Precision Oncology Decision Support: Current Approaches ...ment of precision oncology decision support. Clin Cancer Res; 24(12); 2719–31. 2018 AACR. Introduction The use of targeted

Review

Precision Oncology Decision Support: CurrentApproaches and Strategies for the FutureKatherine C. Kurnit1, Ecaterina E. Ileana Dumbrava2, Beate Litzenburger3,4,Yekaterina B. Khotskaya3, Amber M. Johnson3, Timothy A. Yap2, Jordi Rodon2,Jia Zeng3, Md Abu Shufean3, Ann M. Bailey3, Nora S. S�anchez3, Vijaykumar Holla3,John Mendelsohn3,5, Kenna Mills Shaw3, Elmer V. Bernstam6, Gordon B. Mills3,7, andFunda Meric-Bernstam2,3,8

Abstract

With the increasing availability of genomics, routine anal-ysis of advanced cancers is now feasible. Treatment selectionis frequently guided by the molecular characteristics of apatient's tumor, and an increasing number of trials aregenomically selected. Furthermore, multiple studies havedemonstrated the benefit of therapies that are chosen basedupon the molecular profile of a tumor. However, the rapidevolution of genomic testing platforms and emergence of newtechnologies make interpreting molecular testing reportsmore challenging. More sophisticated precision oncology

decision support services are essential. This review outlinesexisting tools available for health care providers and precisiononcology teams and highlights strategies for optimizing deci-sion support. Specific attention is given to the assays currentlyavailable for molecular testing, as well as considerations forinterpreting alteration information. This article also discussesstrategies for identifying and matching patients to clinicaltrials, current challenges, and proposals for future develop-ment of precision oncology decision support. Clin Cancer Res;24(12); 2719–31. �2018 AACR.

IntroductionThe use of targeted therapies in molecularly selected patients

has led to a paradigm change in cancer medicine. However,although close to 50 targeted therapies have been approved bythe FDA for solid tumors, only half of these targeted therapieshave predictive biomarkers associated with their FDA approval(Table 1).

As the breadth of molecular testing and the number ofbiomarker-selected trials grow, clinicians are less able to rapidlyinterpret molecular reports during a clinical encounterand determine optimal approved or investigational therapies.The complexities of these processes are highlighted in Fig. 1.

A survey at a major cancer center demonstrated that manyoncologists lack confidence regarding their ability to interpretgenomic information (1). As a result, a new field of precisiononcology decision support (PODS) has emerged. The focus ofthis review is to provide novel insights into the current toolsavailable for PODS as well as to help identify opportunities forfuture development.

Molecular TestingGenomic testing

Identifying clinically relevant characteristics of an individualtumor is critical to precision oncology. Historically, oncologistshave chosen therapies and determined prognoses based on site oforigin and histology. In select tumor types, oncologists beganincorporating biomarkers, such as immunohistochemistry (IHC)for HER2 and estrogen/progesterone receptor status in breastcancer (2, 3), into their decision-making. Today, genomic char-acterization is increasingly being used to guide treatment deci-sions, especially in patients with advanced disease.

Genomic characterization has been performed using a varietyof strategies that range from hot spot exon sequencing of a selectpanel of genes to whole genome sequencing (WGS; refs. 4–8).Early genomic evaluations focused primarily on single-nucleotidevariations (SNV). Assessments have now expanded to evaluatefor other alterations, including indels, translocations, and copy-number variations. Although assays to detect these alterations arenot as widely available as those for SNVs, several such alterationshave been successfully targeted (9–12). Furthermore, some ofthese successeswere for alterations present in rare tumor types thatare unlikely to be studied in tumor-specific clinical trials, as in thecase of NTRK fusions (12, 13). This gave rise to the possibility oftumor-agnostic, molecularly driven registration strategies.

1Gynecologic Oncology and Reproductive Medicine, The University of Texas MDAnderson Cancer Center, Houston, Texas. 2Investigational Cancer Therapeutics,The University of Texas MD Anderson Cancer Center, Houston, Texas. 3SheikhKhalifa Bin Zayed Al Nahyan Institute for Personalized Cancer Therapy, TheUniversity of Texas MD Anderson Cancer Center, Houston, Texas. 4Bioinfor-matics, Qiagen Inc., Redwood City, California. 5Genomic Medicine, The Univer-sity of Texas MD Anderson Cancer Center, Houston, Texas. 6School of Biomed-ical Informatics and Medical School, The University of Texas Health ScienceCenter at Houston, Houston, Texas. 7Systems Biology, The University of TexasMD Anderson Cancer Center, Houston, Texas. 8Breast Surgical Oncology, TheUniversity of Texas MD Anderson Cancer Center, Houston, Texas.

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

Corresponding Author: Funda Meric-Bernstam, The University of Texas MDAnderson Cancer Center, 1400 Holcombe Boulevard, Unit 455, Houston, TX77030. Phone: 713-794-1226; Fax: 713-563-0566; E-mail:[email protected]

doi: 10.1158/1078-0432.CCR-17-2494

�2018 American Association for Cancer Research.

ClinicalCancerResearch

www.aacrjournals.org 2719

on June 12, 2020. © 2018 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst February 2, 2018; DOI: 10.1158/1078-0432.CCR-17-2494

http://crossmark.crossref.org/dialog/?doi=10.1158/1078-0432.CCR-17-2494&domain=pdf&date_stamp=2018-5-18

http://clincancerres.aacrjournals.org/

Table

1.FDA-approve

dtargeted

therap

iesfortumors

that

have

anassociated

biomarker

Preferred

name

Direc

tdrugtarget

Compan

yFDA-approve

dindication:

disea

se(s)

FDA-approve

dindication:

biomarke

r(s)

