1/20/2016

1

Practical Molecular Diagnostics in

Lung Cancer: Beyond the NCCN

Guidelines

Lynette M. Sholl, M.D.

Department of Pathology

Brigham and Women’s Hospital

Harvard Medical School

Boston, MA

Disclosures

• Genentech: Scientific Advisory Board

Objectives

• Established and emerging molecular targets in

lung cancer

– Applying advanced technologies in clinical practice

• Role of pathology in current lung cancer

management

– Optimizing histologic diagnoses on small biopsies

– Application of predictive immunohistochemistry

– Accurate reporting of molecular results

1/20/2016

2

Trends in lung cancer histology

Histologic Diagnoses,

SEER data 2008-2012

Meza et al. PLOSOne 2015.

Small cell, 13.1

Squamous cell,

22.8

Adenocarcinoma,

43.2

Large Cell, 2.1

NSCLC, 10.6

Carcinoma,

NOS, 3.1

Other,

5.0

Sarcoma, 0.2

• Change in smoking habits

– Decline in tobacco use since

the 1960s

– Increased use of filters

requiring more vigorous

inhalation

Reasons for increased fraction of adenocarcinoma?

Trends in lung cancer outcomes

• 27% of all cancer deaths

in 2015

• Significant downward

trends in death rates

among both men and

women diagnosed with

lung cancer between

2000-2010

Edwards et al. Cancer 2014.

1/20/2016

3

Reasons for decreasing mortality

rates?

Improved detection:

Reasons for decreasing mortality

rates?

Improved therapy:

Kris et al. JAMA 2014.

TARGETED THERAPIES

Lung adenocarcinoma

1/20/2016

4

EGFR an attractive therapeutic target

• Important pro-growth signaling protein in normal and neoplastic tissues

• Commonly overexpressed in multiple tumor types, including lung cancer

• EGFR TKIs the new wonder drugs in lung cancer??– Poor outcomes from clinical

trials of erlotinib and gefitinib in unselected patients

– Occasional patients with exceptional response to therapy

EGFR mutation is key biomarker

1/20/2016

5

EGFR mutation

…vs. copy number?

What’s the relationship?

• In an EGFR-mutated tumor:

– Step 1: mutation

– Step 2: amplification and

protein overexpression

•Amplification/overexpression

may occur in the absence of

mutation.

ALK translocations in NSCLC:

3 years (!) from bench to bedside

1/20/2016

6

CAP/AMP/IASLC

Clinical Practice Guidelines

• Any advanced stage (IV) patient with ACA or patient with progressive disease should receive EGFR and ALK testing– Reflex testing may be appropriate

in certain environments

– This testing is not recommended for lung tumors lacking evidence of ACA differentiation

– Avoid the term “non small cell lung carcinoma” whenever possible

• Prioritize tissue for EGFR and ALK testing

ROS1 rearrangements in lung cancer

•Rare (1-2% of lung

ACA)

•More common in

never smokers

•Promising responses to

crizotinib therapy

Shaw AT et al. N Engl J Med 2014;371:1963-1971.

1/20/2016

7

RET rearrangements in lung cancer

• Rare (1% of lung ACA)

• Results from small intrachromosomalrearrangement

(inv (10)(p11.22q11.2)

• More common in never smokers

• Reported responses to:– cabozantinib

• (c-Met, VEGFR2 inhibitor)

– Vandetanib • (VEGFR, EGFR, RET inhibitor)

Takeuchi et al Nat Med 2012

67 year old woman with lung mass and

numerous subcutaneous metastases:

EGFR, ALK, ROS1 wild type

Sequencing reveals METex14 mutation &

MET amplification with Exon 14 skipping

Efficacy of crizotinib in MET ex 14 splice tumors

1/20/2016

8

• 64yo M, ~15 pack-year smoker, quit 1 year prior to diagnosis

• May 2014: diagnosed with stage IV lung adenocarcinoma

• Standard testing negative for EGFR, ALK, ROS1

• Started on first-line carboplatin/pemetrexed, with mixed response

• BRAF c.1799T>A (p.V600E) detected by sequencing

Dabrafenib (BRAF inhibitor) +

Trametinib (MEK inhibitor)*

October 2014 March 2015

*60% overall response rate in the phase 2 trial for patients with BRAF V600E mutations.

Planchard et al. J Clin Oncol 33, 2015 (suppl; abstr 8006)

KRAS

33.8

EGFR

19.1ALK

3.9

BRAF

3.8MET

3.0

PIK3CA

2.9

ERBB2

2.5NRAS

1.0

ROS1

1.1

RET

1.0

AKT1

0.6

MAP2K1

0.3

HRAS

0.1

No/unknown driver

29.9

Oncogenic drivers in

non-squamous NSCLC (%)

1/20/2016

9

Reality check.

