-2016Noralane M Lindor MD Genetic Subcommittee of the Cancer Prevention Committee American Society of Clinical

Oncology 2012-2015Federico A Monzon MD Association for Molecular Pathology President 2017

Association for Molecular Pathology President-Elect 2016Association for Molecular Pathology Chair Council of MGP Directors 2014CAP-ASCP-AMP Guideline Development for Colorectal Cancer Molecular Testing AMP Co-chair2011-2013

Association for Molecular Pathology Chair Solid Tumor Subdivision and Council Member2012

College of American Pathologists Technology Assessment Committee (TAC) member2008-2012

Association for Molecular Pathology Clinical Practice Committee 2009-2010Bruce D Minsky MD Chair of the Board American Society for Radiation Oncology

GI Steering Committee National Cancer InstituteJan A Nowak MD PhD CAP Center Committee Pathology and Laboratory Quality Center 2009 e 2015

CAP Public Health Policy Committee (formerly Patient Safety and Performance MeasuresCommittee) 2007- 2011

CAP CGPA Molecular Pathology Working Group 2008- 2012Ad Hoc Committee on Laboratory Quality and Improvement for the 21st Century LQI-21Laboratory Developed Test (LDT) Working Group 2008 e 2010

CAP Molecular Pathology (Molecular Oncology) Resource Committee 2005-2011Measures and Performance Assessment Working Group College of American PathologistsEconomic Affairs Committee 2012 e 2013

Council on Governmental and Professional Affairs - PHC Working Group 2012 e 2016 Collegeof American Pathologists (CAP) Personalized Healthcare Committee (PHC)

Archives of Pathology and Laboratory Medicine Associate Editor for Clinical Pathology2012 e current

CAP Guideline Metrics Expert Panel member 2014 e currentAMA CPT Editorial Panel Member (American Hospital Association) 2015 e currentAMA CPT Molecular Pathology Advisory Group (MPAG) 2015 e currentAMP Professional Relations Committee 2006- 2012AMP Economic Affairs Committee 2009 e current (co-chair 2013 - 2014)AMA CPT Editorial Panel Molecular Pathology Coding Working Tier 1 and Tier 2 Working Groups2009 e 2012

AMP Past President 2010AMP Nominating Committee (chair) 2010Pathology Coding Caucus e AMP Representative 2005-2008 2013 e 2015CMS Medicare Evidence Development and Advisory Committee (MEDCAC) appointee 2012

Daniel J Sargent PhD Alliance for Clinical Trials in Oncology 2011-present

Expert panel members and staff Joseph Willis MD Jennifer Clark SCT(ASCP)CM MBCM Carol Colasacco MLIS R Bryan Rumble MSc Robyn Temple-SmolkinPhD HCLD and Christina Ventura MT(ASCP) have no reported conflicts of interest to discloseThe information above reflects disclosures that were collected and reviewed by the College of American Pathologists the American Society for Clinical

Pathology the American Society of Clinical Oncology and the Association for Molecular Pathology The disclosures that appear in the individual journals of thesocieties may vary based on journal-specific policies and procedures

Molecular Biomarkers for the Evaluation of Colorectal Cancer

Panel Composition


Objective




Inclusion Criteria

Exclusion Criteria


Quality Assessment


Guideline Revision

Disclaimer

Results



Conclusions

Supplemental Data

References