Abem

aciclib

CDK4,C

DK6

EliLilly

andCompan

yBreastcancer

ERpositive

,PRpositive

,HER2ne

gative

Afatinib

EGFR

Boeh

ring

erIngelhe

imNon–

smallc

elllun

gcarcinoma

EGFRdeletionexon19,E

GFRL8

58R

Alectinib

EML4

–ALK

,ALK

,INSR,R

ET

Gen

entech

Non–

smallc

elllun

gcarcinoma

ALK

fusion

Ana

strozo

leAromatase

AstraZen

ecaPha

rmaceu

ticals

Breastcancer

ERpositive

,PRpositive

Bosutinib

ABL1,B

CR–A

BL1,S

RC

Pfizer

Chronicmye

logen

ous

leuk

emia

BCR–A

BL1

Brigatinib

ALK

,CSF1R,INSR,A

BL1,L

CK,IGF1R,

CAMK2G

,FLT

4,R

ET,F

GFR1,FGFR2,

FGFR3,

AURKA,JAK2,

FGFR4,F

YN,H

CK,

LYN,S

RC,E

GFR,E

ML4

–ALK

,FER,F

ES,

FLT

3,FPS,R

OS1,TYK2,

YES1,PTK2B

,HER4,C

AMK2D

,CHEK1,CHEK2

ARIADPha

rmaceu

ticals

Non–

smallc

elllun

gcarcinoma

ALK

fusion

Ceritinib

NPM–A

LK,R

OS1fusion,

ALK

,INSR,IGF1R,

TSSK3,

FLT

3,FGFR2,

RET,F

GFR3,

LCK,

JAK2,

AURKA,L

YN,E

GFR,F

GFR4

NovartisPha

rmaceu

ticals

Non–smallc

elllun

gcarcinoma

ALK

fusion

Cetux

imab

EGFR

EliLilly

andCompan

yColorectal

cancer

KRASwild

type,

EGFRpositive

Cobim

etinib

MAP2K

1F.H

offman

n-La

Roche

Melan

oma

BRAFV600E,B

RAFV600K

Crizo

tinib

ALK

,MET,R

OS1,NTRK1

Pfizer

Non–

smallc

elllun

gcarcinoma

ALK

fusion

Pfizer

Non–

smallc

elllun

gcarcinoma

ROS1positive

Dab

rafenib

BRAF,R

AF1

Glaxo

SmithK

line

Melan

oma

BRAFV600E

Glaxo

SmithK

line

Melan

oma

BRAFV600E,B

RAFV600K

Glaxo

SmithK

line

Non–

smallc

elllun

gcarcinoma

BRAFV600E

Dasatinib

ABL1,K

IT,B

RAF,B

CR–A

BL1,A

BL2

,PDGFRA,

PDGFRB,S

RC,D

DR1,DDR2,

EPHA3

amplifi

cation,EPHA2,FYN,LCK,LYN,YES1

Bristol-Mye

rsSquibb

Chronicmye

logen

ous

leuk

emia

BCR–A

BL1

Acute

lympho

blastic

leuk

emia

BCR–A

BL1

Ena

siden

ibIDH2

AgiosPha

rmaceu

ticals

Acute

mye

loid

leuk

emia

IDH2mutation

Erlotinib

EGFR

Gen

entech

Non–

smallc

elllun

gcarcinoma


GFRL8

58R

Eve

rolim

usMTOR

NovartisPha

rmaceu

ticals

Breastcancer

ERpositive

,PRpositive

,HER2ne

gative

Exemestane

Aromatase

Pfizer

Breastcancer

ERpositive

Fulve

strant

ER

AstraZen

ecaPha

rmaceu

ticals

Breastcancer

ERpositive

,PRpositive

,HER2ne

gative

Breastcancer

ERpositive

,PRpositive

Gefi

tinib

EGFR

AstraZen

ecaPha

rmaceu

ticals

Non–smallc

elllun

gcarcinoma


GFRL8

58R

Gem

tuzumab

ozo

gam

icin

CD33

Pfizer

Acute

mye

loid

leuk

emia

CD33

positive

Ibrutinib

TEC,A

BL1,F

YN,R

IPK2,

SRC,L

YN,P

DGFRA,

HER2,

BTK,E

GFR,B

LK,B

MX,C

SK,F

GR,

PTK6,H

CK,Y

ES1,ITK,JAK3,

FRK,L

CK,

RET,F

LT3

JanssenBiotech/Pha

rmacyclics

Smalllym

pho

cyticlympho

ma,

chroniclympho

cytic

leuk

emia

c.CHR17pdeletion

Imatinib

PDGFRA,K

IT,B

CR–A

BL1,A

BL1,P

DGFRB

NovartisPha

rmaceu

ticals

Gastrointestinal

stromal

tumors

KIT

positive

Chronicmye

loid

leuk

emia,a

cute

lympho

blastic

leuk

emia

BCR–A

BL1

Mye

lodysplastic/m

yeloproliferativedisea

ses

PDGFRAfusion

Chroniceo

sino

philic

leuk

emia

FIP1L1–PDGFRA

Lapatinib

EGFR,H

ER2,

HER4

NovartisPha

rmaceu

ticals

Breastcancer

HER2positive

Breastcancer

ERpositive

,PRpositive

,HER2positive

Letrozo

leAromatase

NovartisPha

rmaceu

ticals

Breastcancer

ERpositive

,PRpositive

Midostau

rin

KDR,F

LT3,

PDGFRA,P

DGFRB,S

YK,A

KT1,

FLT

1,AKT2,

AKT3,

KIT,S

RC,P

RKCA,

PRKCB,P

RKCG,C

DK1,FGR,E

TV6–N

TRK3

NovartisPha

rmaceu

ticals

Acute

mye

loid

leuk

emia

FLT

3mutation

Neratinib

HER2,

EGFR,K

DR

Pum

aBiotechno

logy

Breastcancer

HER2ove

rexp

ression,

HER2am

plifi

cation

(Continue

donthefollowingpag

e)

Kurnit et al.

Clin Cancer Res; 24(12) June 15, 2018 Clinical Cancer Research2720




Panel testing for genomic alterations is becoming more wide-spread. Many treating physicians utilize commercial testingthrough companies such as FoundationOne (Foundation Med-icine, 315 genes) or Caris Molecular Intelligence (Caris LifeSciences, 592 genes). Some institutions have implemented com-mercial platforms such as Ion Torrent AmpliSeq cancer hot spotpanels (Thermo Fisher Scientific, 50 genes) or Oncomine com-prehensive assay (Thermo Fisher Scientific, 143 genes v1, 161genes v3). A small number of institutions have developed theirown platforms, such as the Memorial Sloan Kettering-IntegratedMutation Profiling of Actionable Cancer Targets (MSK-IMPACT)assay (468 genes; refs. 14, 15). Although local testing requireslocalmolecular diagnostic and bioinformatics support (14, 16), italso allows for customization and inclusion of specific rareralterations within a gene panel. As more actionable alterationsare identified, having the ability to alter the available panels tomeet local needs may be advantageous.

Whole exome sequencing (WES) is a technique that sequencesall the protein-coding genes in a genome (known as exome). Thiscan identify the genetic variants that alter protein sequences at alower cost than WGS (the process of sequencing the completegenome). WES andWGS have predominantly been research toolsbut more recently are being introduced to the clinic (17, 18).These strategies have the disadvantage of needing a strongerbioinformatics and decision support pipeline and potentiallyproviding greater information burden on the clinical providers.In contrast, although hot spot sequencing is much cheaper andeasier to support from a bioinformatics perspective, thesesequencing efforts do not provide as much opportunity fordiscovery. Panel testing of 100 to500 cancer-related genes appearsto have the greatest utilization at this juncture. Panel tests have theadvantages of needing relatively small amounts of DNA, havingthe ability to support some discovery, and providing enoughcoverage and sequencingdepth to allow for detectionof subclonalvariants. WES may especially have a role in patients with noactionable alterations on panel testing, in rare tumors, or intumors that infrequently have alterations on commonly testedgenes. The utility of early routineWES versus selectiveWES testingis likely to become more apparent over the next few years.

Treating oncologists must be aware of the relevant actionablealterations in the diseases they treat (Table 2). Knowledge of theexpected genomic landscape will allow oncologists to ordergenomic testing early in diseases that have many genomic driverslinked to approved therapies, such as in the case of non–small celllung carcinoma or melanoma (19). This understanding will alsoensure that oncologists request assays that evaluate for alterationsmore frequently found in the patient's tumor type (e.g., FGFRfusion testing for cholangiocarcinoma and TRK fusions in secre-tory breast cancer), especially when patients are interested ininvestigational therapies.

Tumors may show genomic evolution over time or may haveintratumor and intertumor heterogeneity between primary andmetastatic sites (20, 21); therefore, repeat tumor sampling is beingpursuedmore frequently. Due to greater tissue availability, testingin patients with metastatic cancer is often done on primarytumors. Although one would expect most truncal alterations tobe found in the primary tumor, when more than one tumorsample is available, the most current sample should be used toaccount for genomic evolution. Although biopsies can usually bedone relatively safely, issues surrounding patient inconvenience,procedure-related pain and complications, and biopsy costsmakeTa

ble

1.FDA-approve

dtargeted

therap

iesfortumors

that

have

anassociated

biomarker(Cont'd)

Preferred

name

Direc

tdrugtarget

Compan

yFDA-approve

dindication:

disea

se(s)

FDA-approve

dindication:

biomarke

r(s)

Nilo

tinib

BCR–A

BL1

Nova

rtisPha

rmaceu

ticals

Chronicmye

logen

ous

leuk

emia

BCR–A

BL1

Olaparib

PARP1,PARP2

AstraZen

ecaPha

rmaceu

ticals

Ovarian

cancer

BRCA1(any

deleterious),BRCA2(any

deleterious)

Osimertinib

EGFR,E

GFRT79

0M,E

GFRexon19

deletion

AstraZen

ecaPha

rmaceu

ticals

Non–

smallc

elllun

gcarcinoma

EGFRT79

0M

Palbociclib

CDK4,C

DK6

Pfizer

Breastcancer

ERpositive

,PRpositive

,HER2ne

gative

Pan

itum

umab

EGFR

Amgen

Colorectal

cancer

KRASwild

type,

NRASwild

type

Pertuzumab

HER2

Gen

entech

Breastcancer,inflam

matory

breastcancer

HER2positive

Pona

tinib

PDGFRA,K

DR,SRC,A

BL1,FGFR1,BCR–A

BL1,

KIT,R

ET

ARIADPha

rmaceu

ticals

Acute

lympho

blastic

leuk

emia/lym

pho

blastic

lympho

ma,

chronicmye

loid

leuk

emia

BCR–A

BL1

T315I

Chronicmye

loid

leuk

emia,a

cute

lympho

blastic

leuk

emia

BCR–A

BL1

Ribociclib

CDK4,C

DK6

NovartisPha

rmaceu

ticals

Breastcancer

ERpositive

,PRpositive

,HER2ne

gative

Ritux

imab

CD20

Gen

entech

Non-Hodgkinlympho

ma,

chroniclympho

cyticleuk

emia

CD20

positive

Rucap

arib

PARP1

ClovisOncology

Ovarian

cancer

BRCA1(any

deleterious),BRCA2(any

deleterious)

Tam

oxifen

ER

AstraZen

ecaPha

rmaceu

ticals

Breastcancer

ERpositive

(may

help

predictwhe

ther

therap

ywillbeben

eficial)

Trametinib

MAP2K

1,MAP2K

2Glaxo

SmithK

line

Melan

oma

BRAFV600E,B

RAFV600K

Non–

smallc

elllun

gcarcinoma

BRAFV600E

Trastuzum

abHER2

Gen

entech

Breastcancer,g

astric

cancer,g

astroesopha

gea

ljunction

HER2positive

Trastuzum

abem

tansine

HER2,

p.Tub

ulin

Gen

entech

Breastcancer

HER2positive

Vem

urafen

ibBRAFV600E

F.H

offman

n-La

Roche

Melan

oma

BRAFV600E

Ven

etoclax

BCL2

AbbVie

Chroniclympho

cyticleuk

emia

c.CHR17pdeletion

Precision Oncology Decision Support

www.aacrjournals.org Clin Cancer Res; 24(12) June 15, 2018 2721




proximal or serial tissue testing challenging. Insurance coverageand procedural availability may also vary greatly, making broadrecommendations difficult to provide. Therefore, the role ofrepeat biopsies and sequencing remains controversial. Considera-tions while deciding between analyzing an archival sample andobtaining a new sample include time since sample collection,number of lines of treatment given in the interim, types oftreatment lines, as well as logistical considerations, includingfeasibility and safety of a newbiopsy. Although there is significantconcordance in specific mutations between primary tumors andmetastasis, there may be changes in actionable alterations (21).There are also now many established mechanisms of acquiredresistance to targeted therapies such as acquired mutations inEGFR and ESR1 with EGFR inhibitors and endocrine therapy,respectively, andBRCAmutation reversionswith PARP inhibitors,demonstrating the value of repeat sampling and genomic analysis(22, 23). For this reason, when possible, genomically matched

trial designs should incorporate "progression biopsies": biopsy ofa progressing lesion in patients who experience progression afterinitial response or prolonged stable disease. This can give impor-tant insights into mechanisms of acquired resistance and ways toovercome them early in drug development.