“Non small cell lung carcinoma”

Adenocarcinoma Squamous cell carcinoma

Most common subtype in

nonsmokers

Unique chemosensitivity

profile (pemetrexed)

~60% have a defined oncogenic

driver

-“Targetable”

Smokers

Use of antiangiogenic agents

associated with massive

pulmonary hemorrhage

Minority with defined

oncogenic driver

-Limited targetability

1/20/2016

10

IASLC algorithm for Small Biopsies

Subclassification of morphologic

NSCLC-NOS

IASLC classification of small biopsies,

take home points:

• Distinguish ACA and SQC whenever possible

• The molecular profile of an ACA will dictate

targeted therapy

• Judicious use of IHC is critical

– Two first-line markers:

• TTF1 and p63 or p40

• Less established/less specific markers (napsin, mucin,

CK7, CK5/6) should be considered second line

1/20/2016

11

Make a diagnosis

+

EGFR, KRAS, HER2, BRAF, PIK3CA,

ALK, ROS1, RET, MET, etc. mutations,

copy number alterations, translocations

EGFR mutation-specific

antibodies

(clones 43B2 and 6B6)

Study L858REx19del

Sensitivity Specificity Sensitivity Specificity

Yu et al.92%

(24 of 26)99%

(193 of 195)86%

(23 of 26)100%

(196 of 196)

Kawahara et al.

83%(19 of 23)

100%(16 of 16)

62%(13 of 21)

100%(16 of 16)

Kato et al. 75%

(9 of 12)97%

(56 of 58)50%

(9 of 18)100%

(52 of 52)

Brevet et al.95%

(20 of 21)99%

(171 of 173)74%

(23 of 31)99%

(161 of 163)

Fan et al.93%

(40 of 43)100%

(126 of 126)74%

(17 of 23)99%

(145 of 146)

Bondgaard et al.

80%(8 of 10)

98%(152 of 155)

63%(12 of 19)

99%(153 of 155)

Excellent specificity, limited sensitivity

Informative if positive- may be useful in

scant specimens

Cutz et al. JTO 2014

ALK IHC

93-100% sensitive and specific as compared to FISH

Ventana ALK (D5F3) CDx assay approved as a companion diagnostic for crizotinib

1/20/2016

12

ROS1 IHC

(D4D6 antibody)Ideal screening tool:

Excellent sensitivity,

good specificity

Reference N Sensitivity Specificity Notes

Sholl et al. 2013 210 100 92 Focal expression in a KRAS-

mutated tumor

Strong expression in a FISH

negative tumor

Mescam-Mancini

et al. 2014

221 100 96.9 Expressed in two HER2-

mutated tumors

Cha et al. 2014 330 100 72.6-93.4* ROS1 expression seen in ROS1

WT tumors from ever-smokers

Boyle et al. 2014 33 100 100 As compared to FISH or RT-PCR

Sholl et al. AJSP 2013.

Reference Clone IHC pattern vs.RET

N

WT

NSensitivity Specificity

Tsuta et al.

2014EPR2871 Any

FISH or

RTPCR21 1774 66.7 86.1

Lee et al. 2014 ab134100 Any

Transcript

profiling or

FISH

15 79 100 87.3

Sasaki et al.

2014

3F8 Any RTPCR 3 154 100 70.3

EPR2871 Any RTPCR 3 75 100 33.3

EPR2871Mod to

strong onlyRTPCR 3 75 66.7 77.3

RET IHC• Variable antibody

performance

• Current evidence doesn’t

support routine clinical useTsuta et al. Br. J. Cancer 2014

How has next gen sequencing (NGS)

changed the game?