Liquid biopsiesThe advent of cell-free DNA and circulating tumor cell testing

(i.e., "liquid biopsy") has provided less invasive modalities forgenomic testing. Liquid biopsies were initially undertaken usingallele-specific polymerase chain reaction (PCR) and flow cyto-metry (24). Now, next-generation sequencing panels are increas-ingly used (25). The decision to use panel testing versus PCR-based testing depends on the underlying question.When a knownmutation is being followed during treatment, itmay bemore cost-effective and sensitive to perform focused testing.However, initialtesting or subsequent treatment planning assessments may

© 2018 American Association for Cancer Research

Tumortissue

Bloodsample

Other molecularprofiling

• New biopsy• Archived tissue

• Circulating tumor cells• Cell-free DNA

• IHC• FISH• Tissue microarray• RNA-Seq• Reverse-phase protein array

Resampling and new molecular profile

Next-generationsequencing

Ann

otat

ion

requ

est

Personalized cancer therapyknowledge base

PODS services

Annotation of actionability andfunctional significance

Levels of evidence for that biomarkerin that particular tumor type

Search for the best biomarker-basedtreatment options/clinical trials

Tumor board meeting

Trial Y

Trial ZProgressive

disease

Exceptionalresponder

Trial X

Molecular diagnostics

AGGTTTACAGTTAGTTTTAGTTTTAGGTTTAATGTT ABL1

ACVRL1

AKT1

ALK

APC

APEX1

AR

ARAF

ATM

ATP11B

BAP1

BCL2L1

BCL9

BIRC2

BIRC3

BRAF

BRCA1

BRCA2

BTK

CBL

CCND1

CCNE1

CD274

CD44

CDH1

CDK4

CDK6

CDKN2A

CHEK2

CSF1R

CSNK2A1

CTNNB1

DCUN1D1

DDR2

DNMT3A

EGFR

ERBB2

ERBB3

ERBB4

ESR1

EZH2

FBXW7

FGFR1

FGFR2

FGFR3

FGFR4

FLT3

FOXL2

GAS6

GATA2

GATA3

GNA11

GNAQ

GNAS

HNF1A

HRAS

IFITM1

IFITM3

IDH1

IDH2

IGF1R

IL6

JAK1

JAK2

JAK3

KDR

KIT

KNSTRN

KRAS

MAGOH

MAP2K1

MAP2K2

MAPK1

MAX

MCL1

MDM2

MET

MLH1

MDM4

MED12

MPL

MSH2

MTOR

MYC

MYCL1

MYCN

MYD88

MYO18A

NF1

NF2

NFE2L2

NKX2-1

NKX2-8

NOTCH1

NPM1

NRAS

PDGFRA

PIK3CA

PAX5

PDCD1LG2

PIK3R1

PNP

PPARG

PPP2R1A

PTCH1

PTEN

PTPN11

RAC1

RAF1

RB1

RET

RHEB

RHOA

RPS6KB1

SRC

STAT3

STK11

TERT

TET2

TIAF1

TP53

TSC1

TSC2

U2AF1

VHL

WT1

XPO1

ZNF217

SF3B1

SMAD4

SOX2

SPOP

SMARCB1

SMO

Clinical report

Figure 1.

A flow diagram of precision oncology services that an individual patient may receive while being considered for a targeted agent or investigational therapy. IHC,immunohistochemistry; RNA-Seq, RNA sequencing

Kurnit et al.





Table 2. A list of actionable genes, the alteration types, and the alteration frequencies for several common cancer types

Tumor type Actionable genes Alteration type Frequency Comments

Non–small cell lung cancer BRAF Mutations 5%–10%DDR2 Mutations 1%–6%EGFR Mutations 4%–18%EML4–ALK Fusion 4%ERBB2 Mutations 2%–3%FGFR1 Amplification 2%–17%FGFR3 Fusion 2%KRAS Mutations 1%–32% 1% in adenocarcinoma, 32% in squamous cell

carcinomaMAP2K1 Mutations 1%MET Amplification 1%–4%MET Mutations 3%–8% 3% MET exon 14 mutation in lung adenocarcinomaNF1 Mutations 11%NTRK1 Fusion 2%–4%PIK3CA Mutations 4%–16%PTEN Mutations/deletion 1%–8%RET Fusion 2%–4%RICTOR Amplification 2%–5%ROS1 Fusion 4%–11%STK11 Mutations 2%–17%

Bladder AKT1 Mutations 3%CDKN2A Deletion 47%CDKN2A Mutations 5%EGFR Amplification 11%ERBB2 Amplification 7%ERBB3 Mutations 11%FGFR3 Mutations 45% 60%–80% in non–muscle-invasive bladder cancer;

15–20% in muscle-invasive bladder cancerFGFR3 Amplification 3%FGFR3–TACC3 Fusion 5%KRAS Mutations 4%MDM2 Amplification 9%PIK3CA Mutations 20%PTEN Mutations 3%PTEN Deletion 13%TSC1 Mutations 9%

Biliary BRAF Mutations 7%EGFR Mutations/amplification 5%ERBB2 Mutations/amplification 4%–18% 18% in gallbladder carcinomaFGF19 Amplification 3%FGFR1 Mutations/amplification 4%FGFR2 Fusion 5% 5% in intrahepatic cholangiocarcinomaIDH1/2 Mutations 0%–6% 4%–6% in intrahepatic cholangiocarcinomaKRAS Mutations 18%MDM2 Amplification 5%PIK3CA Mutations 7%PTEN Deletion 1%–7% 7% in gallbladder carcinoma

Gastric EGFR Mutations 3%–5%EGFR Amplification 6%ERBB2 Mutations 5%–7%ERBB2 Amplification 13%ERBB3 Mutations 5%–11%ERBB3 Amplification 4%FGFR1 Mutations 4%FGFR2 Amplification 5%KRAS Mutations 6%MET Mutations 2%MET Amplification 4%PIK3CA Mutations/amplification 24% 42% and 72% in MSI-H and EBVþ gastric cancer,

respectivelyPTEN Mutations 4%–8%PTEN Deletion 4%

Melanoma BRAF Mutations 45%CDKN2A Deletion 13%IDH1 Mutations 6%KDR Amplification 3%KIT Amplification 4%

(Continued on the following page)






Table 2. A list of actionable genes, the alteration types, and the alteration frequencies for several common cancer types (Cont'd )

Tumor type Actionable genes Alteration type Frequency Comments

MAP2K1 Mutations 5%NF1 Mutations 14%NRAS Mutations 10%–25%PDGFRA Amplification 3%

Breast 11q Amplification 15%AKT1 Mutations 2%–4%CDKN2A Deletion 3%–4%ERBB2 Mutations/amplification 13%ESR1 Mutations 10% ERþ breast cancer, metastatic samples and not

primary (marker of resistance to antiestrogentherapy)

FGFR1 Amplification 10%–15%FGFR2 Amplification 4%MAP2K4 Mutations 2%–7%MAP3K1 Mutations 4%–13%NF1 Mutations 2%–4%NTRK3 Fusion 92% Secretory breast cancerPIK3CA Mutations 9%–45%PIK3CA Amplification 4%–5%PIK3R1 Mutations 2%PTEN Mutations/deletion 3%–8%RB1 Mutations/deletion 5%–6% Marker of resistance to CDK4/6 inhibitors

Colorectal AKT1 Mutations 1%–6%BRAF Mutations 3%–47% 47% in MSI-H colorectal cancerERBB2 Mutations/amplification 6%–13%ERBB3 Mutations 4%–20%KRAS Mutations 35%NRAS Mutations 10%PIK3CA Mutations 15%–37% 37% MSI-H colorectal cancerPIK3R1 Mutations 2%–17%PTEN Deletion 4%–20% 20% MSI-H colorectal cancer

Ovarian AKT1 Amplification 3%AKT2 Amplification 2%BRAF Mutations 2%–6% (low-grade

serous ovary)Extremely rare in high-grade ovarian cancer; 2%–6% low-grade serous ovarian cancer (excludingborderline tumors)

BRCA1 (germline or somatic) Mutations 9%BRCA2 (germline or somatic) Mutations 5%CCND1 Amplification 20%CDKN2A Deletion 32%FGFR1 Amplification 5%KRAS Mutations/amplification 19%–33% (low-grade

serous ovary)Extremely rare in high-grade serousovarian cancer;19%–33% low-grade serous ovarian cancer(excluding borderline tumors)

NF1 Mutations/deletion 12%NOTCH3 Mutations/amplification 11%PIK3CA Mutations/amplification 18%PTEN Mutations/deletion 7%

Glioblastoma BRAF Mutations 2%CDK4 Amplification 14%CDK6 Amplification 2%CDKN2A/B Deletion 61%EGFR Mutations 17%–21%EGFR Amplification 41%–44%FGFR1–TACC1 Fusion NAFGFR3–TACC3 Fusion 3%–7%IDH1 Mutations 5%–12%MDM2 Amplification 7%MDM4 Amplification 8%MET Amplification 2%NF1 Mutations 10%NTRK1 Fusion 1%PDGFRA Amplification 10%PIK3CA Mutations/amplification 25%PTEN Mutations/deletion 41%

Abbreviations: EBV, Epstein–Barr virus; MSI-H, microsatellite instability high.

Kurnit et al.





require a broader panel. Furthermore, serial sampling of plasmawith broader sequencing may allow for the detection of acquiredresistance mutations in the targeted gene, as shown with EGFR,FGFR, and HER2 inhibitors (26–28). Larger panels can alsofacilitate identification of alternate resistance mechanisms.

Liquid biopsy also has several limitations. It may miss smallamounts of mutant DNA, as in the cases of lower mutant allelefrequency (subclonal) alterations, patients with limited diseaseburden, and tumor lineages that release small amounts of DNAinto the circulation. It is a strength that liquid biopsy reflects thepool of alterations in a patient, but itmay also bemore difficult toassess mechanisms of resistance in patients with mixed response.Furthermore, certain alteration types may be more difficult todetect, such as copy-number variations (29). With the excitingadvances made possible by single-cell techniques in the assess-ment of tumor heterogeneity and subclonal detection of altera-tions (30, 31), it is likely that liquid biopsies will complementrather than replace tumor testing in the foreseeable future.

Multianalyte testingAlthough genomics has been the primary tool in precision

oncology recently, there is growing recognition that only a subsetof patients has truly compelling genomic alterations, and eventhen, only a subset of patients responds to genotype-matchedtherapy. There is a clear need to move beyond genomics only andexplore the utility of multianalyte testing independently or inconjunction with genomic testing.

As technology has evolved, the field has also begun incorpo-rating transcriptomics and proteomics. Although microarrayshave traditionally been used to evaluate transcriptomics, RNAsequencing (RNA-Seq) is increasingly being used (6, 8, 32).RNA-Seq has already been used to help identify novel alterationtypes that would be missed at the genomics level, such as fusionsand rearrangements (33, 34). RNA-Seq also has the advantage ofdetermining whether the genomic alterations are reflected at theRNA level (e.g., are the oncogenicmutations expressed and are theamplified genes overexpressed?). Newer technologies for prote-omics including reverse-phase protein array, mass spectrometry,and cyclic immunofluorescence have great potential. High-throughput proteomics techniques will likely provide furthernovel insights into potential targets and resistance mechanismsonce these techniques are optimized for the clinical setting (35,36). Many of these novel characterization strategies will likelytransition from the research environment to the clinical setting inthe future. Successful clinical utilization of these rapidly evolvingtechnologies requires a decision support framework to helpinterpret molecular results.

Functional Annotation and DecisionSupportFunctional impact and therapeutic implications

As genomic testing platforms and both preclinical and clinicaldata expand, the interpretation of genomic reports becomesincreasingly complex. Annotation of molecular alterations is theprocess of detailing what change has occurred and the clinicalsignificance of that change.

A change in an amino acid can lead to a change in the activity,expression, or stability of the expressed protein or to no change.Evenwithin the same gene, one alterationmight confer sensitivityto treatment, whereas another may result in resistance (e.g., EGFR

mutations and EGFR inhibitors; refs. 26, 37). Thus, the functionalsignificance of each alteration must be assessed for clinicaldecision-making. In spite of our growing genomic knowledge,larger panels or WES often reveals variants of unknown signifi-cance (VUS) in addition to variants that are well characterized(such as BRAF V600E). The functional impact of these VUS issimply not yet known.

Multiple computational algorithms have been designed topredict the functional impact of specific aberrations; however,their predictive ability remains limited (38). There is growinginterest in functional genomics—generating specific mutationsand assessing the effect of introducing these mutations intoreporter cells, as well as assaying the effect of introducing themutant gene versuswild-type geneon cell growth and survival andonpathway activation.However, it remains unclear if these in vitroassays will be uniformly effective in classifying the functionalimpact of alterations.

A consensus about how to classify therapeutic implications isalso lacking. The use of predetermined levels of evidence may aidin data interpretation, but no universally agreed-upon systemcurrently exists (39–41). Recently, the Association for MolecularPathology, the American Society of Clinical Oncology (ASCO),and the College of American Pathologists have released a con-sensus statement. They propose a four-tiered system for clinicalsignificance of somatic sequence variations, ranging from variantswith strong clinical significance to variants that are benign orlikely benign (42). In practice, however, there are still inconsis-tencies in terms of predicting therapeutic implications (8, 16, 39,43–48). Thus, clinical judgment is still required in addition tointegration of data from large databases and literature reviews.

Actionability of genomic alterations"Actionability" of an alteration has been broadly defined to

mean having potential clinical utility (8, 16, 39, 43–48). Asomatic aberration may be actionable if it alters prognosis orpredicts therapy sensitivity/resistance. In some scenarios, prog-nostic markers may help guide therapeutic choices (e.g., TP53mutations confer worse prognosis in leukemia and may lead toconsideration of transplantation). Germline alterations may alsobe actionable by increasing the risk of cancer or other hereditarydiseases, predicting therapeutic response, or altering drugmetabolism, thereby affecting drug efficacy or toxicity. Overall,actionability implies that knowledge of that specific alteration'spresence would change patient management.

In precision oncology, much of the emphasis has been ondetermining therapeutic actionability. The definitions used at TheUniversity of Texas MD Anderson Cancer Center's PrecisionOncology Program are outlined in Supplementary Table S1.However, there remains no consensus on when a genomic alter-ation should be acted upon. Limiting actionability toFDA-approved therapies for a specific biomarker is restrictive andfails to allow for use of genomic information to inform enroll-ment onto clinical trials or for predicting risk. Alternatively,treating patients with suspected benign alterations or VUS withgenomically targeted agents would decrease the likelihood ofachieving clinical benefit in patients. It would result in patientsreceiving potentially toxic and costly futile therapy that delaysinitiation of effective therapeutic options andwould dilute poten-tial efficacy seen in trials. As the field continues to evolve, ourunderstanding of which alterations are actionable will also grow,and selected VUS will be reclassified as their function and






therapeutic impact is understood.Wewill need to be able to adaptto emerging information quickly.

Use of available resourcesMost decision support teams report using PubMed or other

search engines to review available literature (8, 44, 47, 49–51).Other public databases utilized include COSMIC (cancer.sanger.ac.uk/cosmic), cBio Portal for Cancer Genomics (www.cbioportal.org), ClinGen (www.clinicalgenome.org), UniProt (www.uniprot.org), ClinVar (www.ncbi.nlm.nih.gov/clinvar), IGSR and the1000 Genomes Project (www.internationalgenome.org), anddbSNP (www.ncbi.nlm.nih.gov/projects/SNP). Although theseresources are important, they are not necessarily designed forpoint-of-care clinical use. This underscores the need for special-ized processes to review and interpret the existing molecularknowledge bases. Several groups have therefore developedpublicly available resources to help organize and interpret theimmense amount of existing molecular data.

Some of the most prominent publicly available precisiononcology websites are listed in Table 3. Although most includesimilar general information, important differences between thewebsites include content organization, the ability to search forclinical trials, and the level of detail provided. There are alsodifferences in algorithms for deciding which data to includeand the definitions of actionability. At present, no directcomparisons of these resources exist, and thus the choice of whichwebsite to use depends on users' individual needs and prefer-ences. Although creating andmaintaining these publicly availableresources is a time-consuming and labor-intensive process, suchresources can help bridge the gap between research and practice,particularly when on-site precision oncology support services arenot available.

Many providers or institutions have alternatively sought assis-tance from centralized decision support services. Selected com-mercial services are listed in Supplementary Table S2.

Molecular tumor boardsMany institutions have established molecular tumor boards

(MTB) to help interpret the increasing amount of data available(8, 16, 43–48, 52–54). MTBs are frequentlymultidisciplinary and

include oncologists, research scientists, bioinformaticians, andgenetic counselors (47, 53, 55).