• Increased target coverage with decreased

tissue requirements

ABL1 BMPR1A CDKN1B EP300 FAS HRAS MDM2 NFKBIZ PRAME RPL26 STAG2

AKT1 BRAF CDKN1C EPHA3 FBXW7 ID3 MDM4 NKX2-1 PRDM1 RUNX1 STAT3

AKT2 BRCA1 CDKN2A EPHA5 FGFR1 IDH1 MECOM NOTCH1 PRF1 SBDS STAT6

AKT3 BRCA2 CDKN2B EPHA7 FGFR2 IDH2 MEF2B NOTCH2 PRKAR1A SDHA STK11

ALK BRD4 CDKN2C ERBB2 FGFR3 IGF1R MEN1 NPM1 PRKCI SDHAF2 SUFU

ALOX12B BRIP1 CEBPA ERBB3 FGFR4 IKZF1 MET NPRL2 PRKCZ SDHB SUZ12

APC BUB1B CHEK2 ERBB4 FH IKZF3 MITF NPRL3 PRKDC SDHC SYK

AR CADM2 CIITA ERCC2 FKBP9 INSIG1 MLH1 NRAS PRPF40B SDHD TCF3

ARAF CARD11 CREBBP ERCC3 FLCN JAK2 MLL NTRK1 PRPF8 SETBP1 TCF7L1

ARID1A CBL CRKL ERCC4 FLT1 JAK3 MLL2 NTRK2 PSMD13 SETD2 TCF7L2

ARID1B CBLB CRLF2 ERCC5 FLT3 KCNIP1 MPL NTRK3 PTCH1 SF1 TERC

ARID2 CCND1 CRTC1 ESR1 FLT4 KDM5C MSH2 PALB2 PTEN SF3B1 TERT

ASXL1 CCND2 CRTC2 ETV1 FUS KDM6A MSH6 PARK2 PTK2 SH2B3 TET2

ATM CCND3 CSF1R ETV4 GATA3 KDM6B MTOR PAX5 PTPN11 SLITRK6 TLR4

ATRX CCNE1 CSF3R ETV5 GATA4 KDR MUTYH PBRM1 PTPRD SMAD2 TNFAIP3

AURKA CD274 CTNNB1 ETV6 GATA6 KEAP1 MYB PDCD1LG2 QKI SMAD4 TP53

AURKB CD58 CUX1 EWSR1 GLI1 KIT MYBL1 PDGFRA RAD21 SMARCA4 TSC1

AXL CD79B CYLD EXT1 GLI2 KRAS MYC PDGFRB RAF1 SMARCB1 TSC2

B2M CDC73 DDB2 EXT2 GLI3 LINC00894 MYCL1 PHF6 RARA SMC1A U2AF1

BAP1 CDH1 DDR2 EZH2 GNA11 LMO1 MYCN PHOX2B RB1 SMC3 VHL

BCL2 CDK1 DEPDC5 FAM46C GNAQ LMO2 MYD88 PIK3C2B RBL2 SMO WRN

BCL2L1 CDK2 DICER1 FANCA GNAS LMO3 NBN PIK3CA RECQL4 SOCS1 WT1

BCL2L12 CDK4 DIS3 FANCC GNB2L1 MAP2K1 NEGR1 PIK3R1 REL SOX2 XPA

BCL6 CDK5 DMD FANCD2 GPC3 MAP2K4 NF1 PIM1 RET SOX9 XPC

BCOR CDK6 DNMT3A FANCE GSTM5 MAP3K1 NF2 PMS1 RFWD2 SQSTM1 XPO1

BCORL1 CDK9 EED FANCF H3F3A MAPK1 NFE2L2 PMS2 RHEB SRC ZNF217

BLM CDKN1A EGFR FANCG HNF1A MCL1 NFKBIA PNRC1 RHPN2 SRSF2 ZNF708

ROS1 STAG1 ZRSR2

POPv2 GENE LIST

SNV, INDELS, COPY NUMBER ALTERATIONS

1/20/2016

13

How has next gen sequencing (NGS)

changed the game?• Increased target coverage with

decreased tissue requirements

• Less biased genomic analysis– Detect novel variants in known

oncogenes

– Detect variants outside of coding regions

– Detect variants in genes not typically included on clinical test menus

– Detect common variants in known oncogenes in unusual contexts

NGS vs. IHC vs. FISH for

ALK translocation detection

ALK translocations in Lung Adenocarcinoma

Clinical FISH and/or IHC

ALK + ALK -

Oncopanel

ALK

Fusion + 25 0 Sensitivity

96%

Specificity

100% ALK

Fusion - 1* 190

Total 26 190

* 20% tumor in this specimen

NGS as an arbiter in discrepant cases:

ALK IHC (5A4 clone)

FISH negative IHC positive

1/20/2016

14

Fused Green onlyRed only

NGS showed: CLIP4-ALK fusion AND EML4-ALK fusion

From Vysis ALK FISH probe product insert

CLIP4

FISH cannot detect cryptic EML4-ALK fusions.

Unusual FISH results are considered “negative” but should be followed up.

MET in tumorigenesis

Lai AZ, et al, Trends in Cell Bio, 2009

• Exon 14 deletion removes the juxtamembranedomain of MET

• Tyr1003 phosphorylation site necessary for Cbl binding

• Decreased ubiquitination and impaired downregulation of the activated receptor

Identification of tumor-specific, intronic mutations in Met leading to exon 14 splicing.

Monica Kong-Beltran et al. Cancer Res 2006;66:283-289

©2006 by American Association for Cancer Research

1/20/2016

15

MET: Diverse Deletions & Point Mutations

Awad et al. J Clin Oncol, 2016.

MET juxtamembrane splice mutations:

not uncommon in lung adenocarcinomas

and predict response to MET inhibitors

• 2.4% in Kong-Beltran et al. Cancer Res, 2006

• 3.3% in Onazato et al. J Thor Oncol, 2009

• ~4.5% in TCGA Nature, 2014

• 3% in Frampton et al. Cancer Discovery, 2015

• 4% in Paik et al. Cancer Discovery, 2015

• 3% in Awad et al. J Clin Oncol, 2016

Clinical CharacteristicsClinical Characteristic

MET ex14

(N = 28)

EGFR

(N = 99)

KRAS

(N = 169)

Median age (range), years 72.5 (59-84) 61 (30-93) 65 (42-93)

Sex, N (%)

Male

Female

9 (32%)

19 (68%)

30 (30%)

69 (70%)

62 (37%)

107 (63%)

Smoking history, N (%)‡

Never-smoker

≤10 pack-years

>10 pack-years

10 (36%)

3 (11%)

15 (53%)

57 (58%)

10 (10%)

28 (28%)

6 (4%)

11 (7%)

152 (90%)

Race, N (%)

White, non-Hispanic

Asian

Black

White, Hispanic

Unknown

28 (100%)

0 (0%)

0 (0%)

0 (0%)

0 (0%)

79 (80%)

15 (15%)

1 (1%)

3 (3%)

1 (1%)

157 (93%)

0 (0%)

5 (3%)

3 (2%)

4 (2%)

Histology, N (%)

Adenocarcinoma

Pleomorphic w/ adenocarcinoma component

NSCLC, poorly-differentiated

Squamous

Adenosquamous

18 (64%)

4 (14%)

5 (18%)

0 (0%)

1 (4%)

92 (93%)

0 (0%)

4 (4%)

2 (2%)

1 (1%)

150 (89%)

3 (2%)

10 (6%)

5 (3%)

1 (1%)

Stage at diagnosis, N (%)

I

II

III

IV

13 (46%)

2 (7%)

4 (14%)

9 (32%)

9 (9%)

3 (3%)

9 (9%)

78 (79%)

12 (7%)

12 (7%)

44 (26%)

101 (60%)

Awad et al. J Clin Oncol 2016.

1/20/2016

16

Pulmonary sarcomatoid carcinoma

• Sarcomatoid carcinoma– Pleomorphic carcinoma

• Giant cells or spindled cells comprise ≥ 10% of tumor

• Admixed adenocarcinoma, squamous, or undifferentiated carcinoma

– Carcinosarcoma• Carcinoma + sarcoma with

heterologous elements

– Pulmonary blastoma• Fetal adenocarcinoma+

mesenchymal stroma

• 2-3% of lung cancers

• Predominantly smokers

MET mutations in pulmonary

pleomorphic/sarcomatoid carcinoma

• Lui et al, JCO 2015:

• MET exon 14 splice

mutations in 8/36 (22%)

of pulmonary

sarcomatoid carcinomas

undergoing WES and

targeted MET analysis

• BWH/DFCI:

• MET exon 14 splice

mutations in 4/15 (27%)

pulmonary pleomorphic

carcinoma sequenced by

NGS

• Other alterations include:

– KRAS (13%)

– NRAS (7%)

– no known driver (53%).

Potential for NGS in diagnosis

• Distinguishing multiple primary lung tumors

from metastases

• Defining site of origin for poorly differentiated

carcinomas

1/20/2016

17

RUL

RLL

EGFR c.2573T>G (p.L858R), in 2.8% of 285 reads

KRAS c.35G>A (p.G12D), in 15% of 438 reads

MET c.3028+2T>C (p.D1010_splice) - in 34% of 449 reads

74 year old woman, former smoker,

incidental RUL mass and multiple GGOs

Breast vs. Lung?

• 58 year old woman, h/o

well differentiated

T1aN0 invasive ductal

carcinoma (ER+, PR+,

HER2-) 5 years prior

• Now with lung left

upper lobe mass and

diaphragmatic implants Lung tumor: CK7+, TTF1-, GATA3 weak,

ER weak, mammoglobin -

“The carcinoma in the lung is poorly differentiated with high grade nuclei, abundant

cytoplasm, mucin production, and necrosis. The carcinoma does not resemble the

breast carcinoma in the excision from 2010. The possibility that the patient has a

different breast primary carcinoma should be considered.”