Most MTBs review a patient's clinical, pathologic, andmolecular information, and ultimately make therapyrecommendations that include clinical trial options, use ofFDA-approved agents on label, and occasionally use of inves-tigational agents on a compassionate investigational new drug(IND) or off-label use of FDA-approved therapies (45–47, 49,52, 54, 56, 57). Although clinical trials are potentially the mostdesirable strategy, studies show that the percentage of patientstreated on genomically driven clinical trials is low (2%–18%;refs. 15, 46, 49, 52, 56). In recent studies, the proportion ofpatients whose treatment decision was impacted by genomictesting was 5% to 40% (5, 46, 47, 52, 53, 55, 56), reflectingmany challenges to the routine practice of precision oncology.

As genomic testing becomesmorewidely available, approachessuch as MTBs are unlikely to be scalable. Ideally, available molec-ular data would be incorporated into existing treatment planningconferences and day-to-day clinical workflow, with point-of-caredecision support.

Institutional decision support servicesSeveral large institutions have created teams to centralize

genomic interpretation and provide decision support. At TheUniversity of Texas MD Anderson Cancer Center, we have set upa PODS team to facilitate therapeutic decision-making (39). ThePODS teammaintains highly curated databases for gene variants,functional genomics, drugs, and clinical trials, as previouslydescribed (58, 59).Our current list of actionable genes is availablein Supplementary Table S3 but is continuously evolving.

Most decision support requests are for interpretation of geno-mic results. Clinicians can choose to either interpret the genomicdata on their own or request an annotation by the PODS team. Aconceptual model of the annotation process is shown in Fig. 2.Once completed, the annotation is made available via the elec-tronic health record (59). The PODS team also informs therequesting clinician of any actionable alterations and relevantclinical trials (39). Turnaround time for annotation is critical formaintaining clinical utility, and we have strived to bring this timedown to less than 24 hours, preferably within hours. We expect to

Table 3. Select publicly available PODS websites

Website Maintained byNumberof genes

Organized bygene? Organized by tumor type? Clinical trials?

Personalized Cancer Therapywww.personalizedcancertherapy.org(Accessed: June 26, 2017)

The University of TexasMD Anderson CancerCenter

30 Yes No: tumor-specificinformation listed withinthe gene page

Yes: searchable bygene

My Cancer Genomewww.mycancergenome.org(Accessed: June 26, 2017)

Vanderbilt-IngramCancer Center

823 No: must searchby tumor type(except trials)

Yes Yes: searchable bygene and tumortype

OncoKBoncokb.org (Accessed: June 26, 2017)

Memorial Sloan KetteringCancer Center,partnership with QuestDiagnostics

476 Yes No: frequency data fortumor types listed withinthe gene page

No

Precision Cancer Medicinewww.precisioncancermedicine.org(Accessed: June 26, 2017)

Dana-Farber/Brighamand Women's CancerCenter

17 No Yes: clinical trials organizedprimarily by tumor type

Yes: limited to thoseoffered atDana-Farber

The Drug Gene Interaction Databasedgidb.genome.wustl.edu(Accessed: June 26, 2017)

Washington Universityin St. Louis

8,419 Yes No: would need to reviewreferences fortumor-type information

No

Kurnit et al.




www.cbioportal.org

www.cbioportal.org

www.clinicalgenome.org

www.uniprot.org

www.uniprot.org

www.ncbi.nlm.nih.gov/clinvar

www.internationalgenome.org

www.ncbi.nlm.nih.gov/projects/SNP

www.personalizedcancertherapy.org

www.mycancergenome.org

www.precisioncancermedicine.org


© 2018 American Association for Cancer Research

Are alterations of this type clinically targetable? Is the alteration actionable?What does the specificalteration do?

Levels of evidence

Drug is FDA approved fora specific biomarker in aspecific tumor type or ahistology-agnostic indication

Evidence demonstrating thata biomarker predicts tumorresponse to the drug or that

in a biomarker-selected cohort*(within a specific tumor typeor compelling evidence acrossmultiple tumor types withminimal tumor-typeheterogeneity)

Large-scale, retrospective studydemonstrating that a biomarkeris associated with tumorresponse to the drug**

Clinical data that a biomarkerpredicts tumor response to the

(does not fulfill criteria oflevel 1B or 2A)

Unusual responder(s), eithersingle-patient case studies orsmall case series, showing abiomarker is associated withresponse to the drug and issupported by scientific rationale

Preclinical data demonstratingthat a biomarker predictsresponse of cells or tumors todrug treatment

Actionablegene and

alteration type?Variant-level actionabilityFunctional

significanceConfer therapy

response?

Yes

No

Yes

Inferred

No

STOP

Activating Yes—literature based

Yes—inferred

Yes—functional genomics

Potentially

Unknown

No

Inactivating

Inferredactivating

Inferredinactivating

Unknown

Likelybenign

1A

1B

2A

2B

3A

3B

Step 1 Step 2 Step 3

Figure 2.

Procedural flow used by the PODS team at The University of Texas MD Anderson Cancer Center for annotating an alteration. Our tiers for levels ofevidence, functional significance, and variant level of actionability are included. Terminology is defined in Supplementary Table S1. �For level ofevidence 1B, "Evidence" could be (i) an adequately powered, prospective study with biomarker selection/stratification, (ii) a meta-analysis/overview,(iii) a consensus recommendation for standard of care [as recommended by National Comprehensive Cancer Network (NCCN) guidelines or other consortia].��For level of evidence 2A, "Evidence" could be (i) a prospective trial where biomarker study is the secondary objective, (ii) an adequately poweredretrospective cohort study, or (iii) an adequately powered case–control study.






transition to routine annotation of testing performed in-housesoon but also expect that point-of-care services will still be desiredfor testing performed elsewhere.

At this time, the PODS teamhas received over 3,629 annotationrequests on 2,741 patients with over 50 tumor types (59). The first2,444 PODS annotations performed for 669 patients that includ-ed an actionable variant call were recently reviewed: 32.5% wereactionable, 9.4% potentially actionable, 29.7% unknown, 28.4%nonactionable (59). Of patients with actionable or potentiallyactionable alterations, 20.6% enrolled on a genomically matchedtrial. Trial enrollment was higher for actionable/potentiallyactionable alterations (27.6%) than those with unknown(11.8%) and nonactionable (3%) alterations (P¼ 0.00004). Thisdemonstrates the interest in and perceived value of decisionsupport, even in academic settings with highly specialized oncol-ogists. Decision support needs in the community may be evenhigher, with the caveat that therapeutic options may be morelimited due to lack of access to trials.

Decision Support Efforts in PrecisionOncology TrialsDecision support in precision oncology trials

There are now also several large, multicenter precision oncol-ogy studies. We will highlight four important examples to dem-onstrate decision support approaches utilized.

NCI-MATCH (Molecular Analysis for Therapy Choice,NCT02465060) is a prospective precision oncology trial in whichpatients are assigned to receive treatment based on alterationsfound in their tumors through genomic sequencing (and a fewIHC assays; refs. 57, 60). Treatment decisions are based uponpredetermined algorithms (MATCHBOX; ref. 57), with review ofmatches by amultidisciplinary team including oncologists, infor-maticians, and pathologists.

NCI-MPACT (Molecular Profiling–BasedAssignment ofCancerTherapy, NCT01827384) is a randomized, biomarker-drivenclinical trial. It uses next-generation sequencing to evaluate foralterations impacting RAS/RAF/MEK, PI3K, and DNA damagerepair (61). Prespecified rules designate alterations as beingactionable. For the study, the NCI developed a system calledGeneMed, which helps to determine appropriate treatment allo-cation (61). GeneMed automatically identifies suspected action-able alterations, which are then individually reviewed. These dataare then used to randomize the patient. Automating part of theprocess reduces the burden on the decision support team andstandardizes the definition of actionability across sites.

ASCO's Targeted Agent and Profiling Utilization Registry(TAPUR) is also ongoing. The goal of this registry study is todescribe drug activity and toxicity for novel agents used in thesetting of an actionable variant (62). This study has the potentialto increase access to targeted agents and PODS to communityoncology providers as well as participating academic centers. Inthis study, the provider can make the initial determination forpatient eligibility and receive central confirmation (62).However,an MTB, which is organized by ASCO, is also available forconsultation. All drugs are provided by the pharmaceutical com-panies, and all data are collected and monitored centrally. Inaddition to providing important toxicity and efficacy informa-tion, this study will also assess practitioner knowledge.

Similar trials are also currently being conducted successfullyoutside of the United States. The WINTHER study is one such

study that serves as an interesting model for incorporatingtranscriptomics into treatment decisions (63). WINTHER alsohighlights some of the considerations relevant to future interna-tional collaborations for PODS. Specifically, attention to differ-ences in regulatory testing requirements across countries andbarriers to using similar platforms under different regulatoryenvironments will be important for future trials (63).

Challenges to PODSExpanding available biomarkers

Information about transcriptional output, protein expression,epigenetic modifications, and metabolomics may provide abroader understanding of mechanisms of tumor growth andallow for better treatment decisions (8, 32, 64). Although diffi-culties with RNA preservation in formalin-fixed paraffin-embed-ded tissues persist (65), approaches such as RNA-Seq are nowtransitioning to the Clinical Laboratory Improvement Amend-ments (CLIA) environment. Further investigation into optimizingtechniques for functional proteomics or to operationalize use ofmultiplex IHC will be critical (66). Once these technologies areimplemented in the clinical setting, however, the need for scien-tific and bioinformatics support will only increase.

Patients with multiple alterations remain another challenge.Targets often are prioritized, taking into consideration the allelicfrequency of different alterations, relative copy-number gain/lossfor copy-number alterations, and level of evidence of therapeuticimplications. Sometimes the presence of a second alteration isthought to be a contraindication for acting on the first alteration(as in the case of activating mutation of EGFR and KRAS). Thisproblem of multiple alterations is likely to grow with broadertesting panels. However,multianalyte testingmay further assist indecision-making by elucidating which downstream pathway isactivated, etc.

Immunotherapy is also a growing field. Although many of thedecision support criteriamay be different, the precepts underlyingthe importance of decision support are the same. Biomarkers thatpredict sensitivity or resistance to immunotherapies are starting tobe developed, including tumor mutation burden, neoantigenload, neoantigen expression, tumor-infiltrating lymphocytes,CD8þ cells, PDL1 expression, and immune infiltrate gene expres-sion signatures.However, someare still controversial (e.g., PDL1),and others are not readily available on all tumor types (e.g., tumormutation burden; ref. 67). Although the predictive value ofgenomics and other multianalyte techniques has shown promise,this work is still in its infancy as compared with biomarkerdiscovery for targeted therapies (68, 69). As indications andtherapeutic approaches for immunotherapy continue to grow, itis imperative that PODS services are equipped to interpret theseclasses of biomarkers.

Operational considerationsCost is an ongoing barrier for precision oncology. Although the

cost of next-generation sequencing is decreasing (70), new tech-nologies are continuously being developed. As a field, there isinterest in broadening molecular testing for patients to addressclonal evolution andmolecular heterogeneity. However, this willalso increase both the amount of data and the frequency at whichdata must be analyzed. Furthermore, questions about insurancecoverage of testing persist despite increased use of sequencing inoncologic care (52, 71). Additionally, the cost of maintaining

Kurnit et al.





decision support infrastructure itself must also be taken intoaccount (72).

Access to investigational therapies through clinical trials isalso relevant. The large number of exclusions to eligibility canlimit the ability of patients with genomic aberrations to receivematched therapies. Additionally, insurance concerns aboutenrollment on trials persist. Clinical trial access is also relatedto geographic limitations, as a large proportion of patients withcancer live a burdensome distance from the closest trial site(73). Novel precision oncology initiatives, such as ASCO'sTAPUR study, will enhance access to off-label drugs whileensuring oncologic outcomes are collected.

The institutional nature of most PODS services is both anasset and a weakness. With local review, recommendations arebased upon the specific needs and resources of that institution.However, much of the same work is done independently bymany institutions. Multiple institutions maintain their owndatabases for research, including evolving information aboutactionability and available clinical trials. Although this state ofdecentralized science is not limited to precision oncology, itdoes limit scalability and reproducibility. However, there arealso financial, academic, and logistical incentives for maintain-ing separate databases and knowledge bases. Regardless,increasing efficiency and decreasing cost of decision supportservices will be necessary for precision oncology to become thenew model for oncology care.

ConclusionsPrecision oncology is an exciting and rapidly evolving field.

With a greater number of therapies and diagnostics, there is acritical need for PODS. As our ability to target specific path-ways improves, so too must our processes for identifyingappropriate patients. "All-comer" trials are increasinglyreplaced by molecularly matched clinical trials. While this islikely to improve clinical outcomes, it also means that theknowledge required to practice oncology is becoming increas-ingly complex. Thus, more sophisticated and organized PODSservices are critical to implementing precision oncology inroutine cancer care.

Disclosure of Potential Conflicts of InterestJ. Rodon reports receiving commercial research grants from Bayer and

Novartis and is a consultant/advisory board member for Lilly, Novartis, Orion,Peptomyc, and Servier. E.V. Bernstam reports receiving commercial researchgrants fromAileron, AstraZeneca, Bayer, Boehringer Ingelheim, Calithera, Curis,CytoMx, Debiopharm, Effector Therapeutics, Genentech, Novartis, Pfizer,Puma, Taiho, and Zymeworks and is a consultant/advisory board member forClearlight Diagnostics, Darwin Health, Dialecta, GRAIL, Inflection Biosciences,Pieris, and Sumitomo Dainippon Pharma. G.B. Mills reports receiving com-mercial research grants from AstraZeneca, Critical Outcome Technologies,Illumina, Karus, Nanostring, Pfizer, and Takeda/Millenium Pharmaceuticals,reports receiving speakers bureau honoraria from Allostery, AstraZeneca,ImmunoMet, ISIS Pharmaceuticals, Lilly, MedImmune, Novartis, Pfizer, Sym-phogen, and Tarveda, holds ownership interest (including patents) in CatenaPharmaceuticals, ImmunoMet, Myriad Genetics, PTV Ventures, and SpindletopVentures, and is a consultant/advisory board member for Adventist Health,Allostery, AstraZeneca, Catena Pharmaceuticals, Critical Outcome Technolo-gies, ImmunoMet, ISIS Pharmaceuticals, Lilly,MedImmune,Novartis, PrecisionMedicine, Provista Diagnostics, Signalchem Lifesciences, Symphogen, Takeda/Millenium Pharmaceuticals, Tarveda, and Tau Therapeutics. F. Meric-Bernstamreports receiving commercial research grants from Aileron, AstraZeneca, Bayer,Calithera, Curis, CytoMx, Debio, eFFECTOR, Genentech, Jounce, Novartis,Pfizer, Puma, Taiho, and Zymeworks and is a consultant/advisory boardmember for Clearlight Diagnostics, Darwin Health, Dialecta, GRAIL, InflectionBiosciences, Pieris, and SumitomoDainippon. No potential conflicts of interestwere disclosed by the other authors.

AcknowledgmentsThis work was supported in part by 1U01CA180964 (F.Meric-Bernstam and

E.V. Bernstam), NIH Research Training Grant T32 CA101642 (K.C. Kurnit),The Cancer Prevention and Research Institute of Texas (RP150535; to F. Meric-Bernstam, E.V. Bernstam, A.M. Johnson, J. Zeng, A.M. Bailey, V. Holla,N.S. S�anchez, andK.M. Shaw), the Sheikh Khalifa Bin Zayed Al Nahyan Institutefor PersonalizedCancer Therapy (F.Meric-Bernstam,G.B.Mills, B. Litzenburger,A.M. Johnson, Y.B. Khotskaya, J. Zeng, M.A. Shufean, A.M. Bailey, N.S. S�anchez,V. Holla, J. Mendelsohn, and K.M. Shaw), NCATS GrantUL1 TR000371(Center for Clinical and Translational Sciences; to F. Meric-Bernstam andE.V. Bernstam), The Bosarge Family Foundation (A.M. Bailey, V. Holla, J. Zeng,and G.B. Mills), and The University of Texas MD Anderson Cancer CenterSupport Grant (P30 CA016672; to F. Meric-Bernstam).

The authors would like to thank Ms. Elena Vess, Ms. Amy Simpson, andDr. Lauren Brusco for their administrative support on this article.

Received August 30, 2017; revised November 2, 2017; accepted January 30,2018; published first February 2, 2018.

References1. Gray SW, Hicks-Courant K, Cronin A, Rollins BJ, Weeks JC. Physicians'

attitudes about multiplex tumor genomic testing. J Clin Oncol 2014;32:1317–23.

2. Iwamoto T, Booser D, Valero V, Murray JL, Koenig K, Esteva FJ, et al.Estrogen receptor (ER) mRNA and ER-related gene expression in breastcancers that are 1% to 10% ER-positive by immunohistochemistry. J ClinOncol 2012;30:729–34.

3. Toikkanen S, Helin H, Isola J, Joensuu H. Prognostic significance of HER-2oncoprotein expression in breast cancer: a 30-year follow-up. J Clin Oncol1992;10:1044–8.

4. Hasmats J, GreenH,Orear C, Validire P,HussM, KallerM, et al. Assessmentofwhole genomeamplification for sequence capture andmassively parallelsequencing. PLoS One 2014;9:e84785.

5. Beltran H, Eng K, Mosquera JM, Sigaras A, Romanel A, Rennert H, et al.Whole-exome sequencing of metastatic cancer and biomarkers of treat-ment response. JAMA Oncol 2015;1:466–74.

6. Uzilov AV, Ding W, Fink MY, Antipin Y, Brohl AS, Davis C, et al. Devel-opment and clinical application of an integrative genomic approach topersonalized cancer therapy. Genome Med 2016;8:62.

7. Meric-Bernstam F, Brusco L, Shaw K, Horombe C, Kopetz S, Davies MA,et al. Feasibility of large-scale genomic testing to facilitate enrollment ontogenomically matched clinical trials. J Clin Oncol 2015;33:2753–62.

8. Roychowdhury S, Iyer MK, Robinson DR, Lonigro RJ, Wu YM, Cao X, et al.Personalized oncology through integrative high-throughput sequencing: apilot study. Sci Transl Med 2011;3:111ra21.

9. O'Brien SG, Guilhot F, Larson RA, Gathmann I, Baccarani M, Cervantes F,et al. Imatinib comparedwith interferon and low-dose cytarabine for newlydiagnosed chronic-phase chronic myeloid leukemia. N Engl J Med 2003;348:994–1004.

10. Shaw AT, Ou SH, Bang YJ, Camidge DR, Solomon BJ, Salgia R, et al.Crizotinib in ROS1-rearranged non-small-cell lung cancer. N Engl J Med2014;371:1963–71.

11. Solomon BJ, Mok T, Kim DW, Wu YL, Nakagawa K, Mekhail T, et al. First-line crizotinib versus chemotherapy in ALK-positive lung cancer. N Engl JMed 2014;371:2167–77.

12. Patients with NTRK fusions respond to targeted therapies. Cancer Discov2016;6:566–7.

13. Doebele RC, Davis LE, Vaishnavi A, Le AT, Estrada-Bernal A, Keysar S, et al.An oncogenic NTRK fusion in a patient with soft-tissue sarcoma withresponse to the tropomyosin-related kinase inhibitor LOXO-101. CancerDiscov 2015;5:1049–57.

14. Cheng DT, Mitchell TN, Zehir A, Shah RH, Benayed R, Syed A, et al.Memorial sloan kettering-integrated mutation profiling of actionable can-cer targets (MSK-IMPACT): a hybridization capture-based next-generation






sequencing clinical assay for solid tumor molecular oncology. J Mol Diag2015;17:251–64.

15. Zehir A, Benayed R, Shah RH, Syed A, Middha S, KimHR, et al. Mutationallandscape of metastatic cancer revealed from prospective clinical sequenc-ing of 10,000 patients Nat Med 2017;23:703–13.

16. Pritchard CC, Salipante SJ, Koehler K, Smith C, Scroggins S, Wood B, et al.Validation and implementation of targeted capture and sequencing for thedetection of actionable mutation, copy number variation, and gene rear-rangement in clinical cancer specimens. J Mol Diag 2014;16:56–67.

17. Horak P, Frohling S, Glimm H. Integrating next-generation sequencinginto clinical oncology: strategies, promises and pitfalls. ESMO Open2016;1:e000094.

18. Meyerson M, Gabriel S, Getz G. Advances in understanding cancer gen-omes through second-generation sequencing. Nat Rev Genet 2010;11:685–96.

19. KrisMG, Johnson BE, Berry LD, KwiatkowskiDJ, Iafrate AJ,Wistuba II, et al.Using multiplexed assays of oncogenic drivers in lung cancers to selecttargeted drugs. JAMA 2014;311:1998–2006.

20. Jamal-Hanjani M, Quezada SA, Larkin J, Swanton C. Translational impli-cations of tumor heterogeneity. Clin Cancer Res 2015;21:1258–66.

21. Meric-Bernstam F, Frampton GM, Ferrer-Lozano J, Yelensky R, Perez-Fidalgo JA, Wang Y, et al. Concordance of genomic alterations betweenprimary and recurrent breast cancer. Mol Cancer Ther 2014;13:1382–9.

22. Jeselsohn R, Yelensky R, Buchwalter G, Frampton G, Meric-Bernstam F,Gonzalez-Angulo AM, et al. Emergence of constitutively active estrogenreceptor-alpha mutations in pretreated advanced estrogen receptor-posi-tive breast cancer. Clin Cancer Res 2014;20:1757–67.

23. Domchek SM. Reversion mutations with clinical use of PARP inhibitors:many genes, many versions. Cancer Discov 2017;7:937–9.

24. SorensonGD, PribishDM,Valone FH,Memoli VA, BzikDJ, Yao SL. Solublenormal and mutated DNA sequences from single-copy genes in humanblood. Cancer Epidemiol Biomarkers Prev 1994;3:67–71.

25. SchwaederleM,HusainH, Fanta PT, PiccioniDE, Kesari S, SchwabRB, et al.Detection rateof actionablemutations indiverse cancers using abiopsy-free(blood) circulating tumor cell DNA assay. Oncotarget 2016;7:9707–17.

26. Yu HA, Arcila ME, Rekhtman N, Sima CS, Zakowski MF, Pao W, et al.Analysis of tumor specimens at the time of acquired resistance to EGFR-TKItherapy in 155 patients with EGFR-mutant lung cancers. Clin Cancer Res2013;19:2240–7.

27. Goyal L, Saha SK, Liu LY, Siravegna G, Leshchiner I, Ahronian LG, et al.Polyclonal secondary FGFR2 mutations drive acquired resistance to FGFRinhibition in patients with FGFR2 fusion-positive cholangiocarcinoma.Cancer Discov 2017;7:252–63.

28. Ma C, Bose R, Gao F, Freedman R, Telli M, Kimmick G, et al. Circulatingtumor DNA (ctDNA) sequencing for HER2 mutation (HER2mut)screening and response monitoring to neratinib in metastatic breastcancer (MBC) [abstract]. In: Proceedings of the American Associationfor Cancer Research Annual Meeting 2017; 2017 Apr 1–5; Washington,DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstractnr CT011.

29. Chan KC, Jiang P, Zheng YW, Liao GJ, SunH,Wong J, et al. Cancer genomescanning in plasma: detection of tumor-associated copy number aberra-tions, single-nucleotide variants, and tumoral heterogeneity by massivelyparallel sequencing. Clin Chem 2013;59:211–24.

30. NavinN, Kendall J, Troge J, Andrews P, Rodgers L,McIndoo J, et al. Tumourevolution inferred by single-cell sequencing. Nature 2011;472:90–4.

31. Chung W, Eum HH, Lee HO, Lee KM, Lee HB, Kim KT, et al. Single-cellRNA-seq enables comprehensive tumour and immune cell profiling inprimary breast cancer. Nat Commun 2017;8:15081.

32. LoRusso PM, Boerner SA, PilatMJ, Forman KM, Zuccaro CY, Kiefer JA, et al.Pilot trial of selecting molecularly guided therapy for patients with non-V600 BRAF-mutant metastatic melanoma: experience of the SU2C/MRAmelanoma dream team. Mol Cancer Ther 2015;14:1962–71.

33. Maher CA, Kumar-Sinha C, Cao X, Kalyana-Sundaram S, Han B, Jing X,et al. Transcriptome sequencing to detect gene fusions in cancer. Nature2009;458:97–101.

34. Kohno T, Ichikawa H, Totoki Y, Yasuda K, Hiramoto M, Nammo T, et al.KIF5B-RET fusions in lung adenocarcinoma. Nat Med 2012;18:375–7.

35. Nilsson T, MannM, Aebersold R, Yates JR, Bairoch A 3rd, Bergeron JJ. Massspectrometry in high-throughput proteomics: ready for the big time. NatMethods 2010;7:681–5.

36. Akbani R,NgPK,WernerHM, ShahmoradgoliM, Zhang F, Ju Z, et al. Apan-cancer proteomic perspective on the cancer genome atlas. Nat Commun2014;5:3887.

37. Riely GJ, Pao W, Pham D, Li AR, Rizvi N, Venkatraman ES, et al. Clinicalcourse of patients with non-small cell lung cancer and epidermal growthfactor receptor exon 19 and exon 21 mutations treated with gefitinib orerlotinib. Clin Cancer Res 2006;12:839–44.

38. Tarcic G, Kamal M, Edelheit O, Barbash Z, Vidne M, Miron B, et al.Functional mutational analysis to assess the oncogenic activity of variantof uncertain significance (VUS) detected in patients included in the SHIVAtrial. Eur J Cancer 2016;69:S6–S7.

39. Johnson A, Zeng J, Bailey AM, Holla V, Litzenburger B, Lara-Guerra H, et al.The right drugs at the right time for the right patient: the MD Andersonprecision oncology decision support platform. Drug Discov Today 2015;20:1433–8.

40. Van Allen EM,Wagle N, Stojanov P, Perrin DL, Cibulskis K, Marlow S, et al.Whole-exome sequencing and clinical interpretation of formalin-fixed,paraffin-embedded tumor samples to guide precision cancermedicine.NatMed 2014;20:682–8.

41. Andre F, Mardis E, SalmM, Soria JC, Siu LL, Swanton C. Prioritizing targetsfor precision cancer medicine. Ann Oncol 2014;25:2295–303.

42. Li MM, Datto M, Duncavage EJ, Kulkarni S, Lindeman NI, Roy S, et al.Standards and Guidelines for the Interpretation and reporting of sequencevariants in cancer: a joint consensus recommendationof the association formolecular pathology, American Society of Clinical Oncology, and Collegeof American Pathologists. J Mol Diagn 2017;19:4–23.

43. Bardia A, Iafrate JA, Sundaresan T, Younger J, Nardi V. Metastatic breastcancer with ESR1 mutation: clinical management considerations from themolecular and precision medicine (MAP) tumor board at MassachusettsGeneral Hospital. Oncologist 2016;21:1035–40.

44. Knepper TC, Bell GC, Kevin Hicks J, Padron E, Teer JK, Vo TT, et al. Keylessons learned from moffitt's molecular tumor board: the clinical geno-mics action committee experience. Oncologist 2017;22:144–51.

45. ParsonsHA, Beaver JA,Cimino-MathewsA, Ali SM,Axilbund J, ChuD, et al.Individualized molecular analyses guide efforts (IMAGE): a prospectivestudy of molecular profiling of tissue and blood in metastatic triple-negative breast cancer. Clin Cancer Res 2017;23:379–86.

46. Kaderbhai CG, Boidot R, Beltjens F, Chevrier S, Arnould L, Favier L, et al.Use of dedicated gene panel sequencing using next generation sequencingto improve the personalized care of lung cancer. Oncotarget 2016;7:24860–70.

47. Tafe LJ, Gorlov IP, de Abreu FB, Lefferts JA, Liu X, Pettus JR, et al.Implementation of a molecular tumor board: the impact on treatmentdecisions for 35 patients evaluated at Dartmouth-Hitchcock MedicalCenter. Oncologist 2015;20:1011–8.

48. Wagle N, Berger MF, Davis MJ, Blumenstiel B, Defelice M, Pochanard P,et al. High-throughput detection of actionable genomic alterations inclinical tumor samples by targeted, massively parallel sequencing. CancerDiscov 2012;2:82–93.

49. Mantripragada KC, Olszewski AJ, Schumacher A, Perez K, Birnbaum A,Reagan JL, et al. Clinical trial accrual targeting genomic alterations afternext-generation sequencing at a non-national cancer institute-designatedcancer program. J Oncol Pract 2016;12:e396–404.

50. Bailey AM, Mao Y, Zeng J, Holla V, Johnson A, Brusco L, et al. Implemen-tation of biomarker-driven cancer therapy: existing tools and remaininggaps. Discov Med 2014;17:101–14.

51. Meric-Bernstam F, Farhangfar C, Mendelsohn J, Mills GB. Building apersonalizedmedicine infrastructure at amajor cancer center. J Clin Oncol2013;31:1849–57.

52. McGowanML, Ponsaran RS, Silverman P, Harris LN,Marshall PA. "A risingtide lifts all boats": establishing a multidisciplinary genomic tumor boardfor breast cancer patients with advanced disease. BMC Med Genet 2016;9:71.

53. GundersonCC, RowlandMR,Wright DL, Andrade KL,Mannel RS,McMee-kin DS, et al. Initiation of a formalized precision medicine program ingynecologic oncology. Gynecol Oncol 2016;141:24–8.

54. Hirshfield KM, Tolkunov D, Zhong H, Ali SM, Stein MN, Murphy S, et al.Clinical actionability of comprehensive genomic profiling for manage-ment of rare or refractory cancers. Oncologist 2016;21:1315–25.

55. Parker BA, Schwaederle M, Scur MD, Boles SG, Helsten T, Subramanian R,et al. Breast cancer experience of the molecular tumor board at the


Kurnit et al.




University of California, San Diego Moores Cancer Center. J Oncol Pract2015;11:442–9.

56. Sohal DP, Rini BI, Khorana AA, Dreicer R, Abraham J, Procop GW, et al.Prospective clinical study of precision oncology in solid tumors. J NatlCancer Inst 2015;108. pii: djv332. doi: 10.1093/jnci/djv332.

57. Conley BA, Doroshow JH. Molecular analysis for therapy choice: NCIMATCH. Semin Oncol 2014;41:297–9.

58. Kurnit KC, Bailey AM, Zeng J, Johnson AM, Shufean MA, Brusco L, et al."Personalized cancer therapy": a publically available precision oncologyresource. Cancer Res. In press.

59. Johnson A, Khotskaya YB, Brusco L, Zeng J, Holla V, Bailey AM, et al.Clinical use of precision oncology decision support. JCO Precis Oncol2017;1:1–12.

60. McNeil C. NCI-MATCH launch highlights new trial design in precision-medicine era. J Natl Cancer Inst 2015;107. pii: djv193. doi: 10.1093/jnci/djv193.

61. Zhao Y, Polley EC, Li MC, Lih CJ, Palmisano A, SimsDJ, et al. GeneMed: aninformatics hub for the coordination of next-generation sequencing stud-ies that support precision oncology clinical trials. Cancer Informatics 2015;14:45–55.

62. ASCO TAPUR: Targeted Agent and Profiling Utilization Registry Study[homepage on the Internet]. Alexandria (VA): American Society of ClinicalOncology (ASCO); c2018. [cited 2017 Oct 7]. Available from: https://www.tapur.org.

63. Rodon J, Soria JC, Berger R, Batist G, Tsimberidou A, Bresson C, et al.Challenges in initiating and conducting personalized cancer therapy trials:perspectives from WINTHER, a worldwide innovative network (WIN)consortium trial. Ann Oncol 2015;26:1791–8.

64. Roychowdhury S, Chinnaiyan AM. Translating cancer genomes and tran-scriptomes for precision oncology. CA Cancer J Clin 2016;66:75–88.

65. Staff S, Kujala P, Karhu R, Rokman A, Ilvesaro J, Kares S, et al. Preservationof nucleic acids and tissue morphology in paraffin-embedded clinicalsamples: comparison of five molecular fixatives. J Clin Pathol 2013;66:807–10.

66. Assadi M, Lamerz J, Jarutat T, Farfsing A, Paul H, Gierke B, et al.Multiple protein analysis of formalin-fixed and paraffin-embedded tissuesamples with reverse phase protein arrays. Mol Cell Proteomics 2013;12:2615–22.

67. Rizvi NA, Hellmann MD, Snyder A, Kvistborg P, Makarov V, Havel JJ,et al. Cancer immunology. Mutational landscape determines sensitivityto PD-1 blockade in non-small cell lung cancer. Science 2015;348:124–8.

68. Kato S, Goodman A, Walavalkar V, Barkauskas DA, Sharabi A, Kurzrock R.Hyperprogressors after immunotherapy: analysis of genomic alterationsassociated with accelerated growth rate. Clin Cancer Res 2017;23:4242–50.

69. Plitas G, Konopacki C, Wu K, Bos PD, Morrow M, Putintseva EV, et al.Regulatory T cells exhibit distinct features in human breast cancer. Immu-nity 2016;45:1122–34.

70. van Nimwegen KJ, van Soest RA, Veltman JA, Nelen MR, van der Wilt GJ,Vissers LE, et al. Is the $1000 genome as near as we think? a cost analysis ofnext-generation sequencing. Clin Chem 2016;62:1458–64.

71. Trosman JR, Weldon CB, Kelley RK, Phillips KA. Challenges of coveragepolicy development for next-generation tumor sequencing panels: expertsand payers weigh in. J Nat Compr Cancer Network 2015;13:311–8.

72. Sboner A,MuXJ, GreenbaumD,AuerbachRK,GersteinMB. The real cost ofsequencing: higher than you think! Genome Biol 2011;12:125.

73. Galsky MD, Stensland KD, McBride RB, Latif A, Moshier E, Oh WK, et al.Geographic accessibility to clinical trials for advanced cancer in the UnitedStates. JAMA Intern Med 2015;175:293–5.





https://www.tapur.org

https://www.tapur.org


2018;24:2719-2731. Published OnlineFirst February 2, 2018.Clin Cancer Res Katherine C. Kurnit, Ecaterina E. Ileana Dumbrava, Beate Litzenburger, et al. Strategies for the FuturePrecision Oncology Decision Support: Current Approaches and

Updated version

10.1158/1078-0432.CCR-17-2494doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2018/02/02/1078-0432.CCR-17-2494.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/24/12/2719.full#ref-list-1

This article cites 68 articles, 31 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/24/12/2719.full#related-urls

This article has been cited by 1 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/24/12/2719To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-17-2494

http://clincancerres.aacrjournals.org/content/suppl/2018/02/02/1078-0432.CCR-17-2494.DC1

http://clincancerres.aacrjournals.org/content/24/12/2719.full#ref-list-1

http://clincancerres.aacrjournals.org/content/24/12/2719.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/24/12/2719


Documents

Precision Oncology Decision Support: Current Approaches ...ment of precision oncology decision support. Clin Cancer Res; 24(12); 2719–31. 2018 AACR. Introduction The use of targeted