Back of the envelope:

Gene Frequency

in breast

Frequency

in Lung

PPV

KRAS

mutation

1% 25% 98%

(for lung)

EGFR

mutation

0 13% 100%

(for lung)

ERBB2

amp

16% 0.5% 91%

(for breast)

1/20/2016

18

Breast vs. Lung?

KRAS c.34G>T (p.G12C), exon 1

STK11 c.206C>A (p.S69*), exon 1

TP53 c.499_501CAG>G (p.Q167fs), exon 5

KEAP1 c.1016T>C (p.L339P), exon 3

KEAP1 c.382A>T (p.I128F), exon 2

JAK2 c.3214C>T (p.Q1072*), exon 24

BRIP1 c.434C>G (p.S145C), exon 5

CDC73 c.26G>T (p.R9L), exon 1

CHEK2 c.349A>G (p.R117G), exon 3

DMD c.10567G>A (p.E3523K), exon 75

JAK2 c.3214C>T (p.Q1072*), exon 24

SMARCA4 c.2439_splice (p.S813_splice)

TCF3 c.136G>A (p.G46R), exon 3

Dogan et al. Clin Cancer Res, 2012.

Calles et al. Clin Cancer Res, 2015.

Genomic evidence supports a lung primary.

NGS to explore new targets

KRAS

33.8

EGFR

19.1ALK

3.9

BRAF

3.8MET

3.0

PIK3CA

2.9

ERBB2

2.5

NRAS

1.0

ROS1

1.1

RET

1.0

AKT1

0.6

MAP2K1

0.3

HRAS

0.1

No/unknown

driver

29.9

Jamie Wyeth

“Targeted therapy is dead.”

-Anonymous

PD-1/PD-L1 inhibitors

• Applicable across most

tumor types

• 20-30% response rate on

average

• 18 month response

durability

Mutation-targeted inhibitors

• Applicable to few tumors

(lung, melanoma, GIST)

• 50-70% response rate

• Response durability 7-12

months

From a population standpoint, immune checkpoint blockade outperforms

mutation-targeted therapies.

1/20/2016

19

PD-L1 IHC in practice

nivolumab

pembrolizumab

tremelimumab

atezolizumab

E1L3N

??

Dako 28-8

22C3 pharmDx

Roche SP263

Roche SP142

PD-L1 IHC: mass confusion.

Perhaps other biomarkers will be better…

Characteristics of immune infiltrate?

Expression of other checkpoint inhibitors?

Mutational load?

Mutational signature?

Drug Company FDA approval mAb/Platform Scoring criteria Comment

Pembroluzimab

(Keytruda)

Merck FDA approved for

NSCLC

22C3 (DAKO

pharmDx)/

Link 48 Autostainer

≥50% tumor cells Companion

diagnostic1(as of

Oct 2015)

Nivolumab

(Opdivo)

Bristol- Myers

Squibb

FDA approved for

squamous and non

squamous NSCLC

28-8 (DAKO

pharmDx)/

Link 48 Autostainer

≥1% tumor cells Complementary

diagnostic (as of

Oct 2015)

Predictive only in

non-squamous

carcinomas

Atezolizumab

(MPDL3280)

Roche Expected in 2016 SP142 (Ventana) Tumor cells and/or

tumor infiltrating

immune cells

In development

Durvalumab

(MEDI4736)

Astra Zeneca Expected in 2016 SP263 (Ventana) ≥25% tumor cells In development

Biomarker reporting

• http://www.cap.org/web/home/resources/cancer-

reporting-tools/cancer-protocol-templates

1/20/2016

20

Biomarker reporting

• Optional reporting template

• Available as CAP electronic cancer checklist (eCC)

• Focused on targets with accepted clinical significance– Alteration type

– Methodology

• Updated version available in 2016 covers:– EGFR

– KRAS

– BRAF

– ERBB2

– ALK

– RET

– ROS1

– MET

Important topics given short shrift:

• Tumor types beyond adenocarcinoma

• Mechanisms and significance of acquired resistance in patients receiving EGFR TKIs, MET/ALK/ROS1 inhibitors

• “Liquid biopsy” in solid tumor clinical management

Sequist et al. Sci Trans Med. 2011

Mechanisms of EGFR TKI resistance

Thanks… Questions?

Acknowledgements:

Mark Awad, DFCI

Phil Cagle

Pasi Janne, DFCI

Neal Lindeman, BWH

Geoff Oxnard, DFCI

Members of the

Pulmonary Pathology

